Support this Pinball Repair Website & PHoF. Please purchase the Marvin3m.com/top

Scope.
This document is a repair guide for Bally electronic pinball games made from 1977 to 1985. Since Stern electronic pinball games use nearly identical electronics, these games are covered too.

Internet Availability of this Document.
Updates of this document are available for no cost at http://marvin3m.com/fix.htm if you have Internet access. This document is part three of three (part one is here, and part two is here).

IMPORTANT: Before Starting!
IF YOU HAVE NO EXPERIENCE IN CIRCUIT BOARD REPAIR, YOU SHOULD NOT TRY TO FIX YOUR OWN PINBALL GAME! Before you start any pinball circuit board repair, review the document at http://marvin3m.com/begin, which goes over the basics of circuit board repair. Since these pinball repair documents have been available, repair facilities are reporting a dramatic increase in the number of ruined ("hacked") circuit boards sent in for repair. Most repair facilities will NOT repair your circuit board after it has been unsuccessfully repaired ("hacked").

If you aren't up to repairing pinball circuit boards yourself or need pinball parts or just want to buy a restored game, I recommend seeing the suggested parts & repair sources web page.

Table of Contents

1. Getting Started:
1. Experience, Schematics, Manuals
2. Necessary Tools
3. Parts to have On-Hand
4. Game List
5. Lubrication Notes
6. The Circuit Boards
7. Voltage Test Points on the Boards
8. Fuse Values/Usage & Power Supply Power Distribution

2. Before Turning the Game On:

1. Check the Coil Resistance
2. Removing the MPU Battery and Fixing Corrosion
3. Rebuilding the Power Supply
4. Upgrading the Voltage Regulator/Solenoid Driver Board
5. Upgrading the Ground on the MPU Board
6. Pre Power-On Voltage Checks
7. Connectors
8. Setting Free Play

3. When Things Don't Work:

1. Making a MPU test fixture
2. Fixing the MPU (LED flashes and the such)
3. Test EPROM & Fixing a Dead MPU board
4. Game ROMs, EPROMs, and Jumper - the Basics
5. -35 MPU Game ROMs, EPROMs, and Jumpers
6. -17 MPU Game ROMs, EPROMs, and Jumpers
7. Stern M-200 MPU Jumpers (using a Stern M-200 in a Bally Game)
8. Converting a -133 MPU (Baby Pacman) to a -35 MPU
9. The Built-In Diagnostics/Bookkeeping
10. Coils: Locked-on, Sporadic, or Not Working (Solenoid Driver board) including Flipper Coils
11. Lamps: Locked-on or Not Working Feature Lamps (Lamp Driver board)
12. Solenoid Expander Board Problems
13. Auxiliary Lamp Driver Board Problems
14. Switches and the Switch Matrix
15. High Voltage Section Problems (No Score Display Power)
16. Score Display Fixes and Replacements
17. Reset or Lockup Power Problems
18. Sound Board Problems
19. Flipper Maintenance
20. Miscellaneous Problems and Fixes

Important: Before you Change any Jumpers!!
It is EXTREMELY important that you have a working MPU board before you change any jumper locations! If the MPU board currently has ROMs in it, get it working first before playing with the jumpers. If the MPU board is jumpered incorrectly for the game ROMs installed, the diagnostic LED light will stay on, and the board will not power-up. So it is absolutely important that the jumpers are correct for the ROMs installed. Get the MPU board working first before proceeding.

Also of course, make sure the board being worked on is a AS-2518-35 (or AS-2518-133) MPU (remember the -133 is really a -35 board with R113 changed to a diode CR52). The easiest way to tell (besides looking at the silkscreened part number on the board!) is to examine connector J5. On a -35/-133 MPU board, J5 will have 33 pins (including the removed "key" pin).

Max out ROM Memory on the -35/-133 MPU to use 2732 EPROMs.
Jumpering is not a big problem until the MPU's ROM chips begin to fail, and MPU boards are shifted from game to game. An MPU board can only be jumpered so many times before the traces and jumpers start to lift and strip off the board. So instead of custom jumpering a board to a particular set of ROMs, the best idea is to maximize the board to use the largest EPROM size, and cater the EPROMs themselves to the board (instead of the other way around).

This really makes sense as the largest EPROMs that will fit the -35/-133 MPU board are 2732's, which are commonly and cheaply available. For this reason, the Bally ROM code has been re-formated to fit this size EPROM. The original program size and code is still available from Williams at www.pinball.com/Williams/tech/roms.html. But I highly suggest you down load the ZIP file bly2732.zip instead, as it contains all the Bally ROMs for all games from Freedom (1977) to Cybernaut (1985), and has been converted to 2732 format. Only Baby Pacman and Granny and the Gators ROMs are missing from this file. Using these files will allow you to use a -35 MPU board for ANY Bally game from 1977 to 1985. Click here for a list of the ROM files, and which games use which ROMs (note many games share the same U6 ROM). Text file from Mr. Calahan.

For the -35/-133 MPU board, using all 2732 for ANY game has the advantage of just one set of jumpers. Just download the above file and expand it, and burn your game into 2732 EPROMs. Then jumper your -35 MPU board like this:
1. Jump E4 to E13a
2. Jump E12 to GND (large GND trace next to the ROM sockets, left of E12)
3. Jump E7 to E8
4. Jump E10 to E11
5. Jump E31 to E32
6. Jump E16a to E29 (on -133 "E16 to E29")
7. Jump E33 to E35
Note: any jumpers still installed (from the old ROM setup) that are not listed above should be removed.

Jumpers around the U1 socket, for using two 2732 EPROMs: Jumpers E4 to E13a (top white wire), E12 to GND (right white wire), E7 to E8 (middle left), and E10 to E11 (lower right).

Jumpers around the U6 socket, for using two 2732 EPROMs: E31 to E32, E16a to E29, and E33 to E35.

If you don't want to convert your game to 2732 format, you can also use these jumpers for other types of ROMs in your -35/-133 MPU board.

Using 2532 EPROMs instead of 9332 Masked ROMs.
If the game in question is a later Bally game with 9332 masked ROMs, these can be changed to 2532 EPROMs with NO jumper modifications! This can be handy and convenient if the original black 9332 masked ROMs need changing, but the repair person doesn't want to mess with the jumpers.

The "Mother" Source for Jumper Info.
All the following jumper information came from several sources. The primary source is, of course, the original MPU board schematics. There were also two secondary sources too. First was Williams web site, located at bally_read1st.txt which contains all the jumper info you will probably need. Also Joel Cook's "Pinball Lizard" Tech Tips Guide #1 has this information too. I highly suggest the Joel Cook "lizard" tech books; they have lots of other good pinball repair information too. You can buy these from Marco Specialties at http://www.pinballmachine.com.

Stuff to remember:
• Bally uses a preceding "E" on all jumper numbers. Yes, the "E" has been left out above to save some space in the chart.
• The "dash" between the numbers is the "jump". That is, "1-4" means a jumper from E1 to E4.
• Remove any jumpers not shown above for a given configuration. If it's not mentioned above for your ROM set up, you don't need that jumper!
• Don't trust other Bally jumper charts! The above chart is "the one to use" (using Bally published jumper charts can lead to problems).
• You must know the ROM device type installed at each ROM location. The Bally part number (often printed on the ROM) does not help.
• BLACK masked ROMs, as used in many Bally games, are entirely black and usually have some white part numbers printed on them. These are known as 9316 masked ROMs.
• EPROMs, on the other hand, have a small clear "window" on their top, often with a sticker over the window. The sticker is there for a reason; it prevents UV light from entering the EPROM's clear window (this is how an EPROM is erased! so keep the sticker on the window). EPROMs are labeled as to their size (i.e. "2716").
• The most common EPROMs used on Bally MPU boards are 2716, 2532 and 2732 EPROMs.
• ROMs and EPROMs are game specific. Each game has its own custom set of ROM computer code, stored on that game's ROMs (or EPROMs).
• Some EPROM part numbers are interchangable. For example, 2532 EPROMs, 9332 masked ROMs, and 8332 masked ROMs all use the same jumper setting. But a U2 2532 EPROM from a Kiss game is NOT interchangable with a U2 9332 EPROM from a Strikes and Spares! This also applies to the 16 bit masked ROMs too. That is, 9316 masked ROMs, 8316 masked ROMs, and 8516 masked ROMs all use the same jumper settings. But again a U2 9316 ROM from a Kiss game is NOT interchangable with a U2 8316 ROM from a Strikes and Spares!

Freedom and Night Rider ROMs and Jumper Notes if using U1,U6.
These two games used a strange set of ROMs at U1 and U2. These are 74S474 or 7461 (512 byte) ROMs at U1 and U2, and a 9316 or 2716 (2K byte) at U6. The old Williams tech web site at www.pinball.com states that a U1 2716 EPROM and a U6 2716 EPROM can be used for these two games (and provides the ROM files for download, and the jumper settings for the -35 MPU board).

Note I have also tested both games with 2732 EPROMs at U2 and U6 on a converted -17 and -35 MPU boards (as documented above). This does in fact work fine for both Freedom and Night Rider, and is a better solution in my opinion.

E13, E15 Mistaken Jumper Locations.
There are two jumper pads near the lower right hand corner of the U2 ROM socket labeled "E13" and "E15". There are also two vias (plated through holes) just a little bit further to the right, which are actually closer to the "E13" and "E15" labels. The vias are completely unrelated to the labeled jumper pads. Be careful when using these jumpers that you don't confuse the vias with the jumper pads. They are both round plated through holes, but the jumper pads are a bit bigger.

Important: Before you Change any Jumpers!!
It is EXTREMELY important that you have a working MPU board before you change any jumper locations! If the MPU board currently has ROMs in it, get it working first before playing with the jumpers. If the MPU board is jumpered incorrectly for the game ROMs installed, the diagnostic LED light will stay on, and the board will not power-up. So it is absolutely important that the jumpers are correct for the ROMs installed. Get the MPU board working first before proceeding. If the ROMs are suspect as bad, and the MPU board is set up for 9316 ROMs (most -17 boards are), see the section below, "Making an Adapter to use Two 2716 EPROMs in an Unmodified -17 MPU board that is Jumpered for 9316 ROMs". This will allow the use of two new 2716 EPROMs to replace the failed 9316 ROMs, without any MPU board modifications or new jumper settings.

Also of course, make sure the board being worked on is a AS-2518-17 MPU. The easiest way to tell (besides looking at the silkscreened part number on the board!) is to examine connector J5. On a -17 MPU MPU board, this connector will have 32 pins (including the removed "key" pin).

Bad MPU Board Sockets ("closed frame" sockets).
If a Bally MPU board is using brown or black closed frame sockets (closed frame means the circuit board under the socket can not be seen), or sockets that have "SCANBE" or "RS" impressed on them, it is advised they be changed! These older sockets are very troublesome and cause many intermittent problems. A quick alternative to replacement is to plug a machine pin socket into the brown socket. This is a *temporary* fix, but should work well enough until the board is working, and then later replace the sockets).

A brown and a black closed frame socket. There's no way around it, ALL these sockets will need to be replaced on any Bally MPU board.

Jumpers used on the Early -17 MPU (and Stern M-100).
The 1977 to 1979 Bally -17 MPU boards aren't as versatile as the newer -35 boards. They have limited ROM space, which means they can't be used in the newer 1979 to 1985 games. This can all be rectified, but will require some cutting and jumping of traces on the -17 MPU board.

There are only a few jumper choices on the -17 MPU board. The following jumpers only apply for the early 1977 to 1979 Bally -17 MPU board. Note the configuration that uses a U1 ROM only existed for the first two Bally games, Freedom and Night Rider.

Stuff to remember:

• Bally uses a preceding "E" on all jumper numbers. Yes, the "E" has been left out above to save some space in the chart.
• The "dash" between the numbers is the "jump". That is, "1-4" means a jumper from E1 to E4.
• Remove any jumpers not shown above for a given configuration. If it's not mentioned above for your ROM set up, you don't need that jumper!
• You must know the ROM device type installed at each ROM location. The Bally part number (often printed on the ROM) does not help.
• BLACK masked ROMs, as used in many Bally games, are entirely black and usually have some white part numbers printed on them. These are known as 9316 masked ROMs (though Freedom and Night Rider can also use 74S474 or 7461 masked ROMs).
• EPROMs, on the other hand, have a small clear "window" on their top, often with a sticker over the window. The sticker is there for a reason; it prevents UV light from entering the EPROM's clear window (this is how an EPROM is erased! so keep the sticker on the window). EPROMs are labeled as to their size (i.e. "2716").
• ROMs and EPROMs are game specific. Each game has its own custom set of ROM computer code, stored on that game's ROMs (or EPROMs).
• Some EPROM part numbers are interchangable. For example, 2532=9332=8332. But a U2 2532 EPROM from a Kiss game is NOT interchangable with a U2 9332 EPROM from a Strikes and Spares!

Mod1: 2716 at U2,U6 on a Bally -17 or Stern M-100.
**Freedom/Night Rider with 2716 at U2/U6.
The most common conversion on a -17 or M100 CPU board (other than converting a -17 board to use 2732 EPROMs). In addition to the jumpers listed above (1-2, 3-4, 6-7, 8-10), must also make the following cuts and jumps to use 2716 EPROMs at U2/U6.

1. Make sure jumpers E1-E2, E3-E4, E6-E7, and E8-E10 are in place.
2. On the solder side of the board, cut the trace leading to U18 pin 4.
3. On the solder side of the board, run a jumper from the via above U3 pin 6 (that had its trace cut in the prior step) to U18 pin 5. This connects U18 pin 5 to U2 pin 18.
4. On the solder side of the board, cut the trace leading to U2 pin 21.
5. On the solder side of the board, cut the trace leading to U6 pin 21.
6. On the solder side of the board, run a jumper from U2 pin 21 to U2 pin 24.
7. On the solder side of the board, run a jumper from U6 pin 21 to U6 pin 24.
8. For Freedom and Night Rider, be sure to put the 2716 U1 chip in the U2 socket.

Mod1/2: When using 2716s at U2 & U6, or a single 2732 at U2, make these modifications at chip U18: cut the trace going to U18 pin 4, and connect U18 pin 5 to the cut trace's via (which is above U3 pin 6).

Mod2: The -17/M100 component side modifications for a single 2732 at U2. Make the trace cut shown next to U2's large ground trace, and then connect E4 to the via above the cut trace, and then E3 to the large ground trace.

Mod3: 2716 at U2, 9316 at U6 on a Bally -17 or Stern M-100 MPU board.
In addition to the jumpers listed above (1-2, 3-4, 6-7, 8-10), must also make the following cuts and jumps to use this configuration.
1. On the componet side, cut the trace from U2 pin 18 to U3 pin 18. Best place to do this is where the trace passes between sockets U2 and U3. Use your DMM set to continuity to help figure out the trace to cut.
2. On the solder side, run a jumper from U2 pin 18 to U17 pin 11.
3. On the solder side, cut the trace going to U2 pin 21.
4. On the solder side, run a jumper from U2 pin 21 to U2 pin 24.

Mod4: 2732 EPROMs at U2,U6 on a Bally -17 or Stern M-100 board.
AKA Converting a Bally -17 or Stern M-100 to a Bally -35 MPU.
This modification allows a Bally -17 or Stern M-100 MPU board to be used on any Bally game up to 1985. It doubles the amount of ROM space the older MPU board can use, and essentially makes a Bally -17 or Stern M-100 MPU board a Bally -35 MPU board. Note you can not use a -35 Bally board in a Stern games requiring a M-200 MPU (these boards have two 5101 RAMs instead of one as used on a Bally -35 MPU).

In addition to the jumpers listed above (1-2, 3-5, 6-7, 8-10), you must also make the following cuts and jumps to use this configuration.

1. Make sure jumpers E1-E2, E3-E5, E6-E7, and E8-E10 are in place.
2. Double check jumper E3 connects to E5.
3. On the solder side, cut the trace that runs to U2 pin 21.
4. On the component side, cut the trace that runs to U2 pin 18. Best place to do this is where the trace passes between sockets U2 and U3. Use your DMM set to continuity to help figure out the trace to cut.
5. On the solder side, jump a wire from U2 pin 18 to U2 pin 12.
6. On the soider side, jump a wire from U2 pin 21 to U9 (CPU) pin 24.
7. On the solder side, cut the trace that runs to U6 pin 21.
8. On the component side, cut the trace that runs to U6 pin 18. Best place to do this is where the trace passes between sockets U6 and U5. Use your DMM set to continuity to help figure out the trace to cut.
9. On the solder side, jump a wire from U6 pin 18 to U6 pin 12.
10. On the solder side, jump a wire from U6 pin 21 to U2 pin 21 (this connects both U2 and U6 pins 21 to U9 pin 24).
11. On the solder side, cut the trace that runs to U17 pin 2.
12. On the solder side, cut the trace that runs to U18 pin 4.
13. On the solder side, jump a wire U17 pin 2 to U18 pin 4.

Making an Adapter to use Two 2716 EPROMs in an Unmodified -17 MPU board that is Jumpered for 9316 ROMs.
This is a great adapter to have around when working on a -17 board that you don't want to modify. It will allow you to use a pair of 2716 EPROMs on a stock, unmodified -17 MPU board. This allows testing of the board without making any cuts or jumps.

Basically it takes four good quality (machine pin) 24 pin sockets, and sandwiches two of them together. There are a couple pins that need to be cut, and a couple jumper wires added (see the diagram below):

• On two of the 24 pin sockets, jump pins 21 and 24 together with some wire.
• On the two sockets you modified above, solder a four inch wire to pin 18. Then solder the sockets' pin 18 wires together, and attach a test clip to this wire too.
• On the two sockets you modified above, cut pins 18 and 21 short so they won't plug into anything.
• Plug the two modified sockets into the other two unmodified sockets. Make sure pins 18 and 21 do not contact the unmodified sockets. To be double sure this happens, you can cut pins 18 and 21 off the bottom sockets too.
• Plug the 2716 EPROMs for U2 and U6 into the two modified sockets.
• Plug the sandwiched sockets and EPROMs into the MPU board at positions U2 and U6.
• Connect the test lead coming off pin 18 of the two modified sockets to the right side of R14 (the side nearest the ROM sockets).

Freedom and Night Rider ROMs and Jumper Notes if using U1,U6.
These two games used a strange set of ROMs at U1 and U2. These are 74S474 or 7461 (512 byte) ROMs at U1 and U2, and a 9316 or 2716 (2K byte) at U6. The old Williams tech web site at www.pinball.com states that a U1 2716 EPROM and a U6 2716 EPROM can be used for these two games (and provides the ROM files for download, and the jumper settings for the -35 MPU board).

If using a -17 MPU with a U1 2716 EPROM and U6 2716 EPROM, there are some other cuts and jumps required:

• Cut the trace from U1 pin 18 to U2 pin 18.
• Cut the trace from U1 pin 21 to jumper pad E7.
• Connect U1 pin 21 to U1 pin 24.
• Connect U1 pin 18 to U17 pin 11.
• Connect U1 pin 22 to U2 pin 22.

Note I have tested both games with 2732 EPROMs at U2 and U6 on a converted -17 MPU board (as documented above). This does in fact work fine for both Freedom and Night Rider.

Bally's chart for ROMs in a -17 MPU board (U1,U2,U6 are all 9316 ROMs).

Some Unique MPU-200 Facts.
There some unique relatively unknown features of the Stern MPU-200:
• On an MPU-200, if all 32 dip switches are turned off (open), at power-on it will flash seven times and jump into self-test mode. This will toggle alternatively every solenoid, flashing controlled lamps, and test each digit on the score displays.
• To down-grade the MPU-200 to the older MPU-100, remove jumpers 32-33 and 34-35 (used on all Stern MPU-200 EPROM configurations), and remove the upper 5101 RAM chip from U13.
• The MPU-200, when jumpered for four 2716 EPROMS (U1,U2,U5,U6), will run Bally games using three 2716 (U1,U2,U6) ERPOMs.
• On a MPU-200 configured for four 2716 EPROMs (U1,U2,U5,U6), to run Bally games using two 2716 (U2,U6) EPROMs, change the MPU-200's jumpers 13-14 to 13-15, and 5-7 to 1-5.

Stern M-200 and the 5101 RAM chips (boot up problems).
The Stern M-200 MPU board uses two 5101 RAM chips (instead of just one like Bally and Stern M-100 MPU boards). When buying 5101 chips, the standard speed rating on this chips is 300 ns. This works fine for Bally and M-100 MPU's, but the Stern M-200 (which runs at a higher clock speed) requires a faster 5101 RAM chip, at 100 ns. The faster 100 ns chip is labeled as 5101-1, and the slower 300 ns chip is labeled as 5101-3. If you use the slower 5101-3 RAM chip in a Stern M-200 MPU board, the board may not boot up correctly; you can get the seventh LED flash, but the game just doesn't work right. Often this can be indicated by the game audits and high scores, where a number (like '74') is continually repeated in all the audits or high scores. Remember, if using a M-200 MPU in a Bally game, the second 5101 RAM at U13 can be removed (the Bally ROM software does not use this chip).

In 1982 Bally changed their -35 MPU board very slightly to work in the combination pinball/video games Baby Pacman and Granny and the Gators, and also was used in Grand Slam. If you want to use this MPU in other pinball games, you need to convert it to a -35 MPU.

Zero Crossing Change.
The "Zero Crossing Circuit" is used to protect circuits from damage. It has to do with AC voltage, which alternates from negative voltage to positive voltage 60 times a second (or 50 times a second in Europe). This means 120 times per second AC voltage is momentarily zero, as it passes through the zero crossing point in the AC wave form. It's 120 times per second because the voltage goes from plus to minus and back through zero to plus again twice every cycle. By turning on lamps and solenoids as the power line passes through the zero point, large current surges are avoided, extending their lives. In Bally and Stern games, the CPU monitors the Zero Crossing circuit to know when to switch on CPU controlled lamps and solenoids.

Bally pins from 1977 to 1983 use the 43 volt solenoid voltage to monitor the zero crossing. Now I know what you are thinking, that the 43 solenoid voltage is DC not AC. But it is full wave rectified with no filter capacitor, so the circuit can see the voltage change from about 38 volts DC to 47 volts DC, and use 43 volts DC as the zero crossing point. But starting with Granny and the Gator, Baby Pacman, and Grand Slam, they changed the MPU board slightly to use the 6 volt AC for the General Illumination lights for the zero cross circuit. This was done by changing the 2K ohm resistor at R113 to a 1N4148 diode.

The reason Bally changed the circuit to use AC instead of DC was to be able to double the number of CPU controlled lamps used in a game without adding circuitry. With the half rectified pulsing DC, one could not tell which half cycle was on. By switching to the AC, the MPU board could know if the AC was going positive or going negative. With this, the MPU could turn on the SCRs that drove a particular set of lamps, depending on the phase (negative or positive). To do this each SCR supplies two lamps in series, and the lamps have diodes across them. The diodes are placed such that one diode will be forward conducting to shunt the current around that lamp while the other diode is blocking to light the lamp that is in parallel with that diode (hence the two lamps each have a diode installed in reverse direction to each other). This way one SCR can distinctly control two different lamps in different manner, essentially doubling the number of CPU controlled lamps in the game. Thanks to Dwight for this explaination.

How Do I know If My Game Needs a -133 MPU?
If a -35 or -17 CPU board (with a resistor installed at R113) is put in a game requiring a -133 CPU board, it just won't boot until the resistor is changed to a 1N4148 diode. But if a -133 CPU board (with a diode installed at R113) is installed in an earlier game requiring a -17 or -35 CPU board, damage will be caused to the CPU board. That's because there is no longer a 2k ohm resistor to buffer the 43 volts, and it sends 43 volts across components expecting a much lower voltage, and this of course damages them.

To determine if a game needs a -133 MPU board, measure the voltage at the MPU board's J4 pin 15 (connector removed). If it measures 6.3 volts AC (instead of 43 volts DC), then this particular game should be using a -133 MPU board (Granny and the Gator, Baby Pacman, Grand Slam). Note with a -133 MPU board installed, MPU board TP3 should show about 5 volts DC (instead of 21 volts DC as on a -35 MPU board).

