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

Scope.
This document is a repair guide for Williams and Bally WPC pinball games made from 1990 (Funhouse) to 1999 (Cactus Canyon).

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, what is WPC?, Schematics
2. Necessary Tools
3. Parts to have On-Hand
4. Different WPC Generations
5. Game List
6. Lubrication Notes
7. The Circuit Boards (Board Differences)
8. Introduction to Operation
9. Troubleshooting (quick guide)

2. Before Turning the Game On:

1. Check the Fuses and Power LEDs - Blown Fuses and What Causes them. How to diagnose the "Check Fuses F114/F115" or "F106/F101" error messages. And, "Why at power-on does my game repeatedly fire a coil".
2. Burnt GI Connectors (and WPC-95 GI Diodes)
3. Quick and Dirty Transistor Testing
4. Should I leave my Game Powered On?

3. When Things Don't Work:

1. Removing the Driver board
2. Replacing Components
3. Checking Transistors and Coils (stuck on coils and flashlamps)
4. Game Resets (Bridge Rectifiers, Diodes and Caps)
5. Problems with Flippers
6. The Lamp Matrix
7. The Switch Matrix
8. Infrared Optic Switches
9. Electronic Ball Sensors (Eddy Sensors & Magnetic Reed Switches)
10. Ball Trough Problems (random multi-ball and bad trough LEDs)
11. Dot Matrix/AlphaNumeric Score Displays
12. Power-On LEDs and Sound Beeps
13. "Factory Settings Restored" error (Battery Problems)
14. Lightning Strikes
15. Sound Problems
16. More General Illumination (GI) Problems
17. Test Report & The Diagnostic Dot, Strange Game Behavior
18. Fixing a Dead or Non-Booting CPU board
19. Game Specific & Miscellaneous Repair Tips

4. Finishing Up:

1. Rebuilding Flippers
2. New Coil Sleeves
3. Protecting Slingshot Plastics
4. Cleaning and Waxing the Playfield
5. Playfield Rubber

As early as 1982, Williams started using infrared optic light emitting diodes (LED's) for switches. This is similar technology to what is used in TV remote controls today. These optics have two advantages over conventional mechanical switches: no moving parts, and they can fit in tighter spaces. They also have some disadvantages. They consist of two parts (instead of one part like a micro-switch): a transmitter (the LED that emits the light), and the receiver (the LED that receives and interprets the light). They can also get dirty (from that infamous black pinball dust) and not work. Pin LEDs are always on too. That is, the light emitting half of an opto switch is always powered on, as long as the game is powered on (even when not in play mode). LED's aren't much different than light bulbs; they eventually burn out too.

Several different optos used in Williams games. The "U" shaped slot optos are used for Fliptronics flippers, Twilight Zone clocks, etc. These consist of a transmitter and receiver in one package. The stand-up optos are two parts: the green board opto stand-up is the transmitter, and the blue board opto stand-up is the receiver. The transmitter LED is larger and protrudes further from its case. The single LED shown below is a replacement LED transmitter for the stand-up optos, and for opto boards used in ball troughs, etc. The specs for this infrared LED replacement are also shown in the picture.

Left: Type 1 Flipper Optic board. Again note the orientation of the optics, and how this is different than the Type 2 board, and the vertical metal optic interuptor. This style was seen on games from Addams Family to Twilight Zone. Right: Type 2 Flipper Optic board. Note the orientation of the optics, and the horizontal plastic optic interuptor. This style of flipper optic board was used on WPC games Indy Jones to Cactus Canyon (with only a minor revision around WPC95, using the 5 pin "U" slot Schmitt trigger optic). The plastic activators can be troublesome, as they often warp and don't clear the opto, causing a flipper not to work. Note: When purchasing a replacement flipper optic board, be sure to get the correct style! Many times the newer Type 2 flipper optic board is fitted in older games (all versions of the WPC flipper optic boards are plug compatible)! Replacement flipper opto boards are available from pinballheaven.co.uk and pbliz.com.