Converting a -133 MPU to a -35 MPU.
Converting the -133 MPU to a -35 MPU is very easy do. Just replace the 1N4148 diode at CR52 (R113) with a 2k 1/4 watt resistor. This was the only change made. This diode is on the lower left corner of the MPU board, next to connect J4. On an original -35 MPU, this resistor is labeled R113, not CR52. Likewise converting a -35 MPU board to a -133 just involves changed R113 from a resistor to a 1N4148 diode. The striped end of the diode (cathode) should connect to capacitor C1 (that is the diode's band should be furthest away from J4).

Left: a Baby Pacman -133 board that was converted to a -35 board. Note the 2k resistor installed in place of "CR52".	Right: a -35 MPU board. Note the 2k resistor is labeled "R113".

You may also need to re-jump the board for the ROMs you will be using (typically Baby Pacman came with a 2732 U2 EPROM, and a 2532 U6 EPROM).

Converting a -35 MPU to a -133 MPU for use in Baby Pacman.
If you have a pinball -35 MPU that you want to put into a Baby Pacman game, this of course is easy too. Just remove the 2k resistor R113 and replace it with a 1N4148 or 1N914 diode. The non-banded side of the diode connectors to the J4 header pin 15 (diode band goes towards capacitor C1).

If you game "boots" correctly and goes into attract mode, you can use the built-in diagnostics for the game. This is done by pressing the red test switch mounted inside the coin door. If this switch does not work, alternatively you can short the MPU board connector J3 pin 1 to ground. If self test still does not begin, there is a problem with the MPU board. If diagnostics do begin, the problem is in the wiring or J3 connector, or in the coin door connector.

Here's what happens each time the test button is pressed:

The red self-test diagnostic button inside the coin door.

1. Lamp Test (CPU controlled Lamps). The first test is for all switched feature lamps; they will flash off and on continuously. You can check for burnt bulbs, lights that are never on, or lights that are always on (lamp driver board problems).
2. Score Displays. The second test is for the score displays. Each digit on each score display will cycle from 0 to 9, and repeat continuously. Broken diplays will have digits or segments always on or always off.
3. Solenoid Test. The third test is the solenoid test. Each solenoid will be energized, one at a time, in a continous sequence. Holding both flipper buttons "in" during this test will display the solenoid number being tested on the score display. Correct operation is indicated by the sound of a coil pulling in as its number appears on the display. If a coil makes no sound, note the coil number on the display and check out that coil. The game manual will tell you which coil the solenoid number correspond to.
4. Sound Test. The fourth button push is the sound card test (only on games equipped with one, Lost World and later; "chime" games do not have this test). A tune will be played repeatedly. Improper operation (no sound or distorted sound) can easily be heard.
5. Switch Matrix Test. The last test is the switch test. The MPU will search each switch for stuck contacts. If any are found, the switch number of the first stuck switch is flashed on the score display. Additional switches will be flashed until the last stuck switch is found. The last switch number will remain until the fault is cleared. Then the previous stuck switch is displayed. If no stuck switches are found, the number zero appears in the Match/Ball in Play display.
6. Bookkeeping. After the diagnostic test, the game will go into bookkeeping and adjustments mode. Each press of the red button will display a bookkeeping ID number, and the its value. Refer to your game manual for a description of each ID number. Note some 1982 and later Bally games also have some game adjustments available in addition to the bookkeeping. Again, see your game manual.
To exit the test mode, turn the game off and then back on.

How Coils are Energized (a Technical Introduction).
The basis behind pinball coils is that there is coil voltage at all coils at all times. Coil power is "daisy chained" from one coil to another under the playfield. Each coil is energized by completing its path to ground momentarily using two transistors on the solenoid driver board (a large SE9302/TIP102 style transistor, and a smaller pre-driver transistor contained in a CA3081 chip), allowing the coil to fire. That's why a DMM can be used to check for 43 volts on either coil lug when the game is on and in attract mode (not being played). If coil voltage is missing, either a fuse is blown or the power wire "up stream" has broken.

The following introduction text is thanks to Steve Kulpa and his web page at geocities.com/stevekulpa/bally_sole.htm. He did a really nice job of describing how the system works, so his text is transplanted here with some modifications. This information applies to the AS2518-22 and AS2518-17 solenoid driver boards, used in most Bally 1977 to 1985 games.

The solenoid driver board actually has two functions: The first obviously is to drive the solenoid and relay coils of the pinball, and the second is a voltage regulator. This provides regulated voltages to the other boards, plus high voltage (190v) to the score display driver boards. The voltage regulator stuff won't be discussed here, just the solenoid driver parts. Since the Solenoid Driver board contains the high voltage circuitry for the score displays, there is 190 volts DC on the board. Be careful as this is high voltage, and a shock from 190 volts DC will hurt.

We'll be discussing things from two circuit boards: The MPU board (AS2517-35 or AS2517-17) and the Solenoid Driver board (AS2518-22 or AS2518-17). The solenoid driver gets signals from the MPU board. These signals tell the solenoid driver which solenoid to fire. Up to 15 solenoids plus the flippers (via the flipper relay) can be controlled by the solenoid driver board.

Four signals from the MPU's U11 PIA chip travel out from the MPU board connector J4 pins 5-8,10 to the Solenoid Driver board connector J4 pins 3-7. These four signals tell the Solenoid Driver board which solenoid to fire. This is accomplished by using a 74154 decoder chip that takes the binary pattern of the four signals (16 different patterns) and decodes (or demultiplexes) them into one of sixteen different outputs. The four signals are applied to the decoder then the decoder is strobed. Normally, all sixteen of the decoder output lines are held high (+5 vdc). When strobed, the decoder lowers one of it's sixteen output lines, depending on the pattern of the four input signals.

With no input supplied (strobe is high), the output lines of the 74154 decoder are high (+5 vdc). This puts a voltage at the base of Q1 (this transistor is one of 7 in the CA3081 chip). This turns Q1 "on" and the voltage supplied to it's collector via resistor R1 passes through the transistor to ground. At this point, little or no voltage is present at the base of the large SE9302/TIP102 driver transistor Q2, and hence it is "off". With the SE9302/TIP102 driver transistor off, the 43 vdc at the coil has no place to go, and the coil remains de-energized.

When the MPU board supplies the proper input signals (A-B-C-D) to the decoder, and the decoder is strobed (signal drops to low), the proper output signal will go low, which turns the CA3081's predriver Q1 "off" (notice one of the two strobe lines goes to ground, so it's always low). This allows the +5 vdc at the Q1 CA3081's collector to flow through the diode instead of Q1 on it's way to ground via resistor R3. This also puts a voltage at the base of Q2 and turns this transistor "on". When the TIP102 (Q2) turns on, the 43 vdc at the solenoid now has a path to ground through Q1 and current flows through the coil, thereby energizing it. Then the strobe to the decoder is released, the decoder output goes high again, and everything is back to normal.

A diode, resistor, and capacitor work to slow the speed at which the TIP102 and the solenoid are able to turn off. This is important to prevent the "inductive kick" voltage that builds up when a solenoid is turned off quickly. A solenoid coil can build up hundreds of volts if it is switched off quickly. For example, the spark in the sparkplug of a car is generated from this inductive kick when the ignition coil is turned off quickly. In this case, the diode allows the TIP102 and the solenoid to turn ON quickly (which is OK), because the current that used to be flowing through the CA3081's pre-driver transistor can now flow forward through the diode and turn on the TIP102 on quickly. However, when the decoder output goes back to high and the CA3081's pre-driver transistor turns back on, the diode prevents the charge from the base of the TIP102 driver transistor from being sucked down the CA3081's pre-driver transistor. The charge on the capacitor must drain off (slowly) through the resistor and the base of the TIP102. This takes awhile and slows the turn-off of the TIP102 and the solenoid coil, thus reducing the kick. Also, as the solenoid turns off and the voltage on the collector of the TIP102 starts to rise, this voltage is "fed back" by the capacitor to the base of the TIP102 and tends to keep it on a little longer, slowing the turn-off of the solenoid even more. Another diode across the solenoid works to absorb the solenoid's turn-off kick by conducting when the voltage on the collector of the TIP102 is greater than about 43 volts.

SDB Component Guide Document.
J.McAfee made this nice document of the SDB. Check it out here.

Coin Door Lockout Coil.
The Bally coin door has a coil known as the coin lockout coil. It is a small relay-shaped coil that often energizes when the game is powered on. It triggers a mechanism that allows the game to accept coins. For example if the machine is powered off and a patron puts money in the game, the money is rejected and returned at the coin return slot. But if game is powered on and the coin lockout coil is energized, the game will accept the money and add credits to the game.

Early Bally SS games momentarily turn off the coin lockout coil during game play when the ball enters the outhole, and turns it on again just before serving the next ball or going to Game-Over. Sometimes you can hear this clicking noise at the coin door during game play, and often people will be annoyed by this or think the game has a problem.

The coin lockout coil has no use unless you want the game to accept coins. For this reason many people clip one of the wires (the thinner wire going to the non-banded diode coil lug) to the coil because they don't like the clicking sound or buzzing noise. If your Bally game sometimes clicks at the coin door and is being used at home, you can disconnect this coin lockout coil without any problems.

Also since the coin door lockout coil is on all the time, as a rule, I personally always cut the thinner ground wire (going to the non-banded diode coil lug) to disconnect the coil. Why you ask? Because sometimes this coil can burn (from being on all the time), and cause the solenoid fuse to blow. After checking every other coil in the game and then realizing the problem is the coin door lockout coil, it can get pretty frustrating. Because of this I always disable this coil and set my game to a low replay or have an exterior coin button or a free play option.

Sporadic Coil Firing (and Random Tilts).
Say the game in question has a pop bumper or a slingshot that just fires by itself, on and then off. The coil's switch is good, and is not gapped too closely, and is not closed. But the coil is just going on and off sporadically for no reason. What could this be?

Because Bally felt many high-action coils (slingshots, pop bumpers, etc.) may have quick switch closures, and the relatively slow CPU may not see the switch closure, a capacitor was added to the coil switch. This small green or brown DISC capacitor lengthened the switch closure, helping the CPU to see when that switch was closed. Unfortunately these switch caps often fail, causing "phantom" switch closures. This makes the coil in question fire for no apparent reason during the game.

Switch capacitor failure probably happened because of the high power industrial soldering irons used on the assembly line. The caps used for the switches were actually made for circuit boards. As a result the high powered soldering irons on the assembly line weakened the cap's internal insulation, causing it to leak (not function reliably) with time.

To verify this is the problem, just cut one leg of the disc capacitor off the switch. Be careful when cutting this capacitor; don't cut the small tubular diode from the switch too! (The diode is absolutely needed.) The original switch capacitors were .05 mfd or .047 mfd, 16 volt (or greater), ceramic disc, non-polarized (today the easiest to find replacement is usually .047 mfd 50 volt non-polarized caps). After cutting one leg of the capacitor, see if that fixes the problem. If it is the problem, replace the capacitor next time you're at Radio Shack or ordering electronic parts. The cap helps the game detect quick switch closures, and effectively makes the pop bumpers play better (more sensitive). But it's not absolutely necessary to have the capacitor (it can be cut off completely and not replaced). In most cases the CPU is fast enough to see all switch closures. But the best thing to do is to replace the capacitor.

Another place this switch capacitor is often see is on the plumb-bob tilt. A failed switch capacitor here will make the game tilt for no apparent reason during game play. Definately cut this capacitor off, and don't replace it!

No Coils Working Diagnostics.
If none of the coils work, first look at the power supply:

• Rectifier board test point 5 (TP5, top right test point) should be 43 volts DC. If no 43 volts here, check fuse F4. Remove the fuse and check it with a DMM set to ohms ("buzz out" the fuse).
• The reason we remove fuses is to force a verification that the fuse holder is in good condition. The fuse holder clips can fatigue or burn, and often need to be replaced. If this is the case, 43 volts may not be getting to the rest of the game.
• If fuse F4 is good, the lack of 43 volts at TP5 is probably due to a failed (open) bridge rectifier BR3 on the rectifier board.
• If fuse F4 is bad, and a new fuse immediately blows at power-up, then bridge rectifier BR3 on the rectifier board is probably shorted.
• If 43 volts is found at TP5, check rectifier board connector J1 pin 6 for 43 volts DC. This is usually a brown wire (Bally) and goes directly to the playfield flipper coils.
  • Check rectifier board connector J3 pin 9 for 43 volts. This goes to the solenoid driver board connector J3 pin 5 for the flipper relay.
  • Check rectifier board connector J3 pin 12 for 43 volts. This goes to the MPU board connector J4 pin 15, and then to resistor R113.
  • Check rectifier board connector J3 pin 13 for 43 volts. This goes to the backbox on games that have a knocker in the backbox.
  • Check rectifier board connector J2 pin 2 for 43 volts. This goes to the lower cabinet for the coin lockout coils. Also goes to the knocker coil on games with the knocker in the lower cabinet. On early games this also goes to the chime unit.
• Locate the brown power wire at one of the flipper coils and check for 43 volts DC. If it is present at J1 pin 6 but not at any of the flipper coils, and then there is a wiring problem between the rectifier board connector J1 and the coils.
• If only the flipper coils work, than the 1 amp slow-blow under-the-playfield solenoid fuse is problaby blown. Or perhaps the brown wire from the flipper coils to the 1 amp fuse has broken.
• If the coils still do not work, check the solenoid driver board for 5 volts DC at TP3. If missing, look for a broken jumper wire on the solenoid driver board connector J3 that goes from pin 13 to pin 25.
• If only some coils works (in addition to the flipper coils), look for a broken yellow power wire under the playfield that runs from coil to coil.

Other possible (and more bizarre) problems if coils do not work:

• Possible problem with the game's ROM code that goes from the MPU connector J4 pins 5-8,10 to the solenoid driver board connector J4 pins 3-7. This code selects which of the 16 coils will fire. If one line is missing, coils 1 to 4 will not fire. Check MPU and solenoid driver board connectors J4 for broken wires or bad connector pins.
• Early A8 sounds boards AS2518-32 (games Lost World through Dolly Parton, though sometimes Star Trek to Dolly Parton will have a AS2518-50 sounds board) use the 43 volts. This sound board uses 43 volts DC to make 12 volts DC with a very crude voltage divider/regulator. Sometimes this sound board circuit fails and will short the 43 volts to ground. When trying to diagnose strange 43 volt coil problems in a games using the A8 soundboard, disconnect the power to the sound board before troubleshooting.
• If the 43 volts is missing from the whole game, then the MPU will not complete its 7th flash at power-up. This can be verified by checking for 43VDC on the left side of R113 (located below and to the right of the J4 MPU connector). If the 43VDC is present on the left side of R113 but there is no reading on the right side of the resistor, then replace R113 (2K ohms 1/4 watt). If both sides of R113 show 43 volts then the MPU board may have battery corrosion damaged and the whole area should be repaired.

Only Some Coils Work.


• Check is for power at all the coils in question. Using a DMM set to DC volts, put the black lead on the ground strap in the bottom cabinet of the game. Put the red lead on either lug of any coil. 43 volts DC should be seen. If power is only seen on one coil lug, the coil is bad. In no power is seen at either coil lug, check the power wire 'upstream'. Remember the 43 volts is daisy chained from coil to coil.
• If there is power at the coil, there could be a bad J4 connector (lower right) on solenoid board. A cracked solder joint on the solenoid board J4 header pins or a failed .100" terminal pin in the connector itself. Loosing contact with one of the signal lines can drop a bit, and the solenoid driver board can interpret the instructions incorrectly, firing the wrong coil or no coil at all.
• Bad Solenoid Driver board J4 (upper right side) or MPU board J4 (lower left side) connectors. A cracked solder joint on these header pins or a failed .100" terminal pin in the connector itself. Loosing contact with one of the signal lines can drop a bit, and the solenoid driver board can interpret the instructions incorrectly, firing the wrong coil or no coil at all. The encoding is sent to SDB connector J4 pins 3-6 (PB3-PB0 respectively) and SDB J4 pin 7 (CB2, solenoid bank select). This comes from connector MPU J4 pins 1-8 (PB0-PB7), which go thru resistors R97-R106, and then source at PIA U11 pins 10-17 (PB0-PB7). Any break in this connection stream will drop a bit and cause a solenoid to either not fire, or the wrong solenoid to fire.
• Bad decoder 74LS154 chip on solenoid board (or 74LS138 at U4 on Baby Pacman SDB), which mis-interprets the drive signals PB0-PB7 from the MPU board. This is the least likely problem but it does happen.

Locked On Coil Diagnostics.
If a coil is locked on (Burning! Turn the game off!) or doesn't work, there are several tests you can perform to isolate the problem. This is often seen with the game is first turned on (before the 7 MPU flashes finishes), a coil energizes and won't let go until the power is turned off. First it's good to know the sequence of events in energizing a coil:

• The MPU is told (by a playfield switch or other trigger) to fire a coil.
• The MPU turns on, for just a moment, a solenoid transistor on the solenoid driver board. This completes the power path to ground for the particular coil.
• The coil fires.
There are a series of steps you should take when a coil is not working properly, which we will outline below.

If a Coil is Locked On.
Generally, this is caused by a solenoid driver transistor that is shorted on. If a coil is locked on as soon as the game is powered on, turn the game off immediately (otherwise you'll be replacing more than a bad transistor!). Then follow these steps:

• Check the manual's schematics to figure out which transistor controls the coil in question. This information is on the Solenoid Driver/Voltage Regulator schematic page.
• Look at the connector in the center of the schematic. There the coil name/description will be listed.
• Follow this line back to the first "Q" (transistor) that intersects this line. Write down the transistor number (for example, "Q13"). Also write down the diode number behind it ("CR13"), and the chip number that drives this transistor ("U3"), and the pins of the chip (pins 11, 12). Also note the pin number that connects to the diode (pin 12). All these components could be damaged (but generally it's just the transistor).
Now that you know the transistor in question, you can test it.

Other Coil Diagnosing Techniques.
Another technique is to remove the solenoid connectors J1/J4/J5 from the solenoid driver board to the playfield. These connectors are on the left side of the solenoid driver board, and the one connector at the bottom right of the board too. Now power up the game (with good fuses), and install one connector at a time, until a fuse blows. Which ever connector blows a fuse, look for a little "spark" on one of the connector pins. Now go to the schematics, and see what coil that connector pin goes. Check that coil to make sure it has 2.5 ohms or greater resistance, the coil diode, and the coil's associated solenoid driver board transistor.

Remember there are two fuses that handle power for the coils. The 1 amp underplayfield slow blow fuse will usually only blow if a coil is energized and staying energized. Yet the power supply F4 fast blow fuse usually blows if there is a "hard short" (a dead-short across a diode, or coil that is shorted out (less than 2 ohms), or coil power is shorting directly against a metal ground).

Another thing is if fuse F4 blows, there is often a problem with the backbox knocker, or the cabinet coin door lockout coil, the solenoid bridge rectifier (on the rectifier board), or the rectifier board's varister.

Also try removing connectors J1 and J3 from the rectifier board (this moves the solenoid power back a step further, not allowing it to get any further than the rectifier board). Replace fuse F4 on the solenoid driver board, and turn the game on. If the fuse still blows, the solenoid bridge rectifier (on the rectifier board) or the rectifier board's varister is probably at fault.

Finally, disconnect a wire on each solenoid, and re-attach each wire, one at a time, until the F4 fuse blows. At this point it could be the coil, coil diode, or coil driver transistor at fault.

Testing the Solenoid board Transistors, Game Off.
The transistors on the solenoid driver board are very easy to test. This is done with the game off. You can remove the solenoid driver board, or leave it installed in the game. Using your DMM set to the "diode" setting, do the following:

• Turn the game off.
• On the component side of the board, put the black lead of your meter on the metal tab of a driver transistor.
• Put the red lead of your meter on the center lead of a transistor. Your meter should read zero.
• Put the red lead of your meter on either outside lead of a transistor one at a time. Your meter should read in the .4 to .6 volt range.
• Put the red lead of your meter on the other outside lead of the transistor. Your meter should again read in the .4 to .6 volt range.
If your meter reads anything outside the .4 to .6 range, replace that transistor.

Testing a solenoid driver board transistor. The black lead of the DMM is on the transistor's metal tab. The red lead is put on either outside lead, one at a time. The meter should read in the .4 to .6 range.

Testing a Solenoid board from the Transistor to the Coil, Game On.
If a coil is not locked on, you can test it's solenoid driver board from the transistor to the coil, with the game on and in "attract" mode.
• Attach an alligator wire and clip to ground in the backbox.
• Momentarily touch the other end of the alligator clip to the metal tab on a solenoid driver board transistor.
• The coil that is driven by that transistor should fire.
• Note this doesn't actually test the transistor itself, but just the path from the transistor to the coil.
• Optionally, you can also momentarily touch your ground wire to the solenoid board "U" chip; the pin that does NOT connect to the diode. This will also fire the coil.
If the coil doesn't fire, and the transistor tested properly in the above steps "Testing the Solenoid board Transistors, Game Off", you have either a blown playfield fuse or a broken wire/connector.

To test for a broken solenoid wire or connector pin, do this:

• Turn the game off.
• Put an alligator clip on the coil lug that the NON-banded side of the diode connects to.
• Connect the other end of the alligator clip to one of the test leads on your DMM.
• Set your DMM to continuity ohms setting.
• Refer to the manual and find the "J" connector number and pin number that the solenoid in question connects to on the solenoid driver board.
• Touch the other lead of your DMM to this "J" connector pin on the solenoid driver board.
You should get about 0 ohms. Note if you are testing to the wrong conector pin, you will get about 30 ohms.

Driver Transistor Tested Good, but Coil is still Locked On.
The driver transistor may be OK, but the 1N4004 diode behind it could be bad. Since we wrote down the diode that is behind the driver transistor (in the above steps), refer to that to get the diode number. Here's how to test it:

• Turn the game off.
• Remove the Solenoid driver board from the game.
• Put your DMM on diode setting.
• On the component side of the board, put the DMM leads on the 1N4004 diode. You should get a reading of .4 to .6 volts.
• Reverse the leads, and you should get the same reading you got in the previous step.
• BEST METHOD: remove one lead of the diode from the driver board, and retest. In one direction you should get a zero (null) reading.
If you get any other value, replace the 1N4004 diode.

Test the pre-driver CA3081 transistor chip. The picture on the left is testing the transistor pin that connects to the diode.

Now we need to test the chip that drives the diode and driver transistor. This chip is a CA3081 (NTE916) transistor array. Basically it's several transistors packaged in a chip format. This is known as the "pre-driver transistor pack". You can test this too:
• Turn the game off.
• Remove the Solenoid driver board from the game.
• Put your DMM on diode setting.
• On the component side of the board, put the black lead of your DMM on the GND test point just to the left of the U1 chip.
• Put the red lead of your DMM on the two pins of the pre-driver chip (one at a time) that you noted above.
• For the pin that connects to the 1N4004 CR diode, you should get a reading of .1 to .2 volts.
• For the other pin, you should get a reading of .7 to .8 volts.
• Reverse the DMM leads (red lead now on GND). You will get the same .1 to .2 volt reading for the pin that connects to the CR diode. The other pin will read 1.1 to 1.3 volts. Note this test is far less conclusive than the first test with the black lead on GND.
If you get any other meter readings, replace the pre-driver CA3081 (NTE916) chip.

Always Replace the SE9302 Transistor with a TIP102.
When ever replacing a bad SE9302 driver transistor, always replace it with the more robust TIP102. The TIP102 can handle more current and will just plain last longer.

Replace the Coil Diode and Driver Board Diode/330 Ohm Resistor.
If you had ANY problems with a coil being locked on, ALWAYS replace the coil diode. This 1N4004 diode prevents a "backlash" of current going to the solenoid driver board when the magnetic coil is shut off. This diode is easily damaged and can cause damage to your solenoid driver board if bad. Always make sure you install the new diode with the diode band on the "power" lug of the coil. The power lug is usually the lug with two wires going to it (because the power is daisy chained from coil to coil).