Flipper Opto Board Type List.
If a WPC game is not listed below then the game did not use optic switches for the flippers. Note the type1 and type2 interuptors (either plastic or metal) are not interchangable between type1 and type2 flipper optic boards.

Type 1 (interuptor slot runs vertical). Originally used in:

Addams Family Gold (and some regular Addams Family) Creature From the Black Lagoon Doctor Who Dracula Fish Tales Twilight Zone Whitewater

Type 2 (interuptor slots runs horizontal). Originally used in:

Attack From Mars Cactus Canyon Champion Pub Cirqus Voltaire Congo Corvette Demolition Man Dirty Harry Flintstones Indiana Jones Indianapolis 500 Jack*Bot Johnny Mnemonic Judge Dredd Junkyard

Medieval Madness Monster Bash NBA Fastbreak No Fear No Good Gophers Popeye Roadshow Safecracker Scared Stiff Shadow Star Trek Next Generation Tales of the Arabian Nights Theatre of Magic Who?Dunnit World Cup Soccer 1994

Where Optos are Used.
Williams uses optos for lots of applications. WPC Fliptronics flipper buttons are opto activated. These flipper opto boards were implemented on Addams Family, mid-production (many Addams have them, but early models don't). Often clear ramps have opto ball switches. Many pre-1990 Williams drop targets use optos (they stopped using them there because the LED's leads would break from vibration, and the optos would fall off). All WPC-DCS (1993) and later games use optos to sense balls in the ball trough.

Two parts to a opto switch.
Each opto switch has two parts; a transmitter, and a receiver. The transmitter is a infrared LED (light emitting diode). The receiver is a light sensitive photo transistor. The transmitter (LED) is always on when a game is powered on. If the light beam from the transmitter is interrupted, then this registers the switch as "open". Because the transmitter is always on and producing light (and hence heat), the transmitter is the part that fails 98% of the time in a opto switch. The receiver part rarely fails in comparison.

On non-U shaped optos, usually the transmitter LED is mounted in a WHITE plastic case with a small GREEN printed circuit board. The receiver is usually mounted in a BLACK plastic case with a small BLUE printed circuit board.

Cleaning Optos.
Optos can get dusty from the "black dust" inside a game. To clean an opto, use a Q-tip dipped in glass cleaner. Wipe the opto with the Windex-wet Q-tip, then dry the opto with a clean, dry Q-tip. Do NOT use canned air to blow optos clean! The air in these cans is too cold and can damage an opto.

Testing Opto Switches.
Testing infrared optos switches is no different than testing mechanical switches (to a point). Just use the WPC internal test software. Press the "Begin Test" button inside the coin door, and go to the Test menu. Select the "switch edge" test. Activate an opto switch by passing something in front of it to block the light from its corresponding transmitter. The display will indicate if the switch works. Opto switches that are not activated will be displayed as solid "blocks" in the switch test on the dot matrix display (which is basically reverse what you would expect, compared to a micro switch).

12 Volts to the Optos.
If an opto switch doesn't work, first check that the +12 volts is working. If you have blown the +12 volt fuse (either the unregulated 12 volts which provides power directly to the optos, or the regulated 18/12 volts which provides power to the entire switch matrix), the optos won't work. Check fuses F115 and F116 (F101 and F109 on WPC-95) on the power driver board. Also if the unregulated +12 volts is below about 11 volts, the optic switches can work intermittently! If this is the case, usually it indicates a bad BR5 bridge rectifier on the driver board (or bad 12 volt D3-D6 rectifying diode on WPC-95; see the Reset Section of this document for more information on this). BR5/D3-D6 is the unregulated 12 volts (where BR1/D11-D14 is the regulated 12 volts, which could also be the problem since this powers the entire switch matrix, which ultimately reads the opto switches). Remember there is also a large 10,000 or 15,000 mfd filtering capacitor C30 (C8 on WPC95) associated with the power driver board's unregulated 12 volt rectifiers. Check that too for cracked solder joints around the capacitor's leads from vibration (often I will run jumpers to the capacitors and bridges, as shown in the Reset section of this document).

Testing the Opto Transmitter.
On the transmitter LED (the one emitting light), you usually can not check for 12 volts DC right on the opto with a DMM. Unfortunately in most cases the opto voltage will show only about 1 volt (putting the red DMM lead on each leg of the transmitting LED, and the black DMM lead on ground). A better way is to remove the connector going to the opto, and measuring the voltage at the source connector (usually black and gray wires, where the orange and gray pair go to the receiver). If there is no 12 volts present (and other optos in the game work), there is either a break in the ground or 12 volt connection going to the transmitting LED. Also sometimes the optos get cold solder joints (from vibration) on their associated circuit board. Resoldering the opto leads can fix this (assuming the opto lead going to the LED itself hasn't broken). Heck vibration often breaks the wire off the opto board too.

If there is +12 volts going to the transmitter opto but the switch does not work, there is a good chance the transmitter LED has failed. Radio Shack sells a $5 credit card sized "infrared sensor". MCM Electronics also sells one, #72-6771, for about $7 (800-543-4330 or www.mcmelectronics.com). If you put this card right in front of an opto transmitter, the opto's emitting light can be seen; the light will show on the colored band of the sensor card. Also, a digital camera or a camcorder will usually show infrared light from the transmitting opto, if the digital camera has a small LCD screen used to show images "live" (but personally I like using the opto cards better).

If there is +12 volts (hint: do other optos work?), and the opto switch doesn't register in the diagnostic test, your opto transmitter is probably burnt. The receiver side of an opto switch rarely dies. That's because it only senses light, and doesn't produce light. The transmitter will be the offending unit 98% of the time. Remember the opto transmitter is powered-on all the time the game is turned on, and it can burn out just like a light bulb can burn out.

Reversed Leads on the Transmitter.
Another common fault of the LED opto transmitters is having the wires reversed. Yes it does matter which wire goes where. And don't think you are the only one that can make this mistake. I have seen NOS parts right from Williams where they have soldered the leads reversed on the opto transmitter! Note usually having the leads reversed does not blow the transmitter. There is a flat spot on many LED transmitters too, signifying which side to connect ground or 12 volts. But I have also seen some manufacturers have the flat side reversed! So if in doubt, try reversing the black and gray leads on a non-working opto transmitter.

Testing the Opto Receiver.
The simplest way to test the opto receiver is to first put the game into the "switch edge" test. Then block the opto transmitter with a piece of black electrical tape. Now shine a penlight flash light into the opto receiver, or a TV remote control (which is basically an infrared flashlight). The switch should "close" (go from a solid block to a small dot on the DMD screen). When you remove the light, the switch should "open". If the LED receiver is working properly but the switch does not work, often the opto transmitter has burned out.

Another way to test the opto receiver is using a DMM. First block the opto transmitter with a piece of black electrical tape. Put the black DMM lead on ground (the metal side rail of the game works well). Put the red DMM lead on one leg of the opto receiver (gray wire). One opto receiver leg should show 12 volts DC, and the other opto leg should show close to zero volts (orange wire). Keep the red DMM lead connected to the "low" (zero volt) opto leg. Now shine a flashlight into the opto receiver. The DMM should now go to 12 volts DC, and when the light is removed, go back to near zero volts. If this does not happen, the opto receiver is bad. Or if 12 volts is seen on both opto receiver legs, the receiver is bad (or there is direct light shining into the opto receiver).

Testing the infrared opto transmitters on a 7 LED ball trough assembly. The LED's can be seen lit in this photo, but you won't be so lucky with the naked eye. That's why this Infrared Sensor card or a digital camera/camcorder is so handy. Note in the digital picture below, the red and blue infrared LEDs are lit. With the naked eye, the LEDs do not look lit. This card is available from Radio Shack #276-1099 or MCM Electronics #72-6771, about $7.

Opto Transmitters on Newer WPC games.
Older WPC games use optos with straight resistive photocells. Some newer WPC95 games use a transistor gate photocell. This means the internal transistor can die, even if the photocell part of the opto is OK. Keep this in mind; if an opto transmitter tests good (with your Radio Shack or MCM test card), the opto could still not function properly. Replacing the opto is the only thing that will fix it. This is rare and hard to diagnose, but if everything checks out this could be the problem.