Likewise, on the solenoid driver board, also replace the associate CR diode that connects to the replaced driver transistor. This is also a 1N4004 diode. And it's a good idea to also replace the 330-ohm (1/4 watt) resistor that connects to this diode. Both of these components can really get toasted.

Testing for Power at the Coil.
If a coil doesn't work, and the transistor is good, test for power at the coil. Do this with the game on and in attract mode, and the playfield lifted.
• Put your DMM on DC voltage (100 volt range).
• Put the black lead on the metal side rail (ground) of the game.
• Put the red lead on either terminal of the coil. It should read about 43 volts. On flipper coils, any of the three terminals should also read about 43 volts.
If you are missing voltage at the coil, check for a broken wire/connector, or blown playfield fuse (see below). Remember the power wires are "daisy changed" together. So a break in the power wire in a previous coil will cause the coils further down the line to not work.

Testing a Coil.
You can also test a coil for proper operation. With the game on and in attract mode, and the playfield lifted, try this:

• Connect one end of a alligator clip and wire to the metal side rail of the game.
• Momentarily touch the other end of the alligator clip to the coil's terminal with the non-banded side of the diode connected to it.
• The coil should fire.
Note if you accidentally touch the banded side of the diode to ground, you will probably blow a fuse.

If the coil doesn't fire, you have a damaged coil or no power at the coil. Look for a broken wire going to the coil's terminal. You can also test the resistance of a coil. A good coil should be in the 3 to 15 ohm range.

The Over-Looked Under-Playfield Solenoid Fuse.
Often your Bally game will boot fine and start a game. The flippers work, but no other solenoids on the playfield work. This can often be caused by a blown under-the-playfield solenoid fuse.

If you run the solenoid diagnostic test and the coin lock-out coils, the flipper relay, and the knocker all work, and nothing else works, a dead under the playfield fuse could be the problem. Since the game boots OK, we know the +43 volt fuse on the rectifier board is OK (if this fuse was blown the MPU board won't "flash" the seventh time).

The under the playfield solenoid fuse is usually located on the right hand side by the flippers. Usually it's a 1 amp slo-blo fuse. If this fuse keeps blowing, you have a solenoid problem on the playfield somewhere. This can be caused by a shorted coil, a bad coil diode, or a broken (and shorted) coil wire. A shorted and locked on driver transistor is probably NOT your problem.

If the playfield fuse keeps blowing, there is another procedure you can try to isolate the problem as a last resort. Turn the game off and disconnect the "pull down" wire from EVERY coil under the playfield. The pull down wire is the single wire on each coil, and connects to the NON-banded side of the coil's diode (the power side connects to the banded side of the diode's coil lug). Then power the game on (the fuse should not blow!). Now re-connect each wire to its respective coil. When the fuse blows, you've found your problem coil/diode.

Coil TroubleShooting - More Ideas.
Here's some other ideas on coil problems.

No Coils Work.
• Check TP5 on the power supply for 43 volts. If no voltage, check fuse F4 on the power supply.
• Check solenoid driver board for +5 volts at TP3 on the driver board. If no voltage here, check for a broken jumper on connector J3 from pin 13 to pin 25.

Only the flippers work.

• Check 1 amp slow blow fuse underneath the playfield.
• Check for a broken wire from the above fuse. There should be a brown wire going from the under the playfield 1 amp fuse to the flipper coil.

Flippers work and just some coils.

• Under the playfield, check for a broken yellow wire from coil to coil. This is the coil power wire, and it daisy chains from one coil to another. A break in this wire will stop power from getting to coils "down stream".
• A problem with the connection between connectors J4 of the MPU board and J4 of the solenoid driver board. If one line is missing, coils 1 to 4 will not work.

Flipper does not work or are weak.


• Make sure the EOS (end of stroke) switch is filed clean & shiny with a metal file, and adjusted to be normally closed.
• EOS switch adjustment: When the flipper is fully energized, the EOS switch should open about 1/8". If this switch always stays closed, the flipper coil will burn. If the switch is always open the flippers will be very weak.
• File the cabinet switches with a metal file so they are shiny.
• Make sure the flipper coil wires that are part of the coil have not broken where they attach to the lugs. Remember a flipper coil is actually two coils in one package. The center lug is common, then one outside lug has a thick wire, and the other outside lug has a thin wire. There should also be two 1N4004 diodes on the flipper coil lugs. Often the thin wire will break away from the lug making the flipper "machine gun".
• Check the under playfield fuse and fuse holder. Often the fuse holder will be tarnished or have bad tension. Replace the fuse holder if in doubt.
• With the game on and in game over mode, use a DMM and measure for 43 volts DC at all three flipper coil lugs. If there is only power at 2 of the lugs, the flipper coil or EOS switch is bad. No power at any lugs check the fuses. Power for the flippers is "daisy chained" from other coils and sources. Sometimes the chain breaks "upstream". Look for that.
• Try grounding (momentarily) the center lug of the flipper coil in question - it should fire. If this makes a weak flipper strong, there is a bad ground path from the flipper to the flipper relay.
• Inspect the back of the Solenoid Driver board. The flipper ground path comes from the cabinet flipper switches to the SDB connector J1 (upper left connector). Then it goes to the flipper relay. Sometimes the circuit board traces from the connector J1 to the relay burn right off the SDB! Or the relay's solder joint and/or J1 header pin solder joints crack or go cold, causing the flippers to be weak or intermittent.
• Make sure the flipper relay energizes. If the flipper relay never energizes, the flippers will never work. The flipper relay should energize when a game is started or during the coil diagostic test. Also sometimes the contacts on the flipper relay can burn or pit.

Power supply board coil power distribution.

• J1 pin 6 = brown wire to flipper coils.
  
• J1 pin 9 = to solenoid driver board J3 pin 5 (flipper relay).
  
• J2 pin 2 = to chime unit (in early games) or knocker coil (in later games), and coin lockout coils in all games.
• J2 pin 13 = to backbox knocker on early games.
• J3 pin 12 = to MPU board connector J4 pin 15.

Wrong Coil (or No Coil) Energizes.
This is an interesting problem that isn't very easy to diagnose. For example when a slignshot switch closes, the drop target bank resets. Or perhaps one particular coil never engergizes in the game.

This is often due to the PB0-PB3 encoded lines that come from MPU PIA U11 or the line select. These four bits are encoded by the PIA, sent to the Solenoid Driver board (SDB), then unencoded by the SDB U2 74154 chip, and the appropriate solenoid is energized. If one of these four bits is lost in the data path between the MPU and the SDB, the unencoded number will be wrong, and the incorrect solenoid (or no solenoid) is energized.

Another test of this problem is to put the game in coil test. If coil(s) fire twice (same coil for two solenoid numbers), or does not fire at all, there is potentially a problem with the PB0-PB3 encoded lines.

The data path for this encoding starts at the MPU U11 PIA chip. Unfortunately this chip is right in the battery corrosion zone. So a bad U11 socket or broken trace is very common. Or if the MPU J4 pins 5-8,10 connector or SDB J4 pins 3-7 connector are bad (just one pin missing), the coil that gets fired will be decoded incorrectly by the SDB. Last the SDB U2 chip (74LS154) chip could be bad (though this is the least likely problem). Here's a list of things to check, and a table that will allow easy tracing of the PBx signals.

• Wrong game MPU ROMs installed - ROMs from another game installed at U2/U6.
• Bad U11 PIA on MPU board or bad socket for MPU U11 (which is in the "battery corrosion zone"). Swap PIA chips U10 and U11 to see if anything changes.
• Bad J4 pins 5-8,10 connector on MPU board (lower left corner). A cracked solder joint on the MPU J4 header pins or a failed J4 .100" terminal pin in the connector itself. Loosing contact with one of the signal lines can drop a bit, and the solenoid driver board can interpret the instructions incorrectly, firing the wrong coils. It's a good idea to use a DMM set to continuity, and "buzz out" the wire path from the MPU board to the Solenoid driver board. This should be done with the connectors in place, so any connector issues are identified. (That is, measure continuity from the MPU board components R97-R100 to the SDB component U2, inclusive of the connectors. See the table below for help with that.)
• Bad J4 pin 3-7 connector on solenoid board (upper right corner). Again a cracked solder joint on the solenoid board J4 header pins or a failed J4 .100" terminal pin in the connector itself.
• Swapped wires on the MPU board's switch matrix connector from a previous repair.
• Bad decoder 74LS154 chip on solenoid board, which mis-interprets the drive signals.
• Short from solenoid power (43 volts DV) to a switch matrix line (12 volts DC) somewhere under the playfield. This is sometimes seen on games with three or six bank drop target assemblies. The drop target units have a daisy-chain of 43 volts DC that go from coil to coil on the drop target unit (and sometimes to small memory or knock-down coils). It is very common that the wiring wears from rubbing on metal and eventually shorting 43vdc to the whole metal chassis of the target unit. This can cross over to the switch matrix lines from the switches touching the arms at the bottom of each drop target. A quick check is to unplug the wiring harness connector that connects to drop target bank assembly, and re-boot the game.

Note that PB0-PB3 are the encoded "momentary" solenoid lines, and that CB2 is the bank selection bit. That is PB0 to PB3 are the bits that get encoded and unencoded. There is also PB4-PB7 which are "continuous" solenoid lines (these usually control the coin lockout coil, flipper disable, and potentially two other coils). These bits don't get encoded, and control coils directy from the PIA (they don't go thru the SDB 74154 chip).

Note on Baby Pacman things are a bit different. First only PB0-PB2 are encoded. PB3 and PB7 are not used. PB4-PB6 are the continuous solenoid lines. Also the SDB connector is different: J9 pins 7,6,5 are PB0-PB2 respectively. J9 pin 4 is CB2 (solenoid select bank). J4 pins 10,9,8 are PB4-PB6 respectively. SDB U4 (74LS138) does the signal decoding (but again, this is the least likely problem).

When the wrong coil (or no coil) energizes, I find it best to first trace the PB0-PB7 lines from the MPU U11 PIA to the Solenoid Driver board.

PBx	Signal Type	PIA U11	MPU Resistor	MPU J4	SDB J4*	SDB U2 (74154)*
PB0	Momentary (encoded)	U11 Pin 10	R97 (470 ohm)	MPU J4 pin 4	SDB J4 pin 6	U2 pin 23
PB1	Momentary (encoded)	U11 Pin 11	R98 (470 ohm)	MPU J4 pin 3	SDB J4 pin 5	U2 pin 22
PB2	Momentary (encoded)	U11 Pin 12	R99 (470 ohm)	MPU J4 pin 2	SDB J4 pin 4	U2 pin 21
PB3	Momentary (encoded)	U11 Pin 13	R100 (470 ohm)	MPU J4 pin 1	SDB J4 pin 3	U2 pin 20
PB4	Continuous (not encoded)	U11 Pin 14	R101 (330 ohm)	MPU J4 pin 5	SDB J4 pin 11	n/a
PB5	Continuous (not encoded)	U11 Pin 15	R102 (330 ohm)	MPU J4 pin 6	SDB J4 pin 9	n/a
PB6	Continuous (not encoded)	U11 Pin 16	R104 (330 ohm)	MPU J4 pin 7	SDB J4 pin 8	n/a
PB7	Continuous (not encoded)	U11 Pin 17	R105 (330 ohm)	MPU J4 pin 8	SDB J4 pin 10	n/a
CB2	Coil Bank Select	U11 pin 19	R106 (470 ohm)	MPU J4 pin 10	SDB J4 pin 7	U2 pin 19

* This table's SDB info does not apply to Baby Pacman, Granny & Gators, Grand Slam.

Bally's lamp driver and auxiliary lamp driver boards stayed pretty consistant from 1977 until 1989 (when Bally produced its last game, before being taken over by Williams). These procedures should work on all Bally lamp driver boards from their inception until their end in 1989.

The power for all CPU controlled lamps comes from the rectifier board through fuse F1. On pre-Xenon games (transformer in the backbox), 7 volts AC comes from the transformer to a 10 amp fuse, through a 8 amp bridge rectifier (converted to 6 volts DC), and then to all of the playfield's CPU controlled lamps (there is a lamp power buss wiring going to all CPU controlled lamps under the playfield). On Xenon and later games, it works the same way but the bridge rectifier is larger and so if the F1 fuse (20 amp). The lamp driver board is really a misnomer, as the lamp driver board drives nothing. It has a SCR (Silicon Controlled Rectifier) for each of the CPU controlled lamps. The SCRs switch ground on and off based on decoded information from the MPU board. This completes the power circuit to that particular CPU controlled lamp, turning it on.

If a feature light is continually on, or is never on, you can test the lamp driver board for a component problem. Assuming the wiring is intact, chances are good that the lamp's driving component(s) are bad. This is especially true if a lamp is always on. The internals to the driving components have probably shorted on, leaving the lamp continually on.

All Bally electronic pinball games until Williams bought them out in 1989 used SCR's (Silicon Controlled Rectifiers) to drive feature lamps. SCR's are different than transistors. Instead of a collector, base and emitter like a transistor, they have a cathode, anode and a gate (abbreviated C, A, G respectively, though sometimes the "C" is abbrevated as "K"). Each gate is driven by a CD4514 CMOS decoder output. All SCR cathodes are connected to the feature lamp ground. Each SCR anode is connected to a unique feature lamp.

There are two different SCR's used for lights on the lamp driver board: the larger MCR106-1, and the smaller 2N5060. They serve the same function, just the larger MCR106-1 can handle more current (and sometimes lights two lamps, while the smaller 2N5060 can only light one lamp). There is also a CD4514 CMOS decoder that drives the lamps. Sometimes these go bad too.

Why No Lamp Matrix?
Bally didn't use a lamp matrix like Williams did. Bally's approach was more like Gottlieb's system80, where there was a single transistor or SCR that drives each lamp. There are a total of 24 of the larger MCR106 SCRs, and 36 of the smaller 2N5060 SCRs. This gives a maximum total of 60 discrete CPU controlled lamps in a typical Bally game (the MCR106 could actually control two lamps at the same time, but I count that as one discrete lamp). Some game also had an Auxilary Lamp Driver board which allowed for more than 60 CPU controlled lamps. Because there are 60 lamps and 60 SRCs and no lamp matrix, there are 61 wires going to all the CPU controlled lamps (one power wire, and then a single control wire for each of the 60 lamps). Connectors J1 and J3 on the lamp driver board have 28 pins each, meaning potentially 56 wires go to the playfield for the CPU controlled lamps. The remaining CPU controlled lamps go to the backbox through connector J2 for (at minumum) the Game Over, Tilt, Match, High Score to Date, and Ball in Play lamps, and some other lamps.

Power to the Lamps?
All the playfield CPU controlled lamps have 6 volts DC going to them from the transformer's rectifier board. If none of the CPU controlled lamps work, usually this is a blown rectifier board fuse, bad rectifier board connector, and/or bad rectifier board bridge.

On the bottom of the playfield you will notice a bare wire going to all the CPU controlled lamps. This is the Positive lead of the 6 volts for the lamps. Then on the Tip of each socket there is a color coded wire. This wire goes to the lamp driver board, which grounds this wire, turning the particular lamp on. If none of the lamps work, using a DMM set to DC volts, check the bare playfield wire and make sure there is 5 to 6 volts DC power for the lamps. If not, go to the rectifer board and find why there's no power to the lamps.

Diagnosing Non-Working CPU Controlled Playfield Lamps.
Start the diagnosis by putting the game into the first diagnostic test by pressing the Red test button inside the coin door. This will make all the CPU controlled lamps flash on and off. Note if the game has a Solenoid Expander board under the playfield, this board will turn it's relay on and off, and its accompanying 555 lamp will flash too.

First make sure that a non-working playfield lamp's SOCKET is not the problem! Bally lamp sockets are really crappy, and most non-working lamp problems are related to the socket. Remove the bulb and put it into an empty backbox socket as a test of the bulb. If the bulb is Ok, put it back in the playfield socket and twist it a couple times. Often this will make a non-working lamp turn on. If you want a more permanent solution to a lamp that doesn't want to stay working, either replace the socket with a new one, or solder the socket (as shown in the picture below).

"Fixing" a playfield lamp socket. The wire that powers the tip of the bulb is moved directly to the tip of the socket. The base of the socket is then soldered together so it can not rotate. Be sure to sand the parts before soldering, and to use some Rosin flux on the socket.

If the lamp still does not work, I use a logic probe to see if the lamp is really being driven by the lamp driver board. I know most people don't have a logic probe (but you should, as they are only $20), but it really is a good way to see what is going on. Put the logic probe on a WORKING lamp socket's TIP, and notice how the signal pulses. Now put the logic problem on the Non-working lamp socket's tip, and compare. Is the signal the same? If so, you have a socket problem. If the signal is different, then it's time to move up to the lamp driver board for more diagnosis.

Start by making a list of the non-working lights. If you don't have the schematic, list the non-working lamps AND their wire color that connects to the lamp socket's tip. Now go to the lamp driver board and find the connector with that wire color (game off). Using a DMM set to continuity, you can verify that you have found that wire's connector pin on the lamp driver board. Use a Sharpie pen and mark that pin on the lamp driver board.

Remove the lamp driver connectors, and look at the .100" molex female connector pin in the housing. Sometimes these pins break, and this may be the only problem. Otherwise remove the lamp driver board from the game, and look at the male .100" connector pin. Another common problem is the solder joint of the male connector pin to the board cracks. Touching up the solder at this pin again often fixes a non-working lamp. CAREFUL though, as it is very easy to bridge solder across two pins (these .100" connectors are small and the pins are close together).

Lamp still doesn't work? At this point i use the DMM set to continuity, and find the SCR on the lamp driver board that connects to the male pin for the non-working lamp. Mark the SCR with a Sharpie, then change the DMM to diode function. Using the diode test, the SCR can be checked against a working SCR next to it, to see if the suspect SCR is bad. If the SCR is suspect, replace it. At this point, nearly all non-working lamp problems should be handled by this methodology. (If a lamp or set of lamps still doesn't work, keep reading below.)

Lamp board decodes for 8 Ball Deluxe U1 chip in blue box.

More Complicated Lamp Problems.

AD0-AD3 Lamp Data Lines and PD0-PD3
(or Mass Amounts of Non-Working Lights).
If a particular game has lots of non-working lights, this could be related to a connector issue. To really understand this problem, one must know how the Lamp driver board decodes the lamp info from the CPU board.

There are four lamp data lines (AD0-AD3) which ultimately decide to which lamp gets turned on. These four data lines are binary in nature. That is, they can be 0 or 1, and the combination of these four lines decodes to 16 different binary values. (There is also a toggle line PD0, which is the bit that actually turns U1 chip 'on'.) The AD0-AD3 lamp data lines get decoded by the four 4514 chips on the lamp driver board, and ultimately go to each of the SCRs and to each of the lamps.

For example, below is an Eight Ball Deluxe lamp driver schematic. Now say you run the game's lamp diagnostic test and find a bunch of non-working CPU controlled lamps (Rack 9-15, Target 9-15, "D E L U", 20k,40k,60k saucer values, 50k,70k, 2x,3x,4x,5x,56k,112k bonus values, special saucer, special left lane, shoot again). And the pop bumper lights are always on but go bright/dim/bright/dim. So how do you figure out what is wrong?

Go to the lamp driver schematic and mark all the non-working lamps. Sometimes a pattern can be seen where the lights that don't work are always lamp driver values 2,3, 6,7, 10,11 for all four U1-U4 decoding chips (for now forget about the Aux Lamp Driver board used in 8 Ball Dlx). It's important to remember that AD0 is the right most bit, and AD3 is the left most bit. With this in mind, looking at the binary values for these six missing lamp numbers, another pattern can be seen. That is, the AD1 line (second bit from the right) is always "1" on these six missing lamp lines binary numbers.

Knowing this, we can conclude that the AD1 line going to the lamp driver board is probably broken (or stuck low at the MPU board, meaning a PIA problem). But most likely this is a bad connector pin at the MPU J1 connector (J1-15=AD0, J1-14=AD1, J1-13=AD2, J1-12=AD3). Or at the lamp driver board J4 connector (J4-14=AD0, J4-15=AD1, J4-16=AD2, J4-17=AD3). In this case, because some of the missing lamps are driven by the Auxiliary Lamp Driver board, the problem is almost for certain at the MPU board J1 connector.

There is one more consideration, that is the PD0-PD3 lines. Remember the lamp data AD0-AD3 decodes value from 1 to 16 (in binary). But there are also four PD0-PD3 lines that tell the lamp driver board which U chip the AD0-AD3 lamp data is for. For example say PD2 is toggled. That means lamp data AD0-AD3 applies to the U3 chip (and the lamps which are controlled by the U3 chip). So if you have 16 lamps that don't work and they are all controlled by a single lamp driver U chip, it's most likely a problem with the corresponding broken PD line (PD0 for U1, PD1 for U2, PD2 for U3, PD3 for U4).

The last issue to think about is this: coin door switches. Now normally you would think this has nothing to do with lamps, but it does. If the coin door switches are shorted to the metal coin door skin (ground), or the insulating fishpaper behind the start button is gone (shorting this switch to ground), this can cause all sorts of strange behavior. Like the same type of problem as one of the AD0-AD3 lamp lines missing, where numerous lamps may not work. A quick test of this is to remove MPU connector J3 (lower right). If problems clear, a shorted coin door switch is definately the problem.

Left: the -23 Lamp Driver board with no 4050 buffer chips. Right: the earlier -14 Lamp Driver board with 4050 buffer chips. Bottom: Stern's LDB-100 Rev B lamp driver board. All three of these lamp driver boards are interchangable with each other.

Two Different Lamp Driver Boards.
Bally had two different lamp driver boards, the AS-2518-14 and the AS-2518-23. Both boards have the same number of SCRs, and are completely pin compatible and either board will work in any 1977-1985 Bally solidstate game. The only difference is the earlier -14 Lamp Driver uses four CD4050 buffer chips between the MCR106 SCRs and the CD4514 decoder (the smaller 2N5060 SCRs did not use a buffer chip). But Bally found the lamp driver board to be so robust that the CD4050 buffer chips were not needed. Therefore the -23 version of the Lamp Driver board has just four CD4514 chips and 60 SCRs, with no CD4050 buffer chips. The -14 lamp driver board was only used for from 1977 to 1978 (for example Freedom to Power Play used the -14, but Mati Hari used the -23) when it was replaced by the -23 lamp driver board. The -23 lamp driver even has silk screened on the board, "replacement for AS-2518-14". Even though the -23 board is larger, it was cheaper to produce than the earlier -14 board. The reason was the -23 board is a single-sided board, with circuit board traces only on one side (this is cheaper to make compared to a two-sided board). All the "jumpers" seen on the -23 are not user changable - they are required to move a board trace around other traces on the single-sided board.

Replacement SCR's.
Replacement SCRs are available from a variety of sources. For example, Jameco sells the MCR106-1 as part number C106Y or C106B1, and NTE5411 to NTE5416. The smaller 2N5060's replacement number is 119802 or NTE5400 to NTE5406 (but the NTE version is much more expensive) or at Mouser part# 610-2n5060. Also the larger MCR106-1 can be used in place of the smaller 2N5060, but only if the "A" and "G" legs are "reversed" (twisted to be reversed, when installed). This is not recommended, but it can be done in a pinch.

Mounting a MCR106.
There are many styles of MCR106 rectifiers. When mounting the style without the metal face, note the ANGLE side and the mounting orientation. The angle sided C106 seems to mount "backwards", compared to the metal faced version of the MCR106 (because the manufacturer writing seem like it is on the back of the angled sided SCR). See the picture below.

Mounting of the two styles of MCR105 rectifiers. Note the angled sided SCR compared to the metal faced SCR.

The two styles of SCRs used in Bally games. Left: the SCR106-1. Right: the 2N5060.

This table (thanks to S.Kulpa) shows which SCR goes to
which connector number on any of the lamp driver boards.