Opto Board (the Opto Receiver and Transmitter Tests Good, now what)?
If the receiver tests good with the penlight flashlight, and the transmitter tests good with the infrared sensor card, there is one more thing that could be wrong. This would be the opto board. Usually before messing with the opto board I make darn sure that the optos themselves are not the problem.

I typically do this (for non U-shaped optos) by taking a new opto (receiver or transmitter), and holding its legs to the back of the opto board. For the transmitter I can check it with a digital camera or an opto sensor card. For a receiver I can test it with a penlight (or the other tests given above). Unfortunately if the opto board has a problem, these tests may not work...

Most of the newer WPC games have a seperate board mounted under the playfield called an "opto board". These have some LM339 voltage comparitor chips and diodes and resistors. If this board fails it can really confuse the game. Also games Indiana Jones to Demo Man usually have an opto board under the playfield AND the trough board is essentially a second opto board. Both these board have LM339 chips, which can be problematic. (After Demo Man starting with WCS94, the trough opto boards no longer have LM339 chips, as these were all moved to the under-playfield mounted opto board. So the trough opto board becomes less of an issue.)

There are many different 'flavors' of these opto boards, so it's hard to give an exact test for the opto board. But there are some general things that should be looked at:

• Make sure the CPU board is not the problem. I always put the game in switch test T1, remove all the connectors from the bottom of the CPU board, and cross each switch column/row (this test is described in the switch matrix section). This rules out the CPU board as the problem. I always start there.
• After eliminating the CPU board as a problem (and reinstalling the CPU board connectors), I remove all the connectors from the opto board and make sure the switch matrix test T1 operates cleanly with no errors (other than the missing optic switches). If problem free, then the optic board(s) is starting to look like the problem.
• Don't forget Indiana Jones to Demo Man games essentially have two problematic opto boards: the under PF mounted board and the trough opto board. Games after this WCS94 and later trough boards do not have any LM339 chips on their trough opto boards, so these trough boards don't implode like the Indiana Jones to Demo man trough boards.
• Opto board LED - there's a red LED showing the opto board has power. It should be on.
• Opto board has many IDC connectors. It is not uncommon for these connectors to get a wire broken/pulled, causing an opto (or set of optos) to not work. To test this (game off), pull one female connector just slightly off its male header. Then use a DMM and check for continuity from one male header pin to where the wire goes. Repeat for all pins. No continuity, and you have an IDC connector problem (very common).
• Check the back of the opto board and make sure all the male header pins do not have cracks in the solder where they attach to the circuit board.
• Using a DMM set to diode function, make sure all the 1N4004 diodes on the opto board test correctly. They should read .4 to .6 in one direction, null in the other.
• Check all the resistors with a DMM.
• Check the traces from the header pins to the resistor/diodes. It is not uncommon for a trace to be broken or burned on an opto board. Actually this is a very common problem.

If everything checks out, that only really leaves one thing left: the LM339 chips on the opto board. I generally replace all the LM339 chips (and use sockets!) on the opto board (there are usually two to four of these chips on the opto board). Unfortunately the LM339 chips are not that easy to test, since they're dealing with voltage levels. But as long as the voltage levels on the outputs of the LM339 are stable (not pulsing and not fluctuating), the truth table for the individual comparators can be tested with a DMM (inputs) and a logic probe (output).

Other Problems.
Often the source of ground for the 12 volts going to the opto switches can be tricky to find. And if this ground connection fails, several or all optos will not work. For example on Indiana Jones, the drop target board and flipper opto boards get their ground from the Fliptronics II board's J905 connector. If this connector is bad or removed or off by one pin, there would be no ground optos ground, and none of the optos would work.

WPC Fliptronics Flipper Optos.
Flipper opto boards were implemented on Addams Family, mid-production. If a WPC Fliptronics flipper doesn't work, and it's not a coil, transistor or wiring related problem, you should suspect the flipper opto board. This board has two "U" shaped optos that detects the flipper button. These boards are all made with two optos, even if the game only has two flippers instead of four.

Use the infrared sensor card to determine if the opto is working on the flipper board. If you suspect a problem with this opto (and don't have a infrared sensor card), swap the left and right flipper opto boards, and see if the problem moves to the opposite flipper. Note: both flipper opto boards must be plugged in for this test to work! Flipper opto power is run from the backbox, through the left flipper opto board, to the right flipper opto board. Flipper opto ground is run from the backbox, through the right opto board, then to the left flipper opto board. Hence both opto boards must be plugged in for them to work!

If indeed one of the flipper optos is bad, and the game only has two flippers, reverse the two optos on the bad flipper opto board. One of the optos will be unused since the game only has two flippers, instead of four. Mark the bad opto, and its position on the opto board. As a general rule, the "top" opto on the flipper board (the opto farthest away from the two resistors) is the LOWER flipper opto. Unsolder both optos and move the good opto into the marked (upper) position on the flipper opto board.

The only problem with doing this is a potential switch error with the bad opto. Even though the second flipper board opto is not used, many Williams games check for this switch, and will report it as "bad" in the game's power-on test report (even though the game may not use it). Also some games use the "unused" flipper opto for scrolling through the high-score initials. So ideally it is best to just replace a bad opto instead of swapping.

Weak Flippers and Bad LM339's on the Fliptronics Board.
On WPC fliptronics to WPC-S board, chips U4 and U6 (LM339) on Fliptronics II board can fail. On WPC-95, these LM339 chips are on the CPU board at locations U25 and U26. This will make the flipper opto boards seem like they are not work. Swap the two flipper boards to test this. If the problem doesn't change, suspect the LM339 chip(s). These LM339 chips can also become "leaky". This will make flippers seem very weak. A bad LM339 can also give the indication that the EOS switch is bad.

If there is a marginal flipper switch reading, this causes the high powered side of the flipper to rapidly oscillate between on and off. The holding side of the flipper coil never engages. This problem will cause the flipper coil to get very hot in a short time.

Opto Wavelength.
Optos come in basically two different wavelengths: 880 nM and 940 nM. The 880 nM optos came first, but the opto industry has largely moved to the newer 940 nM wavelength in the last few years. Williams used 880 nM in nearly all applications, but this older wavelength is harder to purchase today. The newer 940 nM standard minimizes false triggering from sunlight and incandescent light, and can operate at longer distances from the opto receiver. Also the newer 940 nM wavelength works better in foul air (high humidity and polution). The only down side to the new standard is if the application has a newer 940 nM transmitter, and an older 880 nM receiver, this can cause problems.

Replacement Infrared LED Optos.
The infrared LED transmitters have the industry part number QED123 (Fairchild, MOT and QT brands). These are 5mm sized LEDs. The color of the LED will range from pink to yellow to blue. They also have one flat size, which denotes the "K" (cathode) lead, which is the shorter lead. The flat side of the LED is usually marked on the circuit board too. The other non-flat side lead should be longer, and is the "A" (anode) lead. Typically in a WPC game, the black switch matrix wire goes to the "K" (flat side) of the infrared LED. The gray wire goes to the "A" lead. Radio Shack sells the infrared LED (transmitter), part number 276-143 (or 276-143c), $1.69 (replaces Williams A-14231). Also Mouser sells Fairchild QED123 LEDs.

Replacement Photo Transistors.
The photo transistor (receiver) have the industry part number QSD124 (Fairchild, MOT and QT brands). These are 5mm sized LEDs. The color of these are usually black. They also have one flat size, which denotes the "E" (emitter) lead, which is the shorter lead. The other non-flat side lead should be longer, and is the "C" (collector) lead. Typically in a WPC game, the orange switch matrix wire goes to the "E" (flat side) of the infrared LED. The gray wire goes to the "C" lead. Radio Shack also sells an infrared transistor (receiver), part number 276-145a (or 276-145), $0.99 (replaces Williams A-14232). When mounting these, the flat edge goes in the hole furthest away from the hole that has the notch drawn on the circuit board. Mouser sells the Fairchild QSD124 photo transistor. Digikey also sells a receiver, part number PN104-ND. When installing this photo transistor remove the center pin before installing. Just wiggled the center lead back and forth until it breaks off at the base. Install this part so the notch at the base lines up with the notch drawn on the circuit board.