SCR Connector Q1 J1-24 Q2 J1-25 Q3 J1-26, J2-21 Q4 J1-28 Q5 J2-16 Q6 J2-14 Q7 J1-27, J2-13 Q8 J1-23 Q9 J1-14 Q10 J1-15 Q11 J1-16 Q12 J1-19 Q13 J1-17 Q14 J1-18 Q15 J2-23

SCR Connector Q16 J2-22 Q17 J1-11 Q18 J2-20 Q19 J2-15 Q20 J1-13 Q21 J1-12, J2-12 Q22 J1-10 Q23 J1-4, J2-8 Q24 J1-5 Q25 J1-6 Q26 J1-7 Q27 J1-9 Q28 J1-8 Q29 J1-1 Q30 J2-6

SCR Connector Q31 J2-2 Q32 J3-27 Q33 J2-11 Q34 J1-2 Q35 J1-3 Q36 J3-26 Q37 J3-23 Q38 J3-25 Q39 J2-4, J3-24 Q40 J2-9, J3-22 Q41 J3-20 Q42 J3-21 Q43 J2-7 Q44 J3-19 Q45 J2-1

SCR Connector Q46 J3-18 Q47 J2-10 Q48 J3-16 Q49 J3-17 Q50 J3-12 Q51 J3-15 Q52 J2-5, J3-13 Q53 J2-3, J3-14 Q54 J3-11 Q55 J3-9 Q56 J3-10 Q57 J3-1 Q58 J3-2 Q59 J3-4 Q60 J3-3

Testing the Lamp Driver SCR's, game On.
If a lamp is permanently stuck on, this procedure won't tell you anything. A lamp that is always on is generally caused because its SCR has internally shorted. Replace the SCR.

Assuming the game powers on, you can test a non-working lamp's SCR's to see if it's working (this assumes you have checked the bulb, the lamp socket, and the wiring to the lamp socket).

• While the game is on and in "attract" mode, press the Self-Test button inside the coin door ONCE. This should put the game into the "Flash All Feature Lamps" test (check your game manual if it does not).
• Note which feature lamps are NOT working. The Bally self test will flash ALL sixty lamps without exception. Write down which lamps do not flash. (You will need this information if several lamps that connect to the same decoder don't work. A decoder has likely failed if 4, 8 or 12 lamps, multiples of 4, are not working.)
• Check the manual's schematics to figure out which SCR number controls the lamp(s) in question. This information is on the Lamp Driver schematic page.
• Look at the connectors at the right of the schematic. There the lamp name/descriptions will be listed.
• Follow this line back to the first "Q" (SCR) that intersects this line. Note the SCR number (for example, "Q8"). If the schematic lists a "**" next to the SCR, this means it's a MCR106-1. Otherwise it's a 2N5060. Also note the chip that drives this SCR (U1 to U4). Both these components could be damaged (but generally it's just the SCR).
• Write down the "Q" number and the lamp name on some paper. Also write down the driving decoder "U" chip number.
• If 4, 8 or 12 lamps that all connect to a single decoder don't work, suspect the decoder "U" chip as faulty.
• Press the game's test switch again to take the game out of lamp test mode. The display test will probaby come up. Leave the game here, as all the playfield lamps should now be turned off.

• With the game in display test, connect an alligator test lead wire to ground. The bare braided wire in the bottom of the back box works well for this.
• Touch the other end of the test lead to the ANODE (A) of the SCR in question. On the larger MCR106, the metal face or metal tab is the anode. On the smaller 2N5060 SCRs, it's the lower right leg. To make sure, the pinout for the SCRs is silk screened on the board for a few selected SCR. Look for the lead marked "A".
• If the lamp does NOT light when the anode is grounded, the problem is NOT on the lamp driver board. Most likely you have a wiring problem, a bad lamp socket, or a bad bulb.

Lamp Always Off.


1. With the game on and in score display test mode, connect one end of an alligator jumper to Lamp Driver board TP3 (goes to R70 2k ohms).
2. Connect the other end of the alligator jumper to the GATE (G) of the SCR in question. On the larger MCR106, this is the left leg. On the smaller 2N5060, this is the right side (center) leg.
3. The lamp in question should light.
4. If the lamp does not light, the SCR is probably bad. Test the SCR with the power off using a DMM in diode test, as described below. If the SCR tests bad, replace it. Repeat steps above 1 to 3.
5. MCR106 and -14 lamp driver board only: If the MCR106 tests good on a -14 lamp driver board with buffer chips, replace the CD4050 buffer chip connecting to the MCR106 in question. Repeat steps above 1 to 3.
6. If the lamp still won't light, replace the 4514 decoder chip connecting to the SCR in question. Repeat steps above 1 to 3.

Lamp Always On.


1. With the game on and in score display test mode, connect one end of an alligator jumper to Lamp Driver board TP2 (ground).
2. Connect the other end of the alligator jumper to the GATE (G) of the SCR in question. On the larger MCR106, this is the left leg. On the smaller 2N5060, this is the right side (center) leg.
3. The lamp in question should turn off.
4. If the lamp does not light, the SCR is probably bad. Test the SCR with the power off using a DMM in diode test, as described below. If the SCR tests bad, replace it. Repeat steps above 1 to 3.
5. MCR106 and a -14 lamp driver board: now ground the input leg of the CD4050 connecting to the SCR in question. If the lamp does not go out, the buffer chip is bad. Repeat steps above 1 to 3.
6. If the lamp still won't go out, replace the U chip connecting to the SCR in question. Repeat steps above 1 to 3.

Testing the Lamp Driver SCRs POWER OFF.
You can also check the lamp driver board's SCR's using your DMM, set to the diode setting.

• Check the manual's schematics to figure out which SCR controls the lamp(s) in question. This information is on the Lamp Driver schematic page. Write down the SCR's "Q" number.
• Look at the connectors at the right of the schematic. There the lamp name/descriptions will be listed.
• Follow this line back to the first "Q" (SCR) that intersects this line. Note the SCR number (for example, "Q8"). If the schematic lists a "**" next to the transistor, this means it's a MCR106-1. Otherwise it's a 2N5060. Also note the chip that drives this SCR ("U1"). Both these components could be damaged (but generally it's just the SCR).
• You can remove the lamp driver board, or leave it installed in the game. Use your DMM set to the "diode" setting.

MCR106-1 Lamp Driver SCR test:

• Put the black lead of your meter on the outside "cathode" leg (labeled "C") of the SCR.
• Put the red lead of your meter on the outside "gate" leg (labeled "G") of the SCR. Your meter should read .4 to .6 volts.
• Swap the meter leads. Now the meter should read 1.4 to 1.6 volts.
If your meter reads anything outside the values above, replace that MCR106-1.

Testing the large MCR106-1 lamp driver SCR.

2N5060 Lamp Driver SCR test:
• Put the black lead of your meter on the "cathode" leg (labeled "C") of the SCR.
• Put the red lead of your meter on the center "gate" leg (labeled "G") of the SCR. Your meter should read .4 to .6 volts.
• Swap the meter leads. Now the meter should read 1.4 to 1.6 volts.
If your meter reads anything outside the values above, replace that 2N5060.

Testing the small 2N5060 lamp driver SCR.

Auxiliary Lamp Driver Boards.
Some games (like Kiss and 8 Ball Deluxe) have more CPU controlled lamps than the lamp driver board can handle. To accomodate these extra lamps, some Bally games use an Auxiliary Lamp Driver board. For example on Kiss, the marquee lights in the backbox for the "KISS" name are CPU driven by an auxililary lamp driver board. Often these boards can develop problems where some (or all) of the lamps it drives do not work.

The most common problem with the Auxiliary Lamp Driver boards are crack .156" header pins. Resoldering the header pins often fixes many problems. Or perhaps a light or two does not work - this is often one or more of the MCR106 silicon controlled rectifiers (SCRs). And lastly the 4013 CMOS chip (U1 on Kiss boards) again fails.

There are two different Auxiliary lamp driver boards that Bally used: AS-2518-43 (12 additional lamps) and AS-2518-52 (28 additional lamps). Below is a table showing which SCR goes to its corresponding connector number. Table thanks to S.Kupla.

The Auxiliary Lamp Driver board AS-2518-43 from a Kiss.

AS-2518-43 Aux Lamp Driver	AS-2518-52 Aux. Lamp Driver
SCR Connector Q1 J2-1 Q2 J2-2 Q3 J2-3 Q4 J2-5 Q5 J2-6 Q6 J2-7 Q7 J2-11 Q8 J2-12 Q9 J2-17 Q10 J2-20 Q11 J2-19 Q12 J2-18	SCR Connector Q1 J2-7 Q2 J2-4 Q3 J2-8 Q4 J2-10 Q5 J2-9 Q6 J2-6 Q7 J2-5 Q8 J2-14 Q9 J2-11 Q10 J2-15 Q11 J2-18 Q12 J2-17 Q13 J2-13 Q14 J2-12 Q15 J3-8 Q16 J3-3 Q17 J3-9 Q18 J3-11 Q19 J3-10 Q20 J3-7 Q21 J3-4 Q22 J3-15 Q23 J3-12 Q24 J3-16 Q25 J3-18 Q26 J3-17 Q27 J3-14 Q28 J3-13

Starting with Eight Ball Deluxe, Bally started designing games that had more solenoids than transistors to drive them on the solenoid driver board. To allow more coils to be used in the game, Bally came up with a "Solenoid Expander Board". This board is a solenoid multiplexer; it allows a single transistor on the solenoid driver board to drive two different devices, instead of just one. This is very similar to what Williams did with System11 (and DataEast) and their A/B relay.

The solenoid expander board's main component is a relay and a 555 lamp. This relay connects to several different coils. When the relay is pulled in, half of the coils can be activated by the solenoid driver board's transistors. When the relay is at rest, the other half of the coils can be activated by the solenoid driver board's transistors.

The real trick to the solenoid expander board is how its relay is pulled in. A relay is basically a small solenoid. To activate a solenoid you normally use a transistor on the solenoid driver board. But wait! We're trying to SAVE transistors on solenoid driver board because we're running out of them to drive all the coils on the game. (That's the whole reason why there's a Solenoid Expander board.)

So instead of using a solenoid driver board transistor to activate the solenoid expander board's relay, a LAMP driver SCR (silicon controlled rectifier) is used on the lamp driver board! This allows the lamp driver board to control which two of the four coils can be controlled by the solenoid driver board's transistor.

The solenoid expander board and its accompanying 555 lamp.

Another way to look at it is this: think of two transistors on the solenoid driver board, say transistors 11 and 12. These normally control two coils (call them coils 11 and 12). But now transistors 11 and 12 control FOUR coils (coils 11a, 11b, 12a, 12b). When the lamp driver board's SCR activates and pulls in the solenoid expander board's relay, the solenoid driver transistors 11 and 12 control coils 11b and 12b. When the lamp driver board and solenoid expander board's relay are at rest, the solenoid driver transistors 11 and 12 control coils 11a and 12a. Note all four coils can not be activated at the same time.

It's also important to note that the solenoid expander board usually controls little used coils, or coils that don't need to act fast. Stuff like the outhole kicker or drop target reset coils. These coils are used far less than say slingshot or pop bumper coils. So if your outhole kicker solenoid is acting up, the expander board may be the cause.

For example, on Eight Ball Deluxe, if at the start of the game the #1/9 drop target falls, but the ball won't kick to the shooter lane, that's a classic example of a bad Solenoid Expander Board. The SEB relay is sharing one driver board transistor for both the outhole kicker and the #1/9 drop target reset. If the SEB won't toggle its relay, the #1/9 drop target falls instead of the outhole kicker working.

But the SEB was also used in some extreme cases too. For example, take Centaur. The solenoid expander on this game is used so a lamp SCR can control the 'Kick Ball to Playfield' coil. This is the most extreme use of a Solenoid Expander, as this is a very high current application (which can cause problems with the SEB, from burnts relay contact points to cracked SEB header pin solder joints).

Here's the 8 Ball Deluxe schematic showing the SEB. Notice the outhole kicker and the 1/9 drop target reset coil share the same driver board transistor thru connector A3J5-12. Which coil is energized by the driver board is determined by the position of the SEB's relay. Notice six coils connect to the SEB on 8 Ball Dlx.

The Solenoid Expander Board's Partner.
The other strange thing about the solenoid expander board is its partner; a 555 lamp! The solenoid expander board is located under the playfield, and right next to it is a lamp. This lamp is VERY important, and it must have a good light bulb installed for the solenoid expander board to function properly.

The reason for the lamp is this; the solenoid expander board has a MOC3011 opto-isolator that actually turns the relay on. However this device doesn't draw enough current for the SCR on the lamp driver board to activate it. To solve this problem, an actual lamp is installed next to the solenoid expander board. This lamp is also connected to the lamp driver board's SCR. The combination of this lamp and the solenoid expander board's MOC3011 gives the lamp driver board's SCR enough current draw for it to work reliably.

Cracked Header Pins on the SEB.
A very common problem with the SEB are the male .156" molex header pins on the SEB's board. Often the solder joint cracks around the pin, giving intermittent connections. Resoldering these pins will fix this problem.

Dim CPU Controlled Lights and the SEB.
If all your CPU controlled lights are dim, this can cause problems with the Solenoid Expander Board (SEB). For example if the SEB relay "buzzes" instead of "clicking", that's indication of a problem. If the CPU controlled lights are dim, I highly suggest replacing the 6 volt bridge rectifier on the transformer's rectifier board. On pre-Xenon games especially, dim CPU controlled lights can be caused by an over-stressed and failing 6 volt bridge rectifier for the CPU controlled lights. Always use a 400 volt 35 amp bridge as a replacement. With a good strong 6 volts DC, the SEB will work better (and all your CPU controlled lights will be brighter too).

Testing the Solenoid Expander Board.
The easiest way to test the solenoid expander board is to use the built in feature lamp diagnostics. After the game is turned on and in attract mode, press the red test button inside the coin door once. This will activate the feature lamp test. All the playfield lamps will turn on and off. The solenoid expander boards relay should also click on and off. If you lift the playfield you should see the solenoid expander board's relay pull in. You should also see the lamp next to the solenoid expander board turn on and off.

If you heard the relay buzz instead of click on and off, this means there is a problem. See the list of problems below to fix this.

Two Versions of the SEB.
Note there are two variants of the solenoid expander board, depending on the game it came from. One version has a jumper, and one version has the jumper cut. If the jumper version is installed in a game designed for the non-jumper version (like say Centaur), the jumper will short the solenoid voltage (43 volts) directly to ground. This will immediately blow the solenoid fuse.

Problems with the Soleniod Expander Board.
If the solenoid expander board buzzes in the above test, this indicates a problem. Also it is important to remember that every coil that connects to the solenoid expander board has TWO diodes; one "normal" diode across the coil lugs, and another diode in series with the power wire going to the coil! If this diode set up is mis-wired, a diode is reversed, missing or broken, you will have problems. Here are some other things to consider too:

• Header pins on the solenoid expander board may have cracked solder joints. Re-flow these header pin solder joints.
• The lamp under the playfield next to the solenoid expander board may be burned out. Replace with a new lamp. Without this lamp, there isn't enough load for the SCR on the lamp driver board to activate the solenoid expander board's relay.
• The lamp driver board's SCR that controls the solenoid expander board may have failed. See Locked-On or Not Working Feature Lights (lamp driver board) section for details on how to test the SCR.
• The 6.5 volts provided by the power supply's rectifier board BR1 bridge may be failing. The 6.5 volts DC provided by BR1 powers the feature lamps. If this voltage is low, the solenoid expander board's relay may not function correctly.
• The power supply's rectifier board R2 resistor (25 ohms 5 watts) may be broken or open. Check it with your DMM set to ohms.
• Connector pins on the power supply's rectifier board may be burned or tarnished. This will create resistance which will lower the 6.5 volts DC provided to the lamp driver board. If this voltage is low, the solenoid expander board's relay may not function correctly.
• Diode on the solenoid expander board's relay may be broken or missing.
• In-series diode going to the power lug of the coil may be broken, missing or reversed.
• Diode across the lugs of the coil may be broken, missing or reversed.
• Switch contacts on the solenoid expander board's relay may be pitted and may need to be filed (or the relay replaced).
• The solenoid expander board's MOC3011 opto-isolator may be bad. Replacements are available from Jameco (part #95020), or replace with NTE3047.

Testing the Transistors and Coils Driven by the Solenoid Expander board.
If the solenoid expander board's relay is working properly, you can test the devices it controls using this test:

• Check the schematics to see which transistor number is controlled through the solenoid expander board.
• Turn the game on and let it boot.
• Press the red test switch once to put the game into diagnostic feature lamp test.
• Using an alligator clip test wire, attach one end to the ground strap in the backbox.
• Touch the other end of the alligator clip test wire to the metal tab on the solenoid driver board's transistor you identified in the first step above.
If you leave the transistor's metal tab grounded while the game is in the feature lamp diagnostic test, the two coils controlled by the above transistor should turn off and on with the feature lamps.

Solenoid Expander Board (SEB) Relay Replacement.
The relay on the SEB are 48 volt relays, and are somewhat expensive and hard to find (this is the same type of relay used on the Solenoid driver board for the flipper relay). But these can be replaced with an inexpensive relay of a different (and more modern) style. The relay is Jameco part# 301065CJ ($16.35) and is Tyco/Potter & Brumfield part# KRPA-11DG-48. This new relay will require a socket, which can be mounted next to the SEB. The socket is Jameco part# 282132CJ ($3.95) and is Potter & Brumfield part# 27E122. The relay socket fits nicely to the bottome of the playfield next to the solenoid expander. There are eight pins (labeled 1-8) on the new relay socket, and the pictures below (thanks to M.Kelly) show how to wire the new relay.

The solenoid expander board modified to use the inexpensive Tyco 48 volt relay. Shown is how the new relay socket is wired to the SEB and mounted under the PF. Pic by B.Kelly.

Later solid state machines by Bally sometimes used an auxiliary lamp driver board to control additional lamps. There were two flavors of this board: AS-2518-43 (12 additional lamps) and AS-2518-52 (28 additional lamps) .

Left: the earlier style AS2518-43 Auxiliary Lamp Driver board.	Right: the later style AS2518-52 Auxiliary Lamp Driver board.

On Bally games with special backbox lighting (Space Invaders, Xenon, and many others), Bally used an auxiliary lamp driver board to run these lights. This saves the lamps on the lamp driver board for use on the playfield. This auxiliary board isn't much different than the lamp driver board itself. It used SCR (silicon controlled rectifiers) just like the lamp driver board. It uses only MCR106-1 devices (no 2N5060 SCR's).

If a lamp is locked on or not working that is controlled by the auxiliary lamp driver board, the procedure for testing and replacing its SCR on the auxiliary lamp driver board is the same as the lamp driver board. Please see lamp driver board section above for more details.

All electronic pinball machines use a technique to read switches called a "switch matrix". Problems with the switch matrix can be difficult to diagnose. The matrix typically consists of eight rows and eight columns, but Bally games use eight rows (returns) and five columns (strobes ST0-ST4) for a total of 40 switches (8x5=40). This makes a checkerboard type pattern with 40 individual switch (or squares). So instead of 40 wires going to 40 switches, there are only 8+5=13 wires going to all the switches.

Bally identifies the switches in the switch matrix starting with row zero, column zero as switch "1", so the switches in column 0 are numbered 1 through 8, column 1 are numbers 9 to 16, and so on. If you look at the schematic it can appear some switches are missing. But most of these "missing" switches are the cabinet switches, which are not shown on the schematics. This includes the the tilt, slam, credit switch, three coin switches and test switch.

Note on some games Bally used a sixth switch column (strobe, aka column5) called ST5, allowing an additional eight switches numbers 41-48. Only some games used this (Spectrum, Centaur, Medusa, Fathom). Normal switch operation is handled by strobes ST0 to ST4 (columns 0-4), coming from MPU connector A4J2 pins 1-5 respectively, and going to PIA U10 port lines PA0-PA4 respectively (U10 pins 2-6). The seldom used sixth switch strobe ST5 comes in at MPU connector A4J4 pin 5, which goes to PB4, a different PIA port at U11 pin 14 (this PIA port is usually used for the continual solenoid data). The return for switch line ST5 is a tie to +5 volts at A4J4 pin 16, thru a playfield mounted 3.3k ohm pull-up resistor.

Switch Rows and Columns.
Here are the switch rows and columns as laid out on the MPU board at connectors J2 and J3. Connector J2 is the entire switch matrix, and connector J3 is just used for the slam, tilt, credit, test and coin switches. Remember "column" is the strobe, and "row" is the return.

• Column 0: J2 pin 1
• Column 1: J2 pin 2
• Column 2: J2 pin 3
• Column 3: J2 pin 4
• Column 4: J2 pin 5
• Column 5: J4 pin 5 (rarely used)
• Row 0: J2 pin 8 / J3 pin 9
• Row 1: J2 pin 9 / J3 pin 10
• Row 2: J2 pin 10 / J3 pin 11
• Row 3: J2 pin 11 / J3 pin 12
• Row 4: J2 pin 12 / J3 pin 13
• Row 5: J2 pin 13 / J3 pin 14
• Row 6: J2 pin 14 / J3 pin 15
• Row 7: J2 pin 15 / J3 pin 16

Testing the MPU Switch Matrix.
The game's switch self test is very handy for testing the switch matrix. Here's a handy procedure that disconnects the MPU from the playfield so any switch matrix problem can be isolated to the playfield or MPU board.

1. With the game off, turn OFF all MPU board DIP switches.
2. Power the game on.
3. Put the game into its switch test. On games with sound cards (Lost World and later), this means pressing the red self-test button inside the coin door FIVE times. For chime games prior to Lost World you only press the test button FOUR times. If MPU connector J3 is removed from the MPU board (disabling the coin door's test button), touch J3 pin 1 to ground four or five times to enter the switch test (each grounding of J3 pin 1 is like pushing the red self-test button).
4. While still in switch test unplug MPU board connectors J2 and J3.
5. Using an aligator clip test lead, attach one end of the test lead to MPU connector J2 pin 1.
6. Connect the other end of the aligator test lead to MPU connector J2 pins 8 to 15 (one at a time), as these are the switch rows. A switch number will show up on the score displays for each of the eight J2 pins.
7. Connect the aligator test lead to MPU connector J3 pins 9 to 16, again these are the switch rows. A switch number should show up on the score displays for each of the eight J3 pins.
8. Move the aligator jumper from MPU connector J2 pin 1 to J2 pin 2 and repeat the above two steps. Continue the test up to J2 pin 5. These are the switch columns (remember Bally only has five switch columns).
9. Power the game off and set all the DIP switches to ON and repeat the above test steps 1 to 8.

If there is a switch matrix problem as indicated in the above test (one pin that does not work), generally it is related to a a J2/J3 connector issue or bad PIA chip/socket. Also check all the resistors and diodes (1N4148) and capacitors (390 pf) connecting to J2 and J3, and look for a broken trace going from the J2/J3 connectors to the resistors/diodes/caps, and then to the PIA chip. The PIA can be easily determined as good or bad - just swap U10 and U11 and see if the switch problem changes. If it does, PIA U10 is bad. If not, there is some other problem.

Stuck Switches.
If a switch is stuck, first remove all the balls from the playfield (to open the trough switches). Also if a switch gets closed as you lift a playfield up, put a piece of tape between the switch contacts to prevent closure. This especially happens on spinner targets.

Put the game into its switch test. On games with sound cards (Lost World and later), this means pressing the red self-test button inside the coin door FIVE times. For chime games prior to Lost World you only press the test button four times. Note the switch number shown in the display. This will be the last switch "bad switch" in the list. Use your game manual to find the actual playfield location of the switch. Clear this switch first, then the next "bad switch" will be shown on the display. Your job isn't over until there are no more switch numbers shown in the display.