Radio Shack also sells a combo package with both the receiver and transmitter, part# 276-142, $1.99. This is essentially the #276-143 and #276-145 parts combined into one package, at a discounted price. The word from Radio Shack is part number 276-142 will change. The the old stock LED style transmitter/receiver is discontinued, and replaced by a "U" shape style opto (though the part number is still the same). This "U" style opto will work on some unique WPC optos (see the "Radio Shack 'U' Opto" section below), but nothing else. But lately the U optos have again been replaced with separate LED style optics.

Lastly, it has been reported that the Radio Shack #276-145a photo transistor is not as sensitive as the stock Williams part. Apparently if the distance is greater than two inches between the two optos, often the photo transistor will not register the infrared LED. In conclusion the #276-145a photo transmitter is not sensitive enough, since using a Radio Shack #276-143 LED and a Williams photo transmitter does seem to work at greater distances. Your mileage may vary, as Radio Shack parts can often be inconsistent.

How Can I tell a Transmitter from a Receiver?
In case you have optos laying around and you don't know if they are transmitters or receivers, a simple diode test with a DMM across the leads will reveal which LED type it is. A transmitter will check about 1.5 across the leads one way, open (no reading) the other way. A receiver will check open (no reading) either direction unless you shine a flashlight on it, then it will check open one way and shorted (0) when the leads are reversed.

WPC-95's Five Leg "U" Shaped Slot Optos.
Starting with WPC-95, Williams changed to a "U" shaped Schmitt Trigger opto (five legs in total, three legs on the receiver, two on the transmitter). The Schmitt trigger optos will not oscillate (turn on and off quickly) when the optics gets dirty/old (they either work, or don't work).

The problem with the older 4 legged flipper optos when dirty/failing was the oscillation. This would cause the flipper coils to get low amounts of power continuously during game play (like the player was pressing the flipper button on and off continuously, and very fast). This would cause the flipper coils to get hot. It would also make the flippers weak (because when the player really did press the button, the oscillation would try and turn the flippers off very quickly too!).

The older 4 legged "U" optos also caused other problems on games that used the flippers to control playfield toys. For example on Indiana Jones, a dirty/failing flipper optic could cause the mini-playfield Path of Adventure (POA) to "stutter" when the player tried to move it right or left with the flipper buttons. This was a confusing error because in test mode, the POA would act normally (because the flipper buttons were not involved in the test - if the POA stutters in both game and test mode, the two 4 legged optos on the POA board could also be bad).

Because of the oscillation problem, Williams changed to a five leg Schmitt trigger "U" shaped opto with WPC-95. This solved the dirty/failing optic flipper problem, and made diagnosing flipper related optic problems easier. The new five leg optos usually either work, or don't work.

Replacement 5-Leg "U" Shaped Slot Optos.
The Williams part number for 5-leg optics is 5490-14575-00 (or QTE734, QT724, QT850, or QT902 has been seen), and is called "IC Opto Integ Schmitt 10mA". Replacement five legged optos are available from Mouser #512-H22LOI, which is a Fairchild Semiconductor part #H22Loi.

Replacement 4-Leg "U" Shaped Slot Optos.
Unfortunately, "U" optos are fairly expensive (compared to micro-switches). For example, if you are repairing your Twilight Zone clock (which means replacing all eight of the "U" shaped optos), this can get costly.

The industry part number for the pre-WPC95 four leg "U" shaped optos is QVE11233, with a standard sensitivity of .0110. Unfortunately, Williams requires a higher sensitivity opto for their applications. This means the cheap $1 optos from most electronic supply houses may not work, as their sensitivity rating isn't high enough. If you are shopping for these "U" optos, keep this in mind. You should be looking for part number QVE11233.0086, where .0086 is the increased sensitivity rating. This is the exact part used in Twilight Zone clocks, one of Williams most sensitive opto applications. This means a QVE11233.0086 "U" opto should work every where else just fine!

As a side note, the original Williams optos were made by Motorola. But around about 1996, they split their opto electronics division into a new company called QT Optoelectronics. Then in early 2001, Fairchild bought QT. What does this all mean? Well it means the "original" Motorola brand "U" optics are all gone, but there is a fairly good stock of QT brand "U" optics around (which are identical to the original Motorola brand, differing in name only). Fairchild unfortunately has discontinued the older optic line, and no longer makes an exact duplicate of the original Motorola/QT "U" slot optos. They do make some similar optos, but the leg spacing and specs are slightly different (but they may work!)

Generic "U" shaped slot optos (QT brand) with the lower .0086 sensitivity are available from Mouser (www.mouser.com, part number 512-QVE11233, $0.90) and Digikey (www.digikey.com, part number QVE11233QT-ND, $0.90). Unfortunately, these most often do not work in Williams pinball applications.

A replacement "U" shaped slot opto that works 100% of the time for sure (and mounts dot on the opto to the dot on the PCB) is available from dragster_73@hotmail.com, Prestige Industries (800-456-7277 www.pinball4u.com) or Competitive Products (800-562-7283 www.competitiveproducts.com). At about $5 each (QT brand, long leads too, for the Twilight Zone clock), these are a very good replacement for nearly every Williams pinball application.

The Radio Shack "U" Shaped Slot Opto.
Radio Shack used to sells a "U" shaped four leg opto, part# 276-142, $1.99. The "new stock" of this part number is NOT a "U" shaped opto, but is essentially the LED style #276-143 and #276-145 receiver/transmitter combined into one package. The word from Radio Shack is the "U" style was discontinued, and replaced by the LED shape style opto (though the part number is still the same). The old R.S. U opto does work in the Twilight Zone clock and in the flipper opto boards (four leg variety, prior to WPC-95), with some minor mounting modifications. The spacing on the bottom part of the "U" of the opto is slightly different, and some mounting adjustments are needed to offset this (especially on the Twilight Zone clock). The old Radio Shack "U" optics is also a perfect replacement for the Indianapolis 500's lighted target. The style of optic used on this target is exactly like the Radio Shack part.

Installing the old Radio Shack "U" Optic.
Installing the Radio Shack optic is "backwards". This opto has a "dot" silkscreened or impressed on its side. Normally, this opto dot should line up with the dot silkscreened on the printed circuit board. But in the case of the Radio Shack #276-142, this dot goes OPPOSITE of the circuit board dot. On the Indy500 targets, the board does not have a dot. Instead the dot on the Radio Shack Opto goes to the "A" terminal (instead of the "C" terminal of the original Williams Opto). If there is any question you can confirm the orientation using your DDM. Testing with the red DMM lead on "A" and the black DMM lead on "K". This will show a reading of about "1". All other combinations get a reading of "0".

On the Radio Shack optos can not be found for an Indy500 fix, drill out the rivets and remove the "R" opto case from the target board. Then take a 4-legged Twilight Zone opto and pry off the case. This will expose the "guts", which can be transplanted to the Indy 500 opto board. Note the cover does NOT need to be put back on the opto.

A Williams flipper switch opto board. The "top" (lower flipper) opto has been replaced. Note the "dot" markings on the flipper opto board. Many replacement optos will have a corresponding "dot" or "notch" in the opto, which aligns with the board's dot. If the new opto does not have a dot/notch, align the "S" and "+" leg of the opto closest to the circuit board's dot.

Installing "U" Shaped Optics (Other than Radio Shack's "U" Optic).
There are two positions that a "U" shaped optic can be installed. Putting the optic in "backwards" usually does not ruin the optic, but it will prevent the optic switch from working! Many replacement optics have a "dot" or "notch" on one side of the optic. This dot/notch should align with the dot silkscreened on the circuit board (there are exceptions to this, such as the Radio Shack #276-142 "U" optic, where the optic's dot goes OPPOSITE of the board's dot, but this is a rare exception, see above). Slot optos use a dot for the collector, and a "S" for the sensor, "E" for the emitter, and unmarked is the cathode.