Switch Capacitors - Stuck Switches and Chattering Pop Bumpers.
The most common stuck switch problem relates to the capacitor on playfield switches. Note all playfield switches have a capacitor, but most do. Often these capacitors short, leak, or just break off. With the game on and in switch test, cut one end of the capacitor and see if the number goes away on the display in switch test. If so, the capacitor will need to be replaced. If the number doesn't go away, then reconnect the capacitor, as this wasn't the problem (check the switch diode next, in the same manner).

Note Bally never sent out a service bulletin telling operators to "clip the caps". Many operators did this though to get their game working. Bally only suggested clipping the caps to trouble shoot a stuck switch problem.

Bally used really cheap switch capacitors that failed prematurely. This probably happened because of the high power industrial soldering irons used on the assembly line. The caps used for the switches were actually made for circuit boards. As a result the high powered soldering irons on the assembly line weakened the cap's internal insulation, causing it to leak (not function reliably). This could result in a stuck switch indication. The other problem was the capacitors' leads were weakened from the high powered soldering irons. This caused the caps to break off the playfield switches.

Chattering pop bumpers is another problem caused by bad switch capacitors. A failing switch cap across the pop bumper switch is very common (due to vibration of the bumper), and can cause the switch to "bounce". This will can make the pop bumper fire more than once. Cut the cap off the pop bumper, and see if that fixes the problem. But make sure to replace the capacitor. It is there to help the game detect quick switch closures, and effectively makes the pop bumpers play better (more sensitive).

The original switch capacitors were .05 mfd or .047 mfd, 16 volt (or greater), ceramic disc, non-polarized (the caps are handling 5 volts, so any voltage will work). The most common, easiest to find replacements today seems to be .047 mfd 50 volt non-polarized caps.

Non-Working Switches.
There can be a few reasons for a missing switch in the switch test diagnostics. If it is only one switch, it could simply be a broken wire or dirty switch contacts. If it's more than one, it could be a broken wire at the top of that switch string. Switches are "daisy chained" together. That is, a row or column wire runs to the first switch in the series, then continues along to the next. If the wire is broken, all the switches behind that break can be affected.

Connectors can also be a problem. All switches run to the MPU board's J2 connector (upper right hand corner). Connector pins can break, or the wire inside the connector can break.

Switch Diodes - Stuck and Flakey Switches.
Just as the switch capacitors can fail, so can the switch diodes. Though not as common a problem, these small diodes can quite often fail or work intermittently. Unlike the switch capacitor, every switch must have a diode. If the diode is missing or fails, the switch will not work correctly.

Switch diodes are 1N4148 (or 1N914), which are known as fast "switching" diodes. Also a 1N4001 or 1N4004 can be used in a pinch. And it does matter which way the band of the diode is installed!

With the game off, you can test a switch diode by unsoldering one lead of the diode, and use your DMM (set to diode setting), and test the diode. In one direction you should get .4 to .6 volts. Reverse the leads and you should get a null or zero reading.

You can also test a diode with the game on and in switch test mode. Do this by jumping the two switch wires together (before the diode). If the switch registers in the self-test, the diode is bad.

What are the Diodes for?
The diode on each switch isolates that switch from the other switches in the switch matrix's column. The diodes "steer" the signal in one direction only. A shorted, broken, leaky or missing diode can havoc in the switch matrix. A broken diode would cause a switch to never be "seen" by the MPU (a broken wire will cause this too!). A shorted diode will violate a switch's uniqueness, and allow the switch strobe signal to show up at the wrong times, and for the wrong switches! Leaky diodes can act some of the time like a shorted diode, and other times be OK. This is the hardest diode problem to locate, but luckily doesn't happen too often.

Using diodes allows the switch matrix to use only 13 wires for 40 switches. The matrix is made up of 8 rows and 5 columns, allowing for a total of 40 switches. But a problem exists in Bally's design; they don't use diodes on every switch! For example the coin door switches do not use diodes.

The reason Bally did this was they felt diodes were not needed on the coin mech switches. Since a coin mech switch would only be activated by itself (and never during game play with the playfield switches), they were not needed. Also some early Bally electronics game playfield switches are also missing diodes. These were omitted because it was thought the ball could only activate one switch at a time.

Missing Diodes: the Problem.
If any of the switches without diodes get stuck closed (for whatever reason), this sends a constant signal to the MPU which totally confuses it. This can cause all kinds of other random game play problems. Bally finally reconized this problem, and put diodes on ALL switches by the mid-1980's. If your game is missing any switch diodes, I recommend installing a 1N4148 or 1N914 or 1N4001 diode on the switch. Orient the band on the diode as it is installed in the other switches.

Under the playfield: The bare slingshot switch wires can break and easily short to the slingshot plunger spring (ground), confusing the switch matrix (like a slingshot switch causes the game to tilt). Or worse, the bare sling wires can touch the coil lugs sending 43 volts down the switch matrix, frying the MPU board PIA chip(s).

Switch Matrix Problems caused by the Coin Door Credit Button.
The start (credit) button is insulated from the grounded coin door with a piece of gray "fish paper". This paper stops the start switch from being permanently grounded. Often this paper gets worn, ripped or torn. This will cause the start button's entire switch strobe (row) to be grounded. This can make your Bally/Stern game do some weird stuff!

If your game has been broken into and the coin door pried open, the start button's fish paper can be easily damaged and causing a switch matrix problem.

For example, one strange problem seen from the start button being permanently grounded was the player one display always showed the same information as the credit/ball display. If the start button gets grounded, other strange game behavior can happen too like a confused game that just won't play right.

If you think the start button may be causing a problem, turn the game off and unplug the coin door's connector plug. Then turn the game back on and see if that clears the problem.

Switches as Interpretted by the MPU.
When a switch is closed, a row and column is crossed. This information goes to the MPU through the wiring harness to the MPU connector J2. Then the U10 PIA chip interprets this signal and directs it to the U9 CPU. Now scoring or other game conditions occur.

The information can get sent to the MPU properly, but a corroded MPU board can mis-direct the information. If acid damage gets underneath connector J2, this can cause havoc. Also connector J3, which at pins 2 and 3 are the shared strobe signals from J2, can be shorted together from corrosion damage and cause more havoc.

MPU Board DIP Switch Problems.
There is another area where the switch matrix can go crazy. This is at the MPU board's 32 mini DIP switches. These switches too have diodes, but if one diode goes bad, the DIP switches can be mis-interpretted, and this can cause havoc in the switch matrix.

If you suspect the MPU board's DIP switches as a problem, just turn them all OFF. This essentially removes them from the switch matrix. Then the DIP switch's diodes can be individually tested using your DMM (set to the diode setting). With the game off, you should get a reading of .4 to .6 for each diode, and when the DMM leads are reversed, a zero or null reading.

Multi-Ball Games and what is Required to Start a Game.
Some Bally multi-ball games require ALL the pinballs to be in the right place before a game will start. If a pinball is stuck or missing, often the game will not start. Sometimes a wire coming off the trough switch can also cause this problem. In either case the game appears to be broken, when the fix is really just unsticking a ball or re-attaching a trough switch wire. Here are the games that are effected by this:

The above switches can be checked in diagnostic switch test. On games with sound cards (Lost World and later), this means pressing the red self-test button inside the coin door FIVE times. For chime games prior to Lost World you only press the test button four times. If the above switches are mis-read by the MPU, this can cause a game to not start, more than one ball being served to the shooter lane, no multi-ball, or some other strange behavior.

Wrong Switch Numbers Showing In Diagnostics.
On a reader's Flash Gordon, was going through the self tests before playing it. On the switch test, none of the switches showed up as stuck, so started closing each switch in turn to make sure it showed up as closed on the display. Got to one column of switches and they all registered but as the wrong switch numbers. I wasn't sure if the problem was with the MPU board or the wiring/playfield so I pulled the J2 connector on the MPU and jumpered the appropriate strobe line to the return lines one at a time. The previously bad switch numbers showed up correct, indicating the MPU board was OK. Some switch matrix lines run through J3 as well, so it's a good idea to disconnect J3 before doing this.

The other thing to try is to test for continuity at the J2 connector between each of the strobe and return lines, and every other strobe and return line, and between each of these lines and ground. With all the switches open, any continuity in any of these tests indicates a short. In my case, two of the strobe lines were shorted together under a thumper bumper bracket. With all the switches open none of the switch matrix lines at the J2/J3 connectors should have continuity with each other or with ground.

The high voltage section of the Solenoid driver board supplies high voltage to the gas score display tubes to light them.

High Voltage Warning.
WARNING! The high voltage circuit contains a large capacitor at C26 which can retain high voltage at dangerous levels. This capacitor will bleed off this high voltage to a relatively safe level after the game has been off for a few minutes. Note this voltage can also be dangerous to the integrated circuits on the solenoid driver board. If you short this cap accidently to the logic circuit, it can ruin those components. Also 155 to 220 volts is high voltage and can do harm to humans!

Adjust the High Voltage down to +170 Volts DC.
Bally and Stern games from 1977 to 1985 have an adjustable high voltage potentiometer ("pot") on the Solenoid driver board. This pot is located in the upper left corner of the board, beneath capacitor C26, and to the left of capacitor C23. This pot allows you to adjust the voltage going to the score displays for optimum voltage. Bally recommends this voltage be adjusted to +190 volts DC. But for increased display life, it's best if you adjust the voltage to +170 volts, or as low as +155 volts. Brightness generally won't be effected since there is a minimum voltage required for the gas inside the score displays to glow. Set your high voltage to whatever value it takes to adequately light all of your score displays, then turn it up just a few volts higher. Keeping this voltage as low as possible will greatly extend the life of these gas display tubes, and the display drive circuits.

Exception to the +170 volt High Voltage Rule.
When installing brand NEW score displays, it may require the full +190 volts DC to light the displays for the first few hours. After the new displays become "seasoned", you can back the voltage down to +170 volts or lower.

The high voltage adjustment pot on the Solenoid driver board AS2518-22. This is the newer SDB with the fuse for the High Voltage. The plastic protector shield is still in place.

The High Voltage (HV) section of a Bally Solenoid Driver Board AS2518-16 (left corner). Note the TP2 and TP4 test points, and the blue trim pot to adjust the voltage. From the factory this area is covered with a flat plastic sheild, which is often removed and lost.

Measuring the High Voltage at TP2 & TP4:
TP2 & TP4 should NOT measure the same!
The high voltage should be measured on the solenoid driver board. Use a DMM on DC volts, and put one lead on the solenoid driver board's TP2 test point. Put the other lead on ground (there is a TP labeled "GND" on the solenoid driver board). This will measure the output high voltage from the solenoid driver board, and should be in the 155 to 190 volt range. Compare this to the measurement of GND and TP4, which is the input voltage to the HV circuit. If TP4 and TP2 measure the same (typically about 220 to 230 volts), the HV circuit is dead and not working! Yes the score display may still be working. But a full 220 to 230 volts is being passed to score displays that only require 155 to 190 volts to operate. This is obviously not a good thing, and it WILL burn your score displays if left unrepaired!

Bally was asked why they designed their HV circuit so when it failed it would "short" instead of going "open" (an open circuit would mean the score displays would not work). They said they had a lot of discussion about this, and decided that operators were smart enough to see the displays burning too bright, and would fix the problem. Unfortunately Bally was wrong about this - most operators did not notice (or care), and left 230 volts going to the 190 volt score displays. This burned the displays quickly and basically ruined them.

A burnt score display - this was caused by not repairing a faulty HV circuit.

What If the Score Displays Do Not Work at All?
First thing to check is fuse F2 on the power supply regulator board. This is a 3/4 amp slow-blow fuse. If it is blown, power will never get to the high voltage section on the driver board.

Typically if regulator board fuse F2 is blowing, one of the 1N4004 diodes on the regulator board has gone open or has shorted. This is a very common problem. Because of the fuse, most times the HV components on the driver board are saved without damage. You can test the four 1N4004 diodes with a DMM set to the diode range. Put the leads on each of the four diodes, and .5 should be seen in one direction and a null reading in the other.

Rebuilding the High Voltage Section.
If your high voltage section of your Solenoid driver board is not working, you may need to rebuild it. Assuming regulator board fuse F2 has been checked and the four regulator board 1N4004 diodes are good, there is a good chance the HV section of the driver board has failed. Check the voltage on the driver board at TP2. If 230 to 250 volts DC is seen, the high voltage section has failed and will need to be rebuilt.

Rebuilding the HV secion is pretty easy to do, and fairly inexpensive. But you should install ALL the parts mentioned below. The design of this circuit was very simple to keep costs low. There is no protection against multiple part failures. Often the parts in this section fail in pairs. So just replacing one will burn out the new part in a short order. So replace all the parts in this circuit to ensure reliability. Steve Kulpa also sells a very nice Bally/Stern high voltage repair kit at bigdaddy-enterprises.com or Great Plains Electronics, which I highly suggest.

High Voltage Rebuild Parts Needed.


• 2N3584 transistor at Q21 (250 volts, 2 amp, TO-66 NPN). Mouser sells these.
• (2) 2N3440 transistors at Q22, Q23 (250 volts, 1 amp, TO-39 NPN).
• 1N5275A zener diode at VR1 (140 volts, 1 watt).
• 1N4004 diode at CR21 (400PIV, 1 amp).
• 25k ohm potentiometer PC mount at RT1
• 3/16 amp fast blow fuse at F1. This is the short fuse (but one end of the fuse holder can be moved to accomodate a standard sized fuse). Note earlier Stern and Bally AS2518-16 solenoid driver boards do not use this fuse.

• 22k ohm 1/2 watt resistor at R51 (often burned).
• 82k ohm 1/2 watt resistor at R56 (often burned).
• 1.2k ohm 1/4 watt resistor at R55.
• 8.2k ohm 1/4 watt resistor at R54.
• 390 ohm 1/4 watt resistor at R52.
• 100k ohm 1 watt resistor at R35.
• (2) .01 mfd 400 vdc metal polyester capacitors at C27, C28.

Definately replace all the transistors and diodes and capacitors specified above, regardless if they test good. You can probably get away with not replacing all the resistors, but make sure you check them all with your DMM. If more than 5% out of spec, replace. It is very common for a resistor to go open in this circuit. The 25k trim pot is of low quality and should probably be replaced. Usually the 3/16 amp short fuse will be blown too (do the modification described below so a normal length fuse can be used). Remember to install the new resistors and transistors slightly above the circuit board to allow a good air flow. Re-use any transistor spacers from the old transistors; they help prevent solder shorts.

High Voltage Still out of Spec?
If the high voltage section still measures out of spec, try replacing C26 (160 mfd 350 volts) on the solenoid driver board. This filter capacitor can wear out, causing the high voltage to be too high or too low.

The 2N3584 Transistor.
Often the high voltage 2N3584 transistor just can't be found. If your high voltage section is showing really high voltage (like 230+ volts), or low volts (150-190 volts), often the 2N3584 and the 2N3440's have failed. The 2N3440's are easy to get. So what can be used to replace the harder to get 2N3584?

For those in North America, Mouser.com sells the 2N3584, part# 610-2N3584. It has a TO-66 case and fits right in. Apparently overseas the 2N3584 is hard to get, as reported by Mark Hurry, so he wanted to come up with a replacement. The replacement transistor is the BUX84. This transistor has a TO-220 style case (instead of a TO-66 case). The heat sink's base/emitter hole may need to be enlarged to make sure there is plenty of clearance for the new BUX84's base and emitter leads. The BUX84's metal tab, which bolts to the circuit board, provides the transistor's collector connection. Because of this, the BUX84's center lead does not need to be used. Remember to also insulate the heat sink from the transistor's (collector) metal tab. Use a nylon washer between the transistor's metal tab and the heat sink. Also use a small piece of heat shrink tubing to go over the bolt that holds the transistor to the heat sink.

When replacing any components in the high voltage section remember to check all the transistors before powering on. Because this high voltage transistor is direct coupled amplifier, when one component fails it usually takes out the rest in a chain reaction.

Replacing the 2N3584 with a BUX84. Note the center lead was cut off and not used. Picture by Marco R.

Testing Your Work.
After the new parts are installed, set the high voltage adjustment pot to the minimum value (turn all the way to the left, counter-clockwise). If the solenoid driver board is the later design, do NOT install the the solenoid driver high voltage fuse!

Now turn the game on and measure the high voltage at TP2 on the solenoid driver board. Now adjust the voltage to +170 volts (or lower, if you can).

Now turn the game off and wait a minute or two for the voltage to bleed off. Re-install the solenoid driver board's 3/16 amp fuse at F1.

Finally, turn the game back on and press the red test button switch inside the coin door twice to run the display test. Let this test run for a few minutes, and make sure all the displays work properly. Check the high voltage again at the solenoid driver board's TP2 to verify everything is Ok.

The "Short" High Voltage Fuse.
The fuse used on the AS2518-22 solenoid driver board for the high voltage section is a smaller 8AG fast blo 3/16 amp style fuse. These are sometimes difficult to find in the smaller 8AG style. If you can't find this fuse, you can solder in a normal 1 1/4" fuse length fuse holder, and use the easier to get 3AG (1.25") fuse.

Replacing the small .75" 8AG fuse with a larger 1.25" fuse. If you can't find a 3/16 amp .75" 8AG fuse, you can solder in a normal size fuse holder in it's place (or move the existing fuse clip, as described below). The larger 1.25" long fuses are much easier to find.

The best approach is to move the top portion of the existing small fuse holder. The existing upper section of the high voltage fuse clip can removed, and repositioned upward, towards the edge of the circuit board, using one of the existing holes. Only one hole will hold this fuse clip in place, but it can now accomodate the larger style fuse easily. Then a jumper wire can be run from the bent up moved upper fuse holder tab to the former fuse hole on the circuit board, to complete the fuse connection (see picture below).

Moving the upper part of the fuse clip to accomodate a larger style fuse. Picture by Jerry.

Some Driver Boards Don't Have a HV Fuse.
Note the earlier AS2518-16 solenoid driver board does not have this fuse. Also the earlier solenoid driver board AS2518-16 (without the fuse) is directly interchangable with the later AS2518-22 solenoid driver board (with the fuse).

Much of this information and pictures came from Steve Kulpa (stevekulpa at yahoo.com) who has some good repair info at www.geocities.com/stevekulpa/bally_disp6.htm. Steve also sells a very nice Bally/Stern score display repair kit at bigdaddy-enterprises.com/repairkits/bally_kits.htm, which I highly suggest.

Bally used two different versions of the 6-digit score displays, the AS-2518-21 and the AS-2518-15. The Bally 7-digit display was part number AS-2518-58. Stern also made their own six digit diaplay board called the DA-100. Functionally all of these display boards are the same and are interchangable. The score displays on these older Bally games can be very tempermental. Here are some tips to help keep them running.

How They Work.
As Steve Kupla describes, each score display has six digits, and each digit consists of seven segments. There are two main parts of a display assembly: The glass score display, and the circuit board which drives the display. The circuit board has a input decoder, a six digit driver circuit, and seven segment driver circuit. There is one digit driver circuit for each of the six digits, and one segment driver circuit for each segment of a digit. The decoder chip takes a number from 0 - 9 as input from the MPU board and determines which segments need to be energized in order to represent this number.

The decoder is the only chip on the Bally/Stern score display circuit board. It's at location U1 and Bally's schematics says it is part number E-620-38, which means nothing to us. In reality it is a MC14543LE, called a "BCD To Seven Segment Decoder". BCD stands for Binary Coded Decimal and is a fancy term for storing a number from 0 - 9 in a half byte of storage (four bits). The input of the decoder is a BCD number from 0 - 9, and the output of the decoder is seven signals. These seven signals are either on or off, and relate to the seven segments of a digit. Each of the seven segment output signals go to a MPS-A42 transistor, which is part of a circuit called a segment driver. This transistor acts like a switch to turn the segments on or off. The outputs of the seven segment drivers go to the seven segment pins of the display glass.

The digit driver circuit consists of an MPS-A42 and a 2N5401 transistor connected in a circuit that acts as a switch. Normally with no input signal applied, the switch is off, keeping the high voltage supplied by the HV Regulator away from the display. There are 6 digit signals provided by the MPU, one for each digit. The MPU will enable one signal at a time, telling the display driver which digit to operate. This signal will then turn on the "switch" for the digit, allowing the high voltage a connection to the proper pins of the glass display to energize the desired digit. These signals from the MPU are simply 6 wires and the MPU will activate one of them at a time. Like the segment signals, the six digit signals go to all of the displays in a dasiy chain fashion.

Burnt Score Display Glass.
Often the score display glass itself is in poor condition, or just completely worn out. There is no way to fix the score glass itself, and these must simply be replaced. Often worn out score glasses are known as "outgassed". This is because of the high voltage involved with score displays, the anode and/or cathode inside the display glass breaks down. This results in the "outgassing" of impurities that eventually change the internal gas properties, so the display can't glow (the gas must be very pure for the display to work). Often the segments that don't light up at power-on will gradually come on as the display warms up. This happens because as the existing gas warms up, it expands. A new display glass will solve this problem.

Most commonly seen are burnt ones and tens digits on Bally score glasses. These score positions are alway in attact mode and game play, so they tend to burn first. Often these digits can be seen as burnt with the game power off!

A burnt score glass. The red circles show beginning problems. Most people can live with this. But this is just the beginning stages. Soon enough this display glass may "out gas" and not light up at all. Picture by Steve.

A burnt score display - this was caused by not repairing a faulty HV circuit. The burn is so bad it can be seen without powering on the display.

Preliminary Repair Step One:
Measuring the High Voltage on the Solenoid Driver Board at TP2 & TP4.
TP2 & TP4 should NOT measure the same!
The high voltage should be measured on the solenoid driver board. Use a DMM on DC volts, and put one lead on the solenoid driver board's TP2 test point. Put the other lead on ground (there is a TP labeled "GND" on the solenoid driver board). This will measure the output high voltage from the solenoid driver board, and should be in the 155 to 190 volt range (dial it to 170 volts for best score display life, using the adjustable trim pot). Compare this to the measurement of GND and TP4, which is the input voltage to the HV circuit. If TP4 and TP2 measure the same (typically about 220 to 230 volts), the HV circuit is dead and not working! Yes the score display may still be working. But a full 220 to 230 volts is being passed to score displays that only require 155 to 190 volts to operate. This is obviously not a good thing, and it WILL burn your score displays if left unrepaired!

Bally was asked why they designed their HV circuit so when it failed it would "short" instead of going "open" (an open circuit would mean the score displays would not work). They said they had a lot of discussion about this, and decided that operators were smart enough to see the displays burning too bright, and would fix the problem. Unfortunately Bally was wrong about this - most operators did not notice (or care), and left 230 volts going to the 190 volt score displays. This burned the displays quickly and basically ruined them.

The High Voltage (HV) section of a Bally solenoid driver board (upper left corner). Note the TP2 and TP4 test points, and the blue trim pot to adjust the voltage. From the factory this area is covered with a flat plastic sheild, which is often removed and lost.

Preliminary Repair Step Two: Check MPU Board Connector J1.
MPU connector J1 (upper left hand corner) is the connection between the MPU board and the score displays. If battery corrosion was a problem on the MPU board, this connector may have problems. If any of the score displays are not working at all or are missing digits, start at this connector first. A good close up examine of the MPU board's J1 header pins should indicate if there are any problems. If there is any doubt, as a quick diagnosis gently run some 400 or 600 grit sandpaper over the face of the J1 header pins. Install the J1 connector on the MPU board and power-up. If some of the display problems went away, power-off and replace the MPU J1 connector header and terminal pins.

Preliminary Repair Step Three: Reflow the Score Display Header Pin Solder Joints.
On all of the score displays in a game, even if they seem to be working fine, reflow the solder on the joints on the score display board's header pins. On the bottom side of the circuit board heat each header pin solder joint with a soldering iron and add some fresh solder. (To really do this correctly, the old solder should be "solder sucked" off the pins, and new solder added to replace it.) Be careful to not create any solder bridges between pins, unless there is already one there.

A 6 digit Bally display AS2518-21. The red circles show the 100k ohm resistors which need to be upgraded to 1/2 watt versions. Picture by Steve.

A 6 digit Bally display AS2518-15. The red circles show the 100k ohm resistors which need to be upgraded to 1/2 watt versions. Picture by Steve.

A 7 digit Bally display.

A 6 digit Stern display DA100. Picture by Steve.

A 6 digit Stern display, unknown part number. Picture by Steve.

Stern's DA-300 seven digit display. Note the nice silkscreening on the board showing what parts do what.

Component/board layout for the Bally AS2518-21 six digit display is available here.

Preliminary Repair Step Four: Check or Replace the 100k Ohm Score Display Resistors.
A very good idea is to check the score display board's six (or seven) 100K ohm (brown-black-yellow) 1/4 watt resistors. If any of the 100k resistors fail, an entire digit will not work on that display. If a digit is not working on a score display, a 100k resistor which has gone open is the likely cause. The original 1/4 watt versions at R1, R3, R5, R7, R9, R11 (R56 too on 7 digit displays) are too small and often overheated. This causes them to change their resistance value, or to go "open" and not work at all. See the pictures above for the locations of these resistors.

More Display Connector Information.
As Steve describes, check all the display driver connectors for loose or broken wires, burnt pins, busted connectors, and replace or repair anything that looks bad. Remember that some of the signals are daisy-chained from display to display, and a bad connection at one display can effect the others "downstream". Here's the things to check:

• If none of the displays work, check solenoid driver board connector J3. There's a high voltage regulator used by the displays on this board, and J3 is where it gets the 230 VDC from the power supply (pins 6 & 3) and feeds the regulated 190 VDC to the displays (pin 8). It is a good idea to dial the 190 volts down to 170 volts by adjusting the solenoid driver board's high voltage circuit pot, to lengthen the life of the score displays.
• Check the score displays themselves for cracked solder joints on the .156" molex connector headers. This is very common and can cause all kind of funky and intermittent bad scoring info.
• On the score display boards check:
  • TP1 = 5 volts DC
  • TP2 = 170 to 190 volts DC.
  • TP3 = ground.
• Check MPU board connector J1 pins 1 to 28 since this is where all the signals originate. Using a DMM set to continuity, check that each MPU connector buzzes to the player 1 display connector. A broken MPU J1 pin (common since these are small .100" molex pins) can cause a problem like say all displays showing "8" instead of "0". Below is a table of continuity.

All Score Displays are Blank.
Assuming the MPU board is fully booting (7 LED flashes), the likely cause of this is no high voltage. Remember high voltage can hurt if you touch it. Here's a check list of things to verify:

• Check rectifier board fuse F2 (3/4 Amp slow blow).
• Check rectifier board TP2 test point for 230 volts DC.
• If fuse F2 is blown and there's no voltage at TP2, often one of the four 1N4004 diodes that rectify the 230 volts is shorted. These can be easily tested in-circuit with a DMM's diode function.
• If the above are OK, check the high voltage fuse on the solenoid driver board (3/16 amp). If this fuse is blown then there's probably a short somewhere on one of your displays or the wiring harness. The high voltage regulator fuse is a bit smaller (8AG) than a standard fuse, and a hard one to find (see the above high voltage section on how to replace this fuse with a normal sized 3/16 amp fuse).
• Check TP2 on the solenoid driver board. It should read between 170 and 190 VDC. If it does, then the regulator is working. (If the display glasses are not new, use the small trimmer pot on near the fuse and adjust the high voltage to be 170 VDC. Older displays work just fine on 170 VDC and the score display glass will last longer.)
• Check is to be sure that the 170 volts is making it to the displays. Check the solenoid driver board connector J3 pin 8 for 170 VDC. Check for a good connector, and check the solder joint on the header pin.
• Check all score displays' conector J1 pin 1 for 170 volts DC and J1 pin 20 for +5 volts DC. Check these connectors and the header pins too. There are also three TP test points on each score display: TP3=ground, TP1=+5 volts, TP2=170-190 volts.
• If there is 170 VDC at all of the displays and they are still blank, then there is probably a problem on the MPU or the MPU connectors. For example this could be a "stuck-on" blanking signal.
• Disconnect the Sound board! Yes this a strange, but just do it before doing anything else. Remove all connectors from the sound board. I had a Cheap Squeek sound board once with a problem, and it would not allow any of the score displays to work. It also caused wacky MPU controlled lighting problems.
• All the score display glasses are outgassed and won't work (even if everything else is Ok). Though this is very unlikely, it can happen (especially if the game was bought from someone that swapped in all bad displays, and sold the game "as-is not working"). It happens more often than you may think.

One or Two Score Displays Blank.
If just one or two of the score displays are not working at all, this could be an "out gassed" display glass. Before replacing the score glass, swap the non-working display into another player position that is currently working. If it doesn't work there either, there is a good chance a new score glass will be needed. There is a chance all the display digit transistors (2N4401) are bad, but that is less likely.

The score display transistors can all be tested with the power off using a DMM and the diode function. This should show .5 to .7 volts with the diode function.

If the test points are showing +5 (TP1) and 170-190 volts (TP2) on the score displays and all the transistors test as good, it is a safe assumption that the score glass is bad. The easiest way to test this is to take a new score glass, and just touch the new display glass' wire leads to the display board in quesiton while the game is on (make sure to properly orient the new display just like the old display). It is important that the new score display's leads be bent straight when doing this, and it takes a careful hand to line it up. But it is not hard to do. If the new glass lights up (even partially), the old display glass is bad. This is an easy test and requires no desoldering and soldering.

All Displays Missing the Same Digit or same Digit Locked On.
Say all the displays have the same digit out, or say all the displays show an "8" instead of "0". The problem is most likely caused by a bad MPU board connector J1 (.100" Molex connector) or a bad terminal pin inside the MPU J1 plastic connector housing. See the table above for the pins used on MPU connector J1. The MPU daisy-chains the digit enable signal to all the displays, and if the same digit is out on all, then it's usually the digit enable signal is not present. Check the connectors on the MPU and all of the displays. If everything is Ok, then one-by-one disconnect each score display connector and see if the problem changes. If it does, then there's a short or a bad connector problem on the score display board just removed. Be sure to turn the power off between connnector dis-connects, and wait 10 seconds for the high voltage filter capacitor to discharge before removing any connectors. Be sure to especially check MPU Connector J1 pins 1 to 6, as that's where the digit enable signals come from, and they are daisy chained to each display.

If the problem still exists, check the MPU board resistors that attach to connector J1 pins 2 to 20. A resistor can go open or have a bad solder joint. If the problem still exists, try swapping chips U10 and U11 and see if the problem changes (U11 is the chip in question though). Also the U11 chip socket could be a problem, check that for corrosion or damage. If the MPU board is rigged up to a power supply on the bench with all 7 flashes, a logic probe can be used on the J1 connector pin in question (see the schematics to figure out exactly which pin). With the board in attract mode, the pin in question should change states. If not trace it all the way back to the U11 chip. Lastly check/replace the MPU board U20 chip.

Diagram of Bally displays and connectors. Picture by Steve K.

Missing a Digit (or Two).
This is very common where a score display is missing a digit or two. It is usually a bad (open) 100k ohm resistor R1, R3, R5, R7, R9, R11 (and R56 on 7-digit displays) on the score display board. Use a DMM set to ohms and test all the 100k ohm resistors. If one is found bad (using your DMM set to ohms), replace them all! They are all 100k ohm resistors. Replace with 1/2 watt versions, instead of the factory installed 1/4 watt size. When installing the new resistor, make sure to leave them slightly above the circuit board for better air flow. Also clean off all the black soot that is attracted to the displays with a 2 inch paint brush.

If the 100k ohm resistors are Ok, next test the level shifter transistors Q1 to Q6 (MPS-A42; and Q20 for millions on 7-digit displays) which lead to the digit driver transistors Q7 to Q12 (2N5401; and Q21 for millions on 7-digit displays). For example, Q6 & Q12 control the 100,000 digits, and Q1 & Q7 control the ones digits. Just to be sure, also check the connector on the display board J1 pin 4 through J1 pin 9. These are the connectors that supply the digit-enable signals from the MPU. The problem may be that the display driver is not getting the enable signal from the MPU. The chart below shows the digit that's blank and its associated two transistors for that digit.

• Game off, test all the 100k ohm resistors. Replace any out of spec or that are open.
• Game on, enter the display test (press the red test button twice).
• Connect a test lead to the score display TP3 (ground).
• Ground the collector of the level shiter transistors Q1 to Q6 (for example, on Q6 this is the junction of Q6 and R11). Remember Q1 controls the ones, Q2 controls the tens, and so on up to Q6 (and Q21 for millions on 7-digit displays). If the digit does not light, replace the appropriate digit driver 2N5401 transistor Q7 to Q12 (Q7 controls the ones, Q8 controls the tens, and so on up to Q12 and Q20 on 7-digit displays).
• If the digit does light, replace the appropriate level shifter MPS-A42 transistor Q1 to Q6 (and Q21 on 7-digit displays).

Digit	100k Resistor	Level Shifter Transistor MPS-A42	Digit Driver Transistor 2N5401
1's	R1	Q1	Q7
10's	R3	Q2	Q8
100's	R5	Q3	Q9
1000's	R7	Q4	Q10
10,000's	R9	Q5	Q11
100,000's	R11	Q6	Q12
1,000,000's*	R56*	Q20*	Q21*

* 7 digit display boards only.



Segment Always OFF.
Segments are the parts each digit that make up the digit. If one segment is not working, some (or most) digits will not display correctly. Here's a test to figure out what is wrong:

• Game on, enter the display test (press the red test button twice).
• Connect a logic probe of the Base of the segment driver transistor Q13 to Q19 (see the table below to determine exactly which segment driver transistor).
• The logic probe should "dance" as the segment is turned on and off in the display test. If the logic probe does "dance", replace the MPS-A42 segment driver transistor.
• If the logic does not dance, cut the Base leg of the MPS-A42 transistor. Use the logic probe and now test the portion of the cut leg coming out of the circuit board (which is now disconnected from the transistor).
• If the logic probe "dances", replace the MPS-A42 transistor.
• If the logic probe does not "dance", replace the U1 decoder chip 4543.

Segment Always ON.
Here's a test to figure out what is wrong:

• Game on, enter the display test (press the red test button twice).
• Connect a logic probe of the Base of the segment driver transistor Q13 to Q19 (see the table below to determine exactly which segment driver transistor).
• The logic probe should "dance" as the segment is turned on and off in the display test. If the logic probe does "dance", replace the MPS-A42 segment driver transistor.
• If the logic probe does not "dance", use a DMM and check the voltage at the Base of the segment driver transistor. It should be 5 volts DC.
• If the Base is greater than 5 volts (like say 110 volts), replace the transistor and the U1 decoder chip (4543).
• If the Base is 5 volts DC, replace the U1 decoder chip (4543).

Segment Segment Driver Transistor MPS-A42 a Q13 b Q14 c Q15 d Q16 e Q17 f Q18 g Q19

Picture by Steve.

Decoder Chip: Some Displays Show Strange Scores.
One Display Missing the Same Segments on all Digits.
This assumes that the MPU connector J1 pins 2 to 20 are Ok and you have check the continuity from J1 pins 2 to 20 to the player one display. And that all displays have .156" molex header pins checked for cracked solder joints.

Each display has it's own decoder chip on the display board, which takes four inputs from the MPU and outputs seven signals to light the seven segments of a digit. If having problems with one or more segments on all the digits, then it's either the score display board segment driver transistor or the decoder chip itself.

The first course of action is to test/replace the segment driver transistor (MPS-A42) on the score display board. If the problem is still there, then suspect a bad MC14543LE decoder chip on one of the score display boards.

Note that just one bad decoder chip on one of the score display boards can cause ALL the score displays to display inproper information! For example, I recently had a bad player-four decoder chip which caused all the displays to show a default high score of 91111 instead of 80000 when the game was first powered on. Also in the game's display test only odd numbers would be shown on the displays (111111, 333333, etc.) If this is the case, disconnect all display but one and see if normal scores return. Then connect each display one by one (power off) until the offending display is found.

Usually either the transistor or decoder chip will fix most strange score problem (unless it is a MPU board problem). Remember when replacing the chip to use a socket. To figure out which driver transistor is for which segment, refer to the chart above. Just find the segment that's having problem, then look up it's corresponding driver transistor.

Digit Always On or Brighter than Other Digits.
Note these steps involve jumping 190 volts DC. If this is done incorrectly, DAMAGE TO YOUR BOARDS WILL OCCUR. So be careful! If done wrong, it can blow the MPU's 6800 and 6821 PIA chips, and the sound board's 6808 and PIA and EPROM chips (depending on the exact sound board).

• Carefully connect a jumper lead from TP2 (190 volts DC) to the BASE of the digit driver transistor Q7 to Q12 (depending on which digit is missing, see chart below).
• If the digit turns off, replace the appropriate Q1 to Q6 level driver transistor (see chart below).
• If the digit does not turn off, replace the appropriate Q7 to Q12 digit driver transistor (see chart below).

Digit	Level Shifter Transistor MPS-A42	Digit Driver Transistor 2N5401
1's	Q1	Q7
10's	Q2	Q8
100's	Q3	Q9
1000's	Q4	Q10
10,000's	Q5	Q11
100,000's	Q6	Q12
1,000,000's*	Q20*	Q21*

* 7 digit display boards only.



Dim Score Displays.
Sometimes the score displays can seem dim. This is often caused by resistors on the score display circuit board. There are many resistors on this board. Test them with your DMM to make sure they are have correct values. Also make sure the header pins and connector pins on the score display board are clean and bright. If they are not, this adds resistance which dims the displays. To correct this, the pins must be replaced. Also sometimes the header pins can have cracked solder joints on the score display circuit board too. Resolder these to correct this problem.

Start a Game and Player One Score Display Shows Credits and Ball in Play.
Problem: first turn the game on, all score displays are fine and working correctly. Start a game and the player one score display shows the credits and ball in play information, just like the normal credit/ball in play score display. Essentially now there are two score displays showing the number of credits and the ball in play. All other score displays are fine. When playing the game everything on the game is working fine, but that player one score display still shows the credit/ball info.

Answer: Shorting any of the switch strobes (columns) to ground will put the credit match info in the wrong score display. Look for a switch that is somehow touching grounded metal. This could happen when connecting the backbox and pinching a wire between the backbox and the lower cabinet for example. Or maybe the ball trough switch is touching the trough metal.

Wacky Score Display Information - Bad 5101 RAM.
I was working on a Silverball Mania where the player one score display was just strange. In attract mode the player one display would give weird scores like "_66_00". If I put the game in display test, only the thousands, ten thousands and hundred thousands digits of the player one display would work. In lamp test "00" flashed on player one as the lamps lit (normally in lamp test the score displays should be blank). The game would not play either, with scoring being very strange at best and some of the coils not working.

I installed a new Alltek replacement MPU board and the game worked perfectly. So it was not the score displays or anything but the MPU board. On a guess I replaced the U8 5101 RAM chip and the problem was fixed! The interesting thing too was the game's power-on self test did not find the 5101 RAM as bad. Turns out a bad 5101 RAM or one that is too slow will not keep up with the CPU's demand for accessing the data. This can cause wacky score display issues. Here's a list of 5101 RAM speeds:

• PCD5101P : 150nS (Philips)
• 5101-1 : 450nS
• 5101L-1 : 450nS
• 5101-2 : 450nS
• 5101L-2 : 450nS
• 5101 : 650nS
• 5101L : 650nS
• 5101-3 : 650nS
• 5101L-3 : 650nS
• 5101-8 : 800nS

The Phillips brand of 5101 is the best, as it's the fastest and will work in all Bally/Stern MPU board applications. Sterns Mpu 200 need 450ns or faster RAM, Bally MPU -35/-17 need 650ns or faster. So you can see that some 5101 RAM just won't work properely in Bally or Stern MPU boards.

Displays Show 2s or 7s in all Digits of all Displays.
Try swapping U10 and U11 PIAs, as there may be a bad U10 port A output. Also MPU board U20 could have failed too. Also a faulty 5101 RAM on the CPU board can cause wacky score display behavior (even though the game passes all boot-up RAM checks). Last look for fried resistors in the top right hand corner of the CPU board.

7 Digit versus 6 Digit Score Displays.
Seven digit score display boards and display glasses started with Skateboard in the fall of 1980. The 7 digit numeric displays were the same physical footprint as 6 digit displays, just the 7 digits are squeezed closer together. In fact 7 digit Bally display boards have the same .156" molex connector, and can be plugged into 6 digit games with no ill effect - only the last digit (1,000,000) will not work.

The 7 digit display circuit board is only slightly different than the 6 digit board. A few added transistors and resistors and a 7 digit numeric display glass being the only difference. The same transistor numbering system is used. For example the level shift transistors are still Q1 to Q6, with the addition of Q20 for the millions digit. Likewise the digit driver transistors are still Q7 to Q11 with the addition of Q21 again for the millions digit.

Bally and Stern Display Units - Interchangable?
Six digit Bally and Stern display units are indeed interchangable. But the seven digit displays are not! But you can modify a Bally or Stern 7 digit display to work in either game.

To do this, jump together pins 11 and 12 of the Bally or Stern 7 digit display with a short piece of wire. This will allow either 7 digit display to work in either game. There is one problem with this, because the Stern display circuit board is deeper than a Bally board. To make this work, you must remount the tray-like mounting bracket on the other side of the backbox door. If you are using a Bally 7 digit display in a Stern game, be careful of shorting the circuit board to the metal bracket since the board will not reach the rubber support buttons at the rear of the bracket.

Finally, the Stern 7 digit display units do not have the circuitry to drive the commas in the display tube. So if installed in a Bally game, the Stern 7 digit display will never show any commas. If a Bally 7 digit display board is installed in a Stern game, the commas will show up! This happens because the commas are generated on the display board and is not part of the data sent by the MPU.

Flickering Displays.
Flickering displays can often be attributed to cracked solder joints on the .156" molex connector J1 on the score displays. While the game is on, put slight pressure on the plastic connector housing attached to the score display. If the flickering goes away momentarily, there are definately cracked solder joints on the score display's .156" header pins. Usually the end pins (pin 1-6 and 14-20) are the ones that crack. It is common for all five displays to have cracked solder joints on the J1 score display header pins! These crack easily from plugging and unplugging the cable on the score display and from vibration. Reflowing the solder and adding a bit of new solder will fix this problem.

Other causes of flickering displays can include the following (but these are a lot less common). Or a broken lead on a component, assuming it is just happening to one or two displays. Here are some other things to check. Often reflowing the solder joints on these components will fix the flickering problem.

• Pin 36 on the display glass.
• Display board R21, R22, R29.
• Display board Q17
• Display board C2 (it can go open).
• Display board chip U1 pin 13 (replace if in doubt).

If all score displays are flickering, there may be a MPU problem (usually the PIA at U10). Also the U10 chip is near the MPU's battery, so battery corrosion can be an issue. I recently repaired flickering displays because a trace going to U10 around the battery area was intermittent due to unfixed battery corrosion (the MPU would sometimes boot fine, but would also only show three LED flashes ocassionally).

The J1 header pins on a score display circuit board. These often need resoldered to fix flickering displays. If you look carefully at this picture, you can barely see the solder joints for pins 20,19 and 18 have cracked (circles around the base of the solder).

MPU board and Sound board problems causing Score Display Problems.
If data from the MPU board is not properly sent to the score displays, this can cause a display not to work. Often the culprit is chip U20 on the MPU board. Move the score displays to a different player to isolate and make sure the problem does not lie within the score glass itself.

Also remove all connectors from the sound board. I have seen sound board problems cause all the score displays to be blank.

Test Counting Sequence Out of Numeric Order.
If during the score display internal test, the counting sequence is strange (i.e. 0,1,0,1,4,5,4,5,8,9 instead of 0,1,2,3,4,5,6 etc), there could be a problem with one of the address lines not being read correctly. If the problem moves when you swap displays, this means the MPU board is not the problem. Often the resistors R49 to R53 (20k) can be the problem (should be within 5% of 20k ohms), but also check for cracked solder joints at the header pins (a VERY common problem). Also a bad decorder chip MC14543LE on one of the score displays can cause this too. The interesting thing is one bad display can affect the other displays. You may have to issolate the displays to find the bad one by plugging them into the game (with the game off) one at a time, and then running diagnostics.

Displays Strobing.
Perhaps all the score displays work but they "strobe" (flash quickly). This can be caused by a number of problems, the most common being cracked header pins on the score display board(s) themselves. But if all the displays strobe, cracked score display board header pins are probably not the problem. Other things to check are the U9, U10, U11 and U20 chips on the MPU board (and the traces that go to these chips, especially if battery corrosion was an issue). The MPU board's 555 timer chip could also be a problem. Or even the MPU board resistors R26 and R27.

Replacing a Score Diplay Tube.
Score display tubes don't last forever. With time and usage, the displays often "outgas". There is no way to fix this short of replacing the glass display tube. Here's how to do it:

• Cut the pins on the old score display tube. Do NOT try and save the old glass! The circuit board is what you are trying to save, not the (bad) glass.
• Remove the old tube's pins from the circuit board. Do this by heating each cut pin, and pulling it out with needle nose pliers. Then use a solder sucker to remove the old solder from the circuit board holes.
• Install the bottom mounting bracket if it was removed for cleaning!
• Feed the new display tube pin into the circuit board starting at one end of the board. This is easiest if all pins are straight first.
• When all the pins have been inserted, carefully bend the tube pins to produce a strain relief bend in all of the pins. This takes a bit of practice so that all the pins don't slip out of the holes in the circuit board! This is very important because as the display tube heats up, the tube can fail where the pins enter the glass if there isn't good strain relief.
• Check display tube alignment before soldering. The bottom plastic display bracket must be installed to check this. If you screw this up, the score digits may not be fully visible through the backglass score windows! Compare your new display to an existing display to make sure things are correct.
• After aligned, gently clamp the tube in place with 2 wooden clothes pins.
• Tack solder the display pins at each end of the circuit board. Re-check alignment.
• Solder all the display tube pins to the circuit board. Make sure the pads are soldered good on both sides of the circuit board. Inspect for any shorts between two pins.
• Clip off the excess display pin leads.
• Inspect again for shorts.
• Put two marble sized dabs of silicon adhesive on the circuit board on both ends of the display glass. Do NOT fully glue the glass to the circuit board!
• Put a big dab of silicon on the display tube's "nipple". This will protect it from breaking and ruining the glass.
• Let the silcon dry overnight and re-install the display.

Often a game will "reset" during play when the flippers are used. The flippers are powerful coils and consume a fair amount of power. On a game with problems, this can causes the 5 volt power supply to drop below 5 volts, causing the MPU board to reset or lock up. Or maybe the game just locks up or resets when it's just sitting there being unused.

First check the flipper coils. There should be two 1N4004 "snubbing" diodes mounted on each flipper coil. When the flipper coil deactivates, it dumps power backwards to the solenoid driver board. The diodes prevent this "backwash". If they are not there, the back flow of power can cause all sorts of strange game behavior (including resetting the game.) Gently tug on these flipper coil diodes, if they are lose or broken the tug should identify that. In my experience these diodes are pretty robust and generally don't break or fail. But if in doubt, replace them. (Note the band on the diodes must be oriented correctly, towards the 43 volt power feed.)

Next check is the C23 filter 12 volt capacitor on the Solenoid driver board (as discussed earlier in this document). If it is not doing its job (old & worn out), the 5 volt power supply will not be as smooth as the game may require. This can cause the MPU board to lock-up and freeze during game play. Use a DMM set to AC volts, and put the DMM leads on both lugs of C23. If more than .250 volts AC is seen, this capacitor definately has failed (actually if more than .150 volts AC is seen, the cap is on it's way out). Replace C23 with a new 10,000 to 15,000 mfd 20 volt capacitor. If your game has more than two flippers, use a 15,000 mfd cap. If the game only has two flippers, 10,000 or 12,000 mfd will work fine.

Next thing to check is the "turn around" power connector on the solenoid driver board (SDB). This is the J3 connector at the upper right side of the SDB. The 12 volts coming in from the transformer rectifier board, which ultimately gets processed to +5 volts, is on this connector. This single ORANGE wire at J3 pin 12 brings all the power from the transformer's rectifier board which becomes +5 volts for the whole game! Because it is just one wire through one .100" connector pin, often this SDB J3 pin 12 becomes intermittent or completely fails. The terminal pins themselves will need to be replaced, along with the board's .100" header pins. Sometimes J3 is an IDC (insulation displacement) connector. In this case the whole 25 pin connector housing should be replaced with a crimp-on style housing (a band-aid fix would be to connect the orange J3 pin 12 orange wire directly to SDB TP5, but I do not recommend that).

Another easy test of the SDB J3 connector are the three test points right next to the J3 connector. The top-most TP5 should be the input 12 volts (12 to 14 volts DC unregulated). Next below this are the two 5 volt test points TP1 and TP3 (should show 4.95 to 5.25 volts DC). If there is any difference in voltages between these two test points, there are connector problems on the SDB J3 connector. To help take some pressure off the J3 connector, I solder a wire between TP1 and TP3. But PLEASE verify these are the 5 volt test points on your SDB. Because if you connect the 12 volt TP5 to TP1/TP3, you will blow up all sorts of chips on the MPU board!

Solenoid Driver board J3 connector:

• *SDB A3J3 pin 12 = 12 volts DC from transformer's rectifier board A2J3 pin 8.
• *SDB A3J3 pin 10 = Ground from transformer's rectifier board A2J3 pin 17.
• SDB A3J3 pin 11 = 12 volts to MPU board.
• SDB A3J3 pin 6 = 230 volts DC from transformer's rectifier board.
• SDB A3J3 pin 8 = 190 volts DC to score displays.
• SDB A3J3 pin 3 = Ground for score display voltage.
• SDB A3J3 pin 5 = 43 volts for flipper enable relay.
• SDB A3J3 pin 23,24 = ground for 43 volts solenoid power.
• SDB A3J3 pins 13-17,25 = +5 volts DC power for entire game.
• SDB A3J3 pins 18-22 = logic ground for entire game.
* Most important connector pins to check/replace in regards to resets.

Now re-test the solenoid driver board TP1 and TP3 voltages - they should both be the same at about 4.95 to 5.25 volts DC. If they are not, again the SDB J3 connector is at fault. Again the best way to fix this is to run a jumper wire between TP1 and TP3 right on the solenoid driver board (be careful not to jump to TP5 which is right above TP1/TP3).

Now check the transformer rectifier board's J3 connector for burnt .156" molex pins. This information mostly applies to pre-Xenon games with the transformer in the backbox (the newer Xenon and later rectifier boards have far less problems). Replace the mail header pins and use new Trifurcon terminal pins in the plastic wire housing. Transformer's rectifier board J3 connector:

• *Transformer A2J3 pin 8 = 12 volts DC to SDB A3J3 pin 12 (which becomes +5 volts).
• *Transformer A2J3 pin 17 = ground to SDB A3J3 pin 10 for 5 volts DC.
• *Transformer A2J3 pin 15 = ground to SDB A3J3 pin 21 for 12 volts DC.
• Transformer A2J3 pin 5 = 230 volts DC to SDB.
• Transformer A2J3 pin 16 = ground to SDB for 230 volts DC.
• Transformer A2J3 pin 2 = 7.3 volts AC GI return.
• Transformer A2J3 pin 10,11 = 7.3 volts AC GI.
• Transformer A2J3 pin 3,4,14 = ground for 5.4 volts DC switched lamp power.
• Transformer A2J3 pin 6 = 5.4 volts DC switched lamp power.
* Most important connector pins to check/replace in regards to resets.

Check the J4 power connector on the MPU board. This is in the battery corrosion zone, so often this connector can develop resistance and fail. The easy way to test this is to check TP1 (5 volts). Again it should be 4.9 to 5.25 volts DC. If it is less than TP1/TP3 at the solenoid driver board, there is a connector problem! Also check MPU board TP2 for 12 to 16 volts DC. Again this should be the same voltage as the solenoid driver board's TP5. Also check the MPU board's TP3 for 43 volts DC. Note that MPU J4 connector pins for 43 and 5 volts are right next to each other! If battery corrosion crosses these two pins, 43 volts will go down the 5 volt buss, and smoke the entire MPU board! Often battery corrosion will get *under* the plastic housing that holds the J4 male header pins on the MPU board, shorting 43 volts to 5 volts. Please keep that in mind. MPU J4 pins connectors:

• *MPU A4J4 pins 16,17 = +5 volts
• MPU A4J4 pin 12 = 12 volts DC
• MPU A4J4 pin 15 = 43 volts DC
• MPU A4J4 pins 18,19 = ground
* Most important connector pins to check/replace in regards to resets.

Another problem is the slam switch, mounted on the coin door. If this switch closes during game play or attract mode, the MPU board will reboot and start its LED flash sequence. The slam switch wires can act like an antenna, picking up noise from the flipper coils during game play. The easy fix to this is cut the wire attaching the slam switch to the MPU board at MPU connector J3 pin 16 (bottom right, bottom most pin). It is the slam switch input to the MPU board. This will often fix a flipper game play reset problem.

The sound board discussed mainly here is the Bally "Squawk and Talk" (S&T;) board (as used in most of their talking games) and the -51/-56 sound boards.

Can You Hear the Amplifier?
All sounds boards have an amplifier. If the volume pot is adjust, the "white noise" the amplifier produces should increase in volume. So before starting any repair, determine if the amplifier is working (can some "white noise" be heard?) If the sound board amplifier is making no noise, usually there is a bad pot, a bad speaker, a removed sound board connector, or a blown fuse.

Bad Speaker?
This happens more often than you think. A bad speaker, or a speaker with one broken lead can drive you crazy thinking the sound problem is a lot more complicated. Check the speaker first before proceeding.

Bad Pot?
There are two potentiometers on the S&Q; board, one for the sound and one for the speech. Often these are mis-adjusted, or have developed a "dead spot". With the game off, turn the pots a few times. Then power the game on and adjust them as needed (usually both area nearly turned up all the way).

Sound Board Chip 86L93
This can be replaced with a 74LS93.

Cheap Squeak Causes No Score Display and bad CPU controlled Lighting.
I recently worked on Bally game with a bad Cheap Squeak. With the sound board plugged in, none of the game's score displays would work. Also the CPU controlled lighting was confused. If the sound board was disconnected, the score displays worked fine as did the CPU controlled lights.

The 2518-51 sound board which, while better than the earlier sound boards which only made mono-tones, is a far cry from the Squalk & Talk board. The difference between the -51 and the -56 board is the -56 model can be connected a vocaliser board (where the -51 can not). Also the -56 boards has some jumpers to convert it into a -51 model. 2518-51 schematics, 2518-56 schematics.

The 2518-51 and -56 boards can be identified by a 6808 (or 6802) CPU 40 pin processor, a 6820 or 6821 40 pin PIA chip, and a single sound ROM at U4. The sound ROM is usually a black ROM, but is equivalent to a 2716 or 2532 type EPROM. In addition the -56 board has a ribbon cable connector so it may be attached to a vocalizer board.

The -51 and -56 boards have a push button which causes a self-test: it plays either a tone or a series of sounds, or possibly does nothing. Pushing the test button with some sound ROMs will make the board beep once. Other ROMs will repeat the test sound. Still others does not respond to the test button (this does not mean the board won't work when the MPU sends signals to the the sound board during normal game play). Here are the games that used the -51/-56 sound board and what happens if the self-test button is pressed:

Game	ROM Size	ROM Number	Checksum	Self Test?	Install Jumper
Nitro Groundshaker	2716	E776-14 or -15		None	Jumper C
Future Spa	2716	E781-02 or -05	7D01	None	Jumper C
Silverball Mania	2716	E786-08 or -11	B62B	Yes	Jumper C
Space Invaders	2716	E792-02 or -07	9F03	None	Jumper C
Rolling Stones	2716	E796-11 or -19	8F9B	Yes	Jumper C
Mystic	2716	E798-05		Yes	Jumper C
Viking	2716	E802-07	A319	Yes	Jumper C
Hog Doggin	2716	E809-07		Yes	Jumper C
Xenon (-56)	2532	E811-36	13B3	None	2532: Jumper D,E 2732: Jumper C,F
Frontier	2716	E819-09		Yes	Jumper C
Skateball	2716	823-02		Yes	Jumper C
Speakeasy	2716	E877-01	40E5	None	Jumper C
BMX	2532	888-02		?	Jumper D
Grand Slam	2532		F338	Yes
Goldball	2716		F273	?

AS-2518-51 and -56 Sound Board Jumpers.
The -51 and -56 sound boards can use either a 6808 and a 6810 RAM, or a 6802 and no 6810 (because the 6802 has internal RAM).

-51 and -56 board U3 CPU jumpers:

• 6802 CPU - install jumper A, remove jumper B, remove U10 RAM 6810
• 6808 CPU- install jumper B, remove jumper A, install U10 RAM 6810

-51 board U4 ROM jumpers:

• 9316 ROM or 2716 EPROM - install jumper C, remove jumper D
• 9332 ROM or 2532 EPROM - install jumper D, remove jumper C

-56 board U4 ROM jumpers:

• 2732 EPROM: install jumper C & E, remove jumpers D & F
• 2532 EPROM: install jumper D & F, remove jumpers E & C.
• No attached volcalizer board: install jumper G and remove cap C27 and remove jumper J.
• Attached volcalizer board: install jumper J and install cap C27.
• Emulate -51 board: install jumper H.
• Most -56 boards will be jumpered with B,C,E,G,H for a 6808 with a 2732 EPROM and no volcalizer board.

Test EPROM for the -51 and -56 Sound Boards.
Leon has made a test EPROM that will bench test these sound boards. This is very handy as the -51 and -56 do not have a built-in power-on LED flash sequence to show problem. Without this flash LED, diagnosing -51 and -56 sound board problems is difficult.

To run Leon's test LED, all that is needed is the test EPROM, a +5 volt DC power supply, and a speaker. The game is not needed, and the sound board can be test "on the bench". The test EPROM will verify the U10 6810 memory chip (or internal 6802 RAM) and the U2 6821 PIA chip, and of course the U3 6808/6802 CPU chip. Note if your sound board uses a 6808 CPU, an external 6810 is required at U10. If you have a 6802 as CPU, the 6810 is not needed, as the 6802 has internal RAM which take cares of the needed memory. Jumpers for either CPU chip are shown above.

The sound board Test EPROM can be downloaded in either 2716 format or in 2532/2732 format. Burn this to the appropriate sized EPROM.

Attach +5 volts at sound board connector J1 pin 5, and the ground at J1 pin 6. The speaker is plugged in at J2. Make a jumper between TP1 and TP2, which jumpers 5 volts to the 12 volt amplifier power. Note we do not need 12 volts for the amplifier, as the amp will work loud enough with 5 volts.



Then take a LED and a 150 ohm resistor connected to the LED leg of the flat side of the LED:

• The resistor side of the LED is connected to ground.
• The non-resistor side of the LED is connected to U3 pin 15.
For a -56 board using a 2732 EPROM and a 6808 CPU chip, apply jumpers B,C,E,G,H.

Place the test EPROM into the socket U4. As 5 volts is applied to the sound board, the test will start immediatly. A control LED connected at pin 15 of the U3 CPU chip and ground will blink if the test EPROM is running properly, and the speaker will make a faint and fast heart-beat sound if the sound board U2 PIA is installed (if you don't have an LED connected, often just the faint and fast heart-beat sound will be enough to tell you the test EPROM is running).

If the LED is not blinking, look at some basic signals at the U3 CPU to find out why:

• U3 pins 2,3,4,40 should have +5 volts
• U3 pins 37 should be 2.5 volts (clock signal)
• U3 pin 39 should be 1.1 to 1.5 volts (clock signal).
• When all the above is seen and the clock is present, the CPU should start. If not continue...
• Remove the U2 PIA and the 6810 memory at U10. Re-boot and look again if the cpu is running (using the test LED on U3 pin 15).
• Look for signals on the address bus U3 pins 9 to 25 (except pin 21 is ground). These address signals should be between 0.5 and 2 volts.
• Look for signals on the data bus U3 pins 26 to 33.

If address/data signals are missing yet the clock and 5 volts are at the U3 CPU chip, either there is a broken trace of the CPU chip is bad.

If we have a case of a CPU that runs fine in stand-alone and with the test EPROM in place, but the test won't work because the control LED does not blink. Remove the test EPROM again and look at the signals arriving on the U3 socket. Are the data and address signals present? If not perhaps the socket is bad. We need a 0.5 volt signal at U3 pin 19 and a 4.5 volt signal at U3 pin 20. If there a missing signal look at the schematic to see wher it comes from, and always suspect sockets if a address or databus signal is missing. Bend up the pin of the CPU chip where the signal comes from and retest. Bend up pin which delivers the signal - that means you have a short on that address or data line.



If all is good now, you can start checking the U2 PIA outputs. Another test LED should be made, as shown in the picture above. This LED is connected between 5 volts and selected pins on the U2 PIA chip (the resistor side of the LED is used to probe the PIA pins, and the non-resistor side of the LED is connected to +5 volts). This will cause the LED to "dance" up and down from 0 to 5 volts. If any of the output PIA pins don't "dance", the PIA is bad. If all are missing, check the socket connections and all the signals to the PIA including the selection signals and data/address bus signals.

• U2 PIA Pins 2-17 high then low (tester LED on and off), alternating every second.
• U2 PIA Pins 19,39 high then low (tester LED on and off), alternating every second.
• U2 PIA Pins 26 to 33 are the data lines, and should be pulsing (use a logic probe for these pins).
• U2 PIA Pin 34 (reset) should be high (tester LED on).

If everything works good so far, push the test button on the sound board, and this lauches the memory test. The control CPU LED should stop blinking for a moment - If the U10 (or 6802) RAM is OK the LED will start blinking again after about two seconds. If the RAM is not good, the LED will stay either off or on and not restart blinking (the test is actually continually testing the memory, and never exits the test, and leaves the LED in its current state at the beginning of the test). The test continuously checks the memory and allow us to measure the arriving signals at the U10 memory chip. At U10 pins 1,14,15 we find ground, and at U10 pin 24 there should be +5 volts. All other pins should show signals between 1.5 and 3 volts. If that measures OK yet the test indicates a bad U10 memory, it should be assumed the U10 chip is bad. Otherwise again suspect the U10 chip socket.

To test the -51 or -56 amplifier, we just put a finger on capacitors C5 and C2. This should result in a good hum in the loudspeaker if the amp is OK. If touching C5 gives hum, the TDA2002 chip U9 is good. If we hear the hum while touching C2 then the pre-amp U8 LM3900 is good. Note the hum is louder at C2, and about half as loud at C5.

When the board still does not work in the pinball the only thing left is the sound generator U1. To test U1 we will need a game sound ROM that supports the self test (see the table above). Place a sound ROM at U4 that supports the self test. Boot the sound board then press the test button. If no sound is hear, replace the AY-3-8910 sound chip.

Strange Note: On the schematic of the -51 board there is a pulse generator formed around chip U7. At the bottom of the schematics the "notes" say that this IC is not always mounted on the board. In fact I have never seen a board with this U7 chip. When you have a board which does have a U7 mounted, control of this circuit is easy - pulses come out continously of U7 pin 3. If no puluses replace the NE555 - the rest of the pulse generator only consists of three resistors and two caps.

Xenon Vocalizer Board.
Xenon uses a secondary sound board called a vocalizer. In the bottom corner of the board CR1-CR3 diodes (1n4004) are used in series to lower the voltage going to the voltage regulator. Because of this these three diodes like to burn up. Jerry uses a small heat sink like used on the solenoid driver board for the HV transistors to keep these three diodes cool (see pic).

CR1-CR3 on the Xenon Vocalizer board replaced, utilizing a heat sink.

Squawk and Talk Sound Board.

The Squawk and Talk (S&T;) sound board was Bally's talking sound board during the early to mid-1980s. It also has a built-in LED which flashed at power-on, making board diagnostics much easier.

Squawk and Talk Distorted Sound and Hum.
Dried out capacitors on the S&T; board is the major cause. The biggest offenders are the final filtering capacitors at C16 and C22. These are 4.7 mfd 25 volt caps, but they should be replaced with 6.8 mfd 16 volts (or higher volts). (Note on some S&T; boards these are Tantalum capacitors and may not need to be replaced, because apparently Bally saw the problem and upgraded these caps to Tantalums to fix it.) Here's are the caps to replace. Note these are related *only* to sound problems (unless otherwise noted). The order is relative to importance:

• C15 (10 mfd 16 volt)
• C16, C22 (4.7 mfd 16 volt, but replace with 6.8 mfd or higher, 16 volt if possible)
• C36, C43 (2.2 mfd 25 volt)
• C19, C24, C25, C28, C31, C34, C42 (1 mfd 25 volt)
• C25, C34 (1 mfd 25 volt) - These may not be currently installed in the S&T; board.
• last resort: C14 (4700 mfd 25 volts). Replace if needed.
• C1 (47mfd 16 volt). RESET capacitor, not sound related.

Make sure you install all the above electrolytic caps in the right "direction". They are polarized, and there should be a "+" on the circuit board to help installation.

As a last resort, replace C14 (4700 mfd 25 volts). There generally is no need to replace the big caps at C27 (100 mfd 16 volt), C29 (470 mfd 16 volt), and C14 (4700 mfd 25 volt). Though these are tempting to replace, they are rarely the problem.

If replacing the above caps doesn't solve your sound problem, it can also be caused by a bad U7 PIA chip. This controls execution of speech via the TMS5200 speech chip. You can swap U7 and U11 (same chip) to see if this makes any difference. If it does, that means U7 is probably bad.

Also check the TMS5200 chip at U8 for oxidation on the chip legs, or a poor connection in the socket. This too can cause faulty speech. Sand off any oxidation on the chip legs and reinstall.

Another thing to try is swaping the PIA chips on the MPU board at U10 and U11. Also a bad potentiometer on the S&T; board can cause bad speech problems.

Don't forget to check for cracked solder joints on the back of connector J1 on the Squawk and Talk board. This too can cause speech problems.

S&T; Board Reset/Boot Problems.
The Squawk and Talk board is not unlike the MPU board. It too is a computer, with a reset section, and an LED which "flashes". The flashes determine if the board has a "boot" (turn on) problem. But before the board boots, it needs to "reset" (hold the CPU's /Reset line low until +5 volts can stablize, and then allow the /Reset line to go high and start the boot process).

Often this reset process never happens, and the S&T; board never boots (LED locked on). Since there is no battery on this board, battery corrosion usually isn't causing any problems (unlike the MPU board!) But the reset section can have faulty parts. One common thing that fails is capacitor C1 (47mfd 16 volt). Replacing this part, or the reset transistors, can often solve a S&T; boot/reset problem.

Checking Your Work.
There is a test button on the S&T; board. Note this button is NOT "debounced", and when pressed it makes multiple closures (the switch was not debounced as a cost savings). This can cause the CPU on the S&T; board to lock up. Sometimes the S&T; board will "stutter" and then lock up. To clear this, turn the game off, and then back on and wait for the game to re-boot. There is no way to avoid the test button bouncing other than luck. The S&T; board's self test button should produce several speech phrases, followed by a long single "sound". Then the board will re-boot itself.

S&T; Test Points.
These are handy to know when repairing a S&T; board.

• TP1 = Ground
• TP2 = +5 vdc
• TP3 = +11.5 vdc
• TP4 = -5 vdc
• TP5 = Speech volume control voltage
• TP6 = Sound volume control voltage
• TP7 = AY38912 output
• TP8 = E
• TP9 = TMS5100 output
• TP10 = VMA
• TP11 = TMS5100 clock
• TP12 = Reset

Bench Testing a Squawk & Talk.
As shown in Making a MPU Test Fixture section, a similar technique can be used on the revision A to D S&T; board as well. Using an old computer power supply, connect ground to TP1, +12 volts to TP3, and -5 volts to TP4. The +5 volts (TP2) comes from the +12 volt source, so do not supply that voltage. Add an 8 ohm speaker to J2 pins 1 and 2.

Now use TP12 jumped to ground to reset the CPU (which saves turning the power off and on all the time). Clip a jumper to TP12 and then briefly touch the other end of the wire to ground (this will reset the CPU without turning the power on and off all the time). TP12 is connected to pin 4 of U15. This works with the board in the game as well. Once the board has booted, press SW1 to make the board talk.

Some repair tips: the same bad socket and corrosion issues seen on the MPU board also apply to S&T; boards. The 6802 CPU (U1), PIAs (U7,U11) and EPROM (U2-U5) sockets go bad, and the CPU, PIA, EPROM legs tarnish and fail to conduct, etc. So look at this common and easy stuff first.

As for specific S&T; stuff, it depends on how many (if any) flashes are seen (see below). A locked on LED is just like a -35 MPU, the CPU chip is not getting or executing code. Check the VMA pin (U1 pin 5) with a logic probe and see if there is a pulse. Also check U1 pins 38 and 39 for a clock signal. Bad crystals are common too. If the crystal is replaced, change the two capacitors C3,C4 as well.

The Squawk and Talk in Technical DETAIL!
Clive Jones wrote and excellent technical document on exactly how the Squawk and Talk works. If you wish to read this document, go to squawk.htm. When fixing a S&T; board, this document should help greatly.

Normal Operation.
The S&T; accepts address signals from the MPU to select one of the sound or speech signals stored in memory. It then plays the request by controlling the sound generator chip U12, or the D/A converter U10 for sounds, or the speech generator chip U8 for speech. The S&T; is notified of a sound/speech request by an interrupt from the MPU.

Power Needs.
The S&T; requires several voltages for operation. A +12 vdc at 3 amps (unregulated) is required for the LED, VR1, and audio amplifier U18. Also +5 vdc is required for the remaining components, and comes from 6.3 vdc through voltage regulator VR2 (which sources with +12 volts). Also -5 vdc is needed for the speech generator chip, and also comes from the 6.3 vdc input voltage.

Audio Control.
The S&T; generates speech via the U8 chip. Commands and speech data are passed to this chip through U7 PIA. The speech chip uses the information it receives to control an electronic vocal tract that produces a speech signal across R14. This signal contains unwanted high frequencies that are removed in a low pass filter through U13, C19, C20, C21, R11, R15, R16, R27 and R81. The filtered speech signal is mixed with an optional off-card audio signal and is given to the speech voltage controlled amplifier (VCA). The output of the speech VCA is fed into the U18 power amplifier (8 watts).

No Sound and Loud Buzz from Speakers.
This isn't always a complicated problem. Sometimes no sound and/or loud buzzing from the speakers can be as simple as a bad ground to the sound board. Check the Molex connectors, and make sure the ground wires are making good contact to the sound board.

The S&T; has a series of LED power-on flashes much like the MPU does. Here's a description of the flashes.

First Flicker.
Fakers Guide: No flicker means a bad U5, flakey U15, leaky C1, open R1, leaky CR1, or bad U17.
Techno Guide: On power up, U1 requires +5 volts to be applied before the reset line is allowed to go high. If this condition is met, the LED does a quick "flicker". At power-on, C1 slowly charges via R1. The voltage across C1 is monitored by U15. When it reaches 1.7 volts DC, U15 take the reset line high. Diode CR1 across R1 provides a quick discharge path for C1 in the event the +5 momentarily disappears.

First Flash.
Fakers Guide: No first flash means a bad U6, U15 or U17.
Techno Guide: The U1 chip tests the U6 RAM. It attempts to write then read back all 256 patterns in each of the 128 scratch pad RAM memory locations. If U1 completes the 256 x 128 = 32,768 tests, the LED is flashed.

Second Flash.
Fakers Guide: No second flash means PIA U7 is bad.
Techno Guide: The U1 chip now tests the first PIA U7. Each of the two PIA chips U7 and U11 are interchangable. The test is the same for both. If it determines the two PIAs are good, the U1 chip performs some test. This includes testing each of the two full byte port initialization registers, testing the two full byte I/O registers, and testing the CA2 and CB2 ports. If all checks out, the LED is flashed.

Third Flash.
Fakers Guide: No third flash means PIA U11 is bad.
Techno Guide: The same test above is performed on PIA U11.

Fourth Flash.
Fakers Guide: No fourth flash means sound generator U12 is bad.
Techno Guide: The U1 chip performs a test on the sound generator chip U12. The U12 chip is controlled by PIA U11. If the sound chip passes the LED is flashed the fourth time. A bad PIA at U11 can also cause no fourth flash! Or a bad connection between the three chips U1 (microproocessor), U7 (PIA), and U12 (sound generator).

Fifth Flash*.
Fakers Guide: NOTE: Some games, namely Fathom, will not have a fifth flash! See the explaination below. No fifth flash means speech chip U8 is bad. Also no fifth flash can be as simple as no -5 volts at TP4. Note the -5 volts comes from the G.I. (General Illumination) circuit. So if the GI fuse is blown, the sound/speech won't work either (because the -5 volts will be missing on the sound board). Thanks Cliffy!
Techno Guide: The speech generator U8 chip requires an initialization sequence at power-on. Since the chip is a "slow" device, there is an acknowledgement signal from the speech chip to the U7 PIA. Every time a write to the speech chip is done, the speech chip acknowledges. The U7 PIA attempts to send 9 bytes of initialization data to the speech chip, one at a time, waiting for acknowledgement. If it is sucessful, the speech chip is considered functional and the LED is flashed the fifth time. A bad PIA at U7 can also cause no fifth flash! Or a bad connection between the three chips U1 (microproocessor), U7 (PIA), and U8 (speech chip).

*No Fifth S&T; Flash on Fathom (and some other games).
Fathom does not use the AY-8912 PSG (Programable Sound Generator) chip at U12 (and has the EE jumpers installed on the S&T; board indicating it is not used). Because of this, only *four* flashes of the S&T; LED will be seen (after the initial power-on flicker). This is normal and correct. Disassembly of the self-test ROM code for Fathom reveals there are only four calls to the LED flash routine instead of five.

Volume Lowers as the Game Warms Up.
After power-up the voice volume is initially lower (about half as loud) than the sound effects volume. After the game is on for about 15 minutes or so, then the volumes are equalized.

To fix this, try spinning the two pots on the S&T; board back and forth several times rapidly to clean up the internal wiper contacts. These can get oxidized and develop "dead" spots over time. After "cleaning", adjust the volume accordingly. If problem persists, replace the pots. If problem still persists, there may be a bad capacitor(s) on the S&T; board, as this is common too. But this usually gives more symptoms than just volume issues

Stern SB-100 Sound Board.
Problem: Stern Wild Fyre and was having trouble with the SB100 sound board. Whenever the game was turned on, would get a screeching sound from the sound board. The sound would continue until a game was started and a "game sound" was initiated. At this point, the screeching would stop and the game would function normally. However, once the ball drained, the screeching noise would return again until another game sound was initiated.

Solution: replaced the 7402 chip at U1 on the SB-100 and that fixed the problem.

When flipper don't work there are two avenues to travel: is the problem electrical or mechanical? This is usually easy to figure out. For example, say the flipper stays "up" (in the energized position) after the cabinet flipper button is released. At this point, turn the game off. If the flipper returns to its normal resting position, then it's an electrical problem. If the flipper stays up, then it's a mechanical problem.

Flipper coil on a Bally Eight Ball. Here are the important parts of the flipper assembly.

Electrical Problems.
Check the cabinet flipper switches. Sometimes the contacts for these switches can actually weld themselves together! Either replace the switch blades with new ones, or separate them and file the contact points clean with a metal file.

The End of Stroke (EOS) switches on each flipper coil are also problematic. If these are mis-adjusted or broken, the flipper will either burn up and seize, or be extremely weak.

What does the EOS switch do? Well a flipper coil is actually *two* coils in one package. A high power (low resistance) coil that initially kicks hard. And a low power (high resistance) coil that allows the player to hold the cabinet flipper button in without burning the flipper coil. The EOS switch is the determiner as to which part of the flipper coil gets power.

Note the EOS switch OPENS when a flipper coil is energized.

When the EOS switch is closed (as it should be when the flipper is not energized), the low-power part of the flipper coil is shorted out (not in the power circuit). But when the flipper coil is energized, the EOS switch opens up, putting the low-power side of the flipper coil into the power circuit (which prevents the flipper coil from burning).

So if the EOS switch is broken (missing a blade for example), the flipper will be very weak (since the low-power side of the flipper coil is always in the power circuit). If the EOS switch is mis-adjusted and never opens, then the low-power side of the flipper coil is always shorted out (not in the power circuit), and the flipper coil will eventually burn (or blow a fuse). An easy way to tell if the flipper coil has been heat stressed is to try and remove the nylon coil sleeve. If it comes out easily, the coil is probably fine. If the sleeve won't come out, it's time to replace the flipper coil.

Another common electrical problem is on the solenoid driver board (SDB) in the backbox. The top left connector J1 handles the ground returns to the flipper relay on the SDB. It is very common for the J1 .156" male connector pins to crack where they solder to the board (resolder these points). Also the traces on the back of the SDB going from J1 to the relay can actually burn off the board! And the solder points for the flipper relay can go "cold" and need to be resoldered.

If a flipper coil mechanically binds (stays in the energized position), this often happens because the end of the coil plunger "mushrooms", and rubs on the inside of the nylon coil sleeve. The plunger/link can be removed from the assembly, and the muchroom filed or ground off) To do this, the coil stop will need to be removed, the flipper coil pushed asided, and the two set screws on the pawl will need to loosened so the pawl/plunger/link can easily be removed.

Power to the flipper coils is handled by the brown power wire going to the coils. Note this wire goes to the outside flipper coil lug with the BANDED side of the diode. I have seen these brown wires break internally, causing one of the flipper to not work. Remember the under-playfield fuse is wired after the flippers (so if this fuse if blown it should not affect the flippers).

Something often seen on flippers is "machine gunning". This happens when the player presses and holds the flipper cabinet button, and the flipper machine-guns up and down repeatedly. This happens because the hold side of the flipper coil has a broken winding. Replacing the coil is usually the only recourse.

Why coil diodes? The 1N4004 coil diodes are very important on the flipper coil. Also note the BAND on the diodes - these must be positioned correctly on the flipper coil, and have the correct wires attaching to the correct flipper coil lugs and diodes (the brown power wire always goes to the outside flipper coil lug with the diode band). The coil diodes prevent EMF power from "back flushing" through the power circuit. Without these flipper coil diodes, the game can damage itself (and at minimum act strange).

The two set screws in the flipper pawl that "bite" the flipper shaft determine the resting flipper angles. I like to position the resting flippers so the balls drain off the inlanes symmetrically to the flippers. (Note the flipper bat is parallel to the blue inlane line, and both flipper bats are symmetrical and positioned the same.) Also remember when tightning the set screws on the pawl, that the flipper bat has some up and down 'play'. That is, look at the flipper bat and the pawl as the bread in a sandwich, and the flipper shaft as the 'meat'. The 'meat' should be able to move up and down about 1/16" inside the nylon flipper bushing between the break slices. Otherwise the bread will bind the meat and make the flipper bind (stick) on the nylon flipper bushing, preventing the flipper bat from moving freely.

When both flippers are in the energized position, again I like them adjusted so they are symmetric. If both flippers are symmetric when in the resting position, but are uneven in the energized position, something is worn (like the plunger&link; and/or the coil stop). A complete flipper rebuild kit available from the Pinball Resource should correct this problem, and will also drammatically increase the flipper power.

These are a number of strange problems and solutions that people have mentioned to me.

• "My Bally Globetrotters reboots (resets) while I'm playing a game. The GI lights never go out, but the displays go blank and the MPU does all its seven flashes, and the game powers back on. What is causing this?"
  This common problem is usually caused by one of three things. First, a failing C23 filter capacitor on the solenoid board. Second, poor connectors between the solenoid board and the rectifier board (J3 on both boards). Third, bad MPU board chip sockets or oxidized chip pins. All the repairs for these three problems are covered in the "Before Turning the Game On" sections of this repair guide.
• "On my Eight Ball Deluxe (EBD), the score displays would flicker during attract mode, and a bit during game play. Also occasionally the flipper would cut out, or the flipper enable relay would chatter. I did all the suggested power supply and solenoid board modifications in this document too."
  The problem turned out to be a bad ground path. The later Bally games used a foil covered cardboard as the grounding method for all the backbox boards. In this case, the foil was not consistently making a good ground to the MPU board. Also the green masking on the back of the MPU board was preventing the board from touching the mounting bracket, so the ground mod in this document did not help. To fix this, wires were run from one mounting bracket on each board to a true ground. Also the green solder mask was scraped from the circuit boards around the mounting bracket. Also make sure the "earth-ground-cable" going to the backbox is connected.
• "On my Elecktra, the player one display does not work."
  The backbox "earth-ground-cable" was mistakenly not connected. Attaching this cable fixed the problem.
• "On my Centaur I recently changed the battery because the game wouldn't remember previous credits. After installing the battery, I went to the self-test button to turn the background sound back on. When I pressed the self-test button, it cycled through the game tests. But the next time I pressed the button to go to the bookkeeping functions, the displays went blank and the game locked up. The game boots just fine and gets seven LED flashes too."
  What has happened is the memory in the 5101 RAM has become corrupt. Replacing the 5101 will fix this problem. Alternatively, you can try grounding all the pins together on the existing 5101. Sometimes that works too.
• "My Bally game works fine except for the flippers. They do nothing. All fuses are Ok, as are the flipper EOS switches."
  This is a very common problem. There is a relay on the solenoid driver board that turns on the flippers during a game. The power to this relay is through brown jumper wires between two header pins, adjacent to the relay. If one of the wires break, comes loose, or their solder points go cold, the flippers will no longer work. On the back side of the solenoid driver board, look for the two brown wires and resolder them to the connector header pins.
• "How do I reset the replay values and high scores?"
  This can be done by pressing the red button on the MPU board when the game is powered on. Note this is the button on the MPU board, not the red (self-test) button inside the coin door!
• "I changed the battery or 5101 RAM on my MPU board, and now the game's sound doesn't work; I only get "chime" sounds, where before I got electronic sound."
  On some 1982 to 1985 Bally games, there are several different types of "sound themes" available. These are kept in memory and not set by a DIP switch. If you change the battery on your MPU board, the game will default to the chime sound. See your game's manual for details on how to change to another sound theme.
• "My Centaur powers on but does not finish 'booting'."
  Centaur is a multi-ball game that requires all balls to be in the ball trough and the ball trough switches working before it will finish its power-on.
• "All playfield solenoids don't work."
  First thing to check is the under the playfield fuse might be blown. Next check fuse F4 on the power supply regulator board. Also check it's fuse clip is in good condition with good tension, and is not brown. Now check TP5 (test point 5) on the power supply regulator board. You should get about 43 vdc. If no voltage at TP5, assume the bridge BR3 on this board is bad and replace it.
  After getting +43 vdc at TP5, then check connector J1, pin 6 on the power supply regulator board. This brown wire goes directly to the playfield flipper coils. If you have +43 volts at the connector, but not at the brown wire on the flipper coils, you have a problem in the wiring.
  Also note, +43 volts on some games is also used on the early A8 sound board (Lost World to Dolly Parton). A problem on this sound board (or a bad connector there) can cause problems.
  If your game is not getting the 7th MPU LED flash, that means +43 volts is missing. If you have checked all the above, verify you have +43 vdc on the MPU board on the left (connector) side of R113. Now check the right side of R113. If no voltage there, then replace R113 (2k, 1/4 watt) and retest. If still no voltage, you probably have battery acid damage in this area on the MPU board.
• "A playfield target feature lamp won't turn off."
  A bad SCR on lamp driver board, or bad capacitor on the switch that activates the lamp.
• "Only even numbered score displays show numbers."
  A bad 5101 on the MPU board can cause this.
• "No game credits shown in the credit display."
  There is a DIP switch setting to turn this back on!
• "Player one display always displays the same information that's in the credit/ball display."
  Maybe a problem with the start button! The start button has a piece of "fish paper" (insulating paper) behind the last leaf. This stops the switch leaf from touching the metal support behind the switch. The fish paper had torn, and allowed the switch blade to touch the metal (grounded) support. This has the effect of grounding a row or column of the switch matrix. This will cause all kinds of strange problems! A new piece of fish paper, and the player one display worked fine.
• "My Bally game boots with 7 LED flashes, but I just can't start a new game, even though I have credits."
  Again, if the grey insulating fish paper on the coin door's start button is worn, ripped, or damaged, this will ground that entire switch matrix strobe (row). This can cause all kinds of strange game behavior.
• "All four of the 7-digit score displays on a 1980 Xenon have the same problem; the first digit and the comma right after it are all dimly lit and visible whenever the displays are on."
  Try disconnecting the score display modules one by one and see if the problem goes away. If so it's related to one of the score display units. If this doesn't correct the problem, most like the problem is the MPU board. Check the MPU board's J1 connector for a poor connection. Also U20 or it's associated components like U11 PIA could be faulty. Try swapping U10 and U11 and see the problem changes. Also check the +190 volts at the display board and make sure it is indeed from 155 to 190 volts.
• "Game starts but won't serve the ball to the shooter lane."
  Make sure there are enough balls in the ball trough, and that all the ball trough switches are working. Also make sure the tilt switch is not stuck closed (a game will not start if the tilt or slam switch is closed). Also check for a bad capacitor on these switches. If that is not the problem, chances are this game uses a "solenoid expander board". This board multiplexes a solenoid driver board transistor so it drives TWO coils instead of one. If the solenoid expander board develops a problem, certain coils may not work. See the section titled Solenoid Expander Board section above for more details.
• "I have an EBD that was running fine and on power up the following happens: the ball doesn't move from the trough to the plunger area and the one ball target drops as soon as a new game is started.
  You have a problem related to the Solenoid Expander Relay board under the playfield. The expander switches power between two banks of coils. If the expander isn't working, then only one bank of coils will have power, and the other bank will not. Consequently, when the game goes to fire a specific coil, only the coil on the bank that has power will fire. This is why the #1 target is falling instead of the outhole kicker kicking. First thing to check on an expander problem like this is your feature lamp fuses in the game. The expander relay is powered by the feature lamp buss (6vdc), and if power is absent from part of the feature lamp buss, then the expander won't work and you'll get the above symptoms. Also check the 20A fuse on the rectifier board in main cabinet, which supplies 6vac to the rectifier, then out to the feature lamp buss. If fuse is ok, check for 6vac going into the corresponding rectifier, and 6vdc coming out of the rectifier. Often this rectifier fails due to overheating- I've even seen them get so hot they desolder themselves from the board and open up the circuit. If rectifier shows any signs of old age or dying, replace with a newer heavy duty unit. After this, check for blown fuses under the playfield- fuses and location will vary depending on what version of the game you have- 1981 original, or 1984 remake (standard and limited editions). On later games there was a bank of 3 fuses under the front right corner of playfield. So if you have any feature lamps off on the playfield, that is your clue that power is absent from the expander relay. There are other possible causes for this problem, including burned out or missing lamp next to the expander relay board, bad MOC3011 IC on the expander relay board, bad relay, or bad SCR on the lamp driver board which controls the MOC3011. These are the most common causes, but check for the fuse/power issue first.
• "The wrong coil activates."
  Make sure you have the correct game ROM software installed in the MPU board. Aside from that, chances are this game uses a "solenoid expander board". This board multiplexes a solenoid driver board transistor so it drives TWO coils instead of one. If the solenoid expander board develops a problem, the wrong coils can be activated instead of the correct coil. See the section titled Solenoid Expander Board section above for more details.
• "Coil will not work, yet transistor check out good, and there's power at the coil."
  Bad or flakey connector on the solenoid driver board.
• "Can't get the last flash from the MPU (Star Trek)."
  Solenoid fuse on the rectifier board was good, but a bad diode on the sound board lowered the solenoid voltage enough to prevent the MPU from flashing the last time.
• "Solenoids fire "out of order" in the diagnostic test mode."
  Bad PIA on the MPU board. Try swapping U10 and U11 and see if the problem changes. If so, U10 or U11 are bad.
• "Game boots up and everything is "slow" (power up tune is half speed, and there's a delay from when the ball hits a bumper until it scores)."
  Check all passive components that connect to U12, the 555 timer chip. For example, capacitor C16 can go bad and slow the 555 timer frequency, and increase the service interrupt period. This will "slow" the game down.
• "I am working on a Mr/Mrs Pacman, and I notice the GI (general illumination) lamps flash on and off at certain times during attract mode, but does not come on during game play."
  To make the GI lamps flash on and off, Bally used a Triac bolted to the power supply board in the cabinet. The triac is cycled by the GI lamp flasher board mounted under the playfield. This flasher board is similar to a solenoid expander board. The output of a lamp SCR driver is used to drive a MOC3011 opto coupler that drives the Triac. As with the solenoid expander board, a dummy 555 lamp is connected to the SCR output on the flasher board to insure there is enough current draw to drive the MOC3011. Most likely this 555 bulb is burned out, and causing your problem. Also check the flasher board for bad (cracked or cold) solder connections. Check all the cable connections to the triac too. Lastly you can replace the MOC3011 chip.
• "My Xenon won't run with the sound card and vocalizer plugged in. If I unplug the vocalizer from the sound card, and the sound card plugged in, the game works fine (but no sound)."
  Probably a problem with bad capacitors or a bad voice ROM on the volcalizer board. Also check the 3 vertically mounted diodes in the lower right corner of the vocalizer board. When these go bad the diodes run hot and the game has hum and noise. If you replace the diodes, also replace the voltage regulator to the left of them.
• "My Xenon keeps kicking two balls into play. When one ball drains, it starts adding up the bonus even though the second ball is still in play. What is wrong?"
  If the MPU battery died, and the game is restarted with a ball in the outhole and one in the ball trough, this problem can occur. To fix it, replace the MPU battery. Then remove the 5101 RAM from the game. Put both balls in the outhole. Now put the 5101 RAM back into the MPU socket, and power the game on. This should fix the problem.
• "My Mati Hari displays are strobing wildly. I swapped in a known good driver board, but no change. These displays do not strobe in another game. Game works and plays perfectly otherwise. What is wrong?"
  Connector J1 in the upper left corner of the MPU board could be the problem. Make sure there are no cracked solder joints on the MPU board's J1 connector header pins. This problem could also be related to the rate the displays are being updated. It could be a bad capacitor in the MPU's circuitry for the U12 (555) chip, perhaps the "big green" capacitor.
• "My Bally Space Invaders continues to blow the 20A Fuse for the GI on the Rectifier Board and I'm stumped. It doesn't blow the fuse iimmediately it usually does it shortly after game play begins. Are there any components on the Rectifier Board that may be making it short out? If so what?"
  This is never an easy problem. Break the GI lamp circuits down into sections and isolate the problem that way. Maybe start by disconnecting GI to the backbox, to the cabinet, and/or to the playfield. Eventually you should be able to isolate which part of GI is at fault. Once you've localized it, start disconnecting different portions of the GI circuits on the components in question (playfield, cabinet, backbox). Keep in mind there are two GI lines- red/white, and green/orange. Each one usually runs a couple of lines of bulbs on playfield/backbox. Just start disconnecting a wire from individual lines until problem goes away, then you've got it narrowed down even further. The problem could be shorted wires or sockets, to wires swapped (red to white and vice-versa), even a #44 bulb that had a filament wire sticking out of the bottom of it that was just long enough to short out against the side of the bulb socket when it was installed (this is why I personally like to replace light bulbs with the power on, so you can immediately see a problem bulb). There really aren't any components on the rectifier board that control the GI circuit, unless it's a short on the back side of the board or something. All that really is in the GI circuit on the rectifier board is the 20A fuse. GI power comes off the transformer and goes through fuse. Then it goes to the various pins on J1, J2, and J3, and the test point. Just look for wires rubbing, solder splash on traces/pins on back of board, etc. Of course, if this was the case, then the fuse would still blow even when J1 and J3 were disconnected from the rectifier board. If this isn't happening, then keep looking elsewhere in the game- backbox, playfield, cabinet (coin door), etc.
• "When I turn on my Stern game with a M-200 MPU board, the game will energize coils, blink the lamps, and the displays will count 1,2,3,etc real fast. What is the problem?"
  If all 32 DIP switches on the Stern M-200 MPU board are in the OFF position, the Stern MPU is designed to go into diagnostic test mode at boot-up. So either all your DIP switches are off, or PIA U10 is bad and telling the CPU that all the switches are open (off).
• "My Stern Sea Witch with a M-200 MPU has a problem where it will not boot on occassion. Also the bookkeeping and high score reset button has no effect. There is no MPU board corrosion, and the reset section of the MPU has been totally replaced."
  This person solved their problem by replacing the U9 6800 CPU chip. Apparently the NMI (non-maskable interupt) circuit inside the CPU chip was damaged, maybe by static from a poorly grounded coin door. This CPU problem did not show itself immediately, and let the MPU board function for the most part. The non-functioning reset switch was the give-away here, since this switch connects to the NMI circuit on the CPU chip.
• "My Baby Pacman's playfield GI {General Illumination} lights do not come on when the pinball mode starts, why?"
  The playfield GI lamps are controlled by a Triac, so the lamps can be switched on and off (as the game moves from video to pinball and back to video modes). There is a small driver board which holds the Triac. This board is connected to a switched lamp SCR on the lamp driver board. If all the switched lamps do not work (fuse?), or the SCR on the lamp driver board has failed, the playfield GI lamps will never turn on.
• K.McCormick reports CPU board reboots after the 7th flash. It would run through a normal seven flash boot and then reboot again, always resetting after the seventh flash. Lifted U9 pin 40 and tied it high and it still did it, so it wasn't the reset section. Turns out it was R134, the IRQ pullup down by U11. Solder job looked good, but it didn't have good contact with the trace. Reflowed the connection and fixed the problem.
• "Fathom has worked fine for the past 5 years. Upon powering it up for the first time in over a year it's exhibiting some strange behavior (MPU clean and no acid damage. U6 is a masked rom): If you hit nothing and just lose the ball, no bonus is scored and the ball is served again to the shoooter lane and the ball count is not advanced (this is normal behavior). But if you do hit a target and achieve bonus then lose the ball, the bonus scores. Then just after the ball is served to the shooter lane, the 5X (and sometimes the 4X as well) green cave trap drop targets drop down and the cave trap saucer kicks out the ball that is locked in there trapping the ball against the rear of 4X or 3X green cave trap drop target(s). The game now thinks it is in 2-ball multi-ball and scores double points on everything (as it should during 2-ball multi-ball) However when losing that one ball that you can bat around, the other ball is still trapped behind that 3X (or 4X) green drop target and the game thinks a ball is still in play (and playable). At this point the only way to continue a game is to remove glass to free the ball.
  Answer: It was one of those eight pesky double-dioded coils that was causing my Fathom to drop the 5x target and prematurely spit out the ball from the cave trap lock. At first I approached it from "what coil activating would cause this 5x target to behave this way?" type of approach and started replacing both pairs of diodes on the 3x,4x,5x memory coils and 1,2,3 drop target reset coil and the diode on the solenoid expander board. When all those efforts didn't fix the problem, I replaced ALL of the diodes on the remaining double-dioded coils. Wouldn't you know it but that finally fixed it. DrDave.
• "On my Rolling Stones I can play for about 5 minutes, but then the flippers no longer work. I changed the MPU board and the problem went away. How do I fix my original MPU board?"
  Answer: The MPU connector J4 ties the MPU board to the flipper relay on the Solenoid driver board. Inspect J4 header pins for corrosion, cracks, or other problems (all pins should be shiny). Also check the connector pins on the solenoid driver board at J1, as these pins can crack or tarnish, or have burnt board traces. If that does not fix the problem try swapping MPU chips U10 and U11. (The MPU chip swap idea fixed the problem.)