If the new opto does not have a dot/notch, there should be "S", "E" and "+" markings on the top of the two legs of the optic. In this case, align the "S/+" leg of the opto closest to the circuit board's dot.

After the new optic is installed and in the game with the power on, use the Radio Shack infrared card to find the transmitter leg of the optic. The newly installed optic should have its transmitter leg in the same relative position as the other original adjacent optic(s).

The "U" optic on the left is an original base mounted Williams optic (this one from No Good Gofers). This style of "U" optic case is sometimes hard to find. But the case can be pried apart, reused, and new optic guts placed inside. The optics on the right are the replacement "guts" for the "U" shaped optic (taken from a regular "U" shaped optic). The original case is then set over top and snapped into place. Alternatively, the plastic case can be discarded, as shown here!

"U" Optic Replacment Alternative: Reusing the "U" Optic Case.
The U shaped optic's black plastic case can be reused, and just loaded with new optics "guts". Guts can be taken from other new "U" optics (that use a different style case), or the optic guts can be purchased separately. The "U" case pries apart from the bottom, using a small screw driver. The new guts are then placed inside. When doing this be careful to identify which is the transmitter *before* taking the original optic apart. This way the new transmitter and receiver can be inserted in the same positions, and the case cover installed with the "dot" in the correct location. In some applications, the black plastic case may not even need to be replaced (if there is no risk of a pinball hitting the optics, and no risk of stray ambient light). Gregg Woodcock sells these individual optic guts at www.ClassicCoinOps.com/wmsoptos.htm. The transmitter (Gregg's are red) goes into the spots marked "A" and "K". The receiver (Gregg's are clear) goes into the spots marked "C" and "E".

Starting in 1993, Williams starting using "eddy sensors" to determine when a pinball rolled under a portion of the playfield. A eddy sensor is a electronic switch; it has NO moving parts. It can sense when a steel pinball passes over it, and acts like regular mechanical switch. Star Trek Next Generation (STNG), Road Show and Theatre of Magic (ToM) uses these eddy sensors. These electronic switches are used in playfield areas where a regular mechanical switch is not practical or visually pleasing.

An under the playfield eddy sensor control board as used on Roadshow, STNG, Theatre of Magic. Note the potentiometer and LED. The connector on the left goes to the actual under-the-playfiled mounted "sensor" (see pictures below) that tell this board there is a ball above it.

Adjusting Eddy Sensor Boards.
Often eddy sensors can go out of adjustment and become less sensitive. This can cause the eddy sensor to not activate when a ball passes above it on the playfield. To adjust an eddy sensor do this:
• On the under the playfield eddy sensor control board, turn the potentiometer counter-clockwise until the LED just turns on.
• Now turn the potentiometer back clockwise until the LED just turns off.

That is all that is required to adjust an eddy sensor. To test the sensor, put the game into WPC diagnostic's first switch test. Then move a pinball over the playfield area where the eddy sensor is located. The switch should activate on display. Also from the bottom of the playfield, the eddy board LED should go ON as a ball passes in front of the eddy board's senssor (this can be seen anytime, the game does not need to be in switch test.)

Different R1 Eddy Sensor Values (Fine Tuning).
Because the ball sensors are different on some games, the value for R1 on the Eddy sensor boards can be different. For example, on Star Trek Next Generation and (two of the eddys on) Theatre of Magic), R1 is 4.7k ohms (these games uses the small ball sensor). But on Roadshow and the ToM trunk, which uses a much larger ball sensor, R1 is 2k ohms. So if you switch an Eddy board between these games, the Eddy R1 resistor may need to be changed to the correct value.

The purpose of the R1 resistor is to make the adjustment pot "centered" for the particular ball sensor. For example, if you use a 2k ohm R1 eddy board in STNG, the adjustment pot will be turned almost all the way up (with very little adjustment range). It still works most of the time, just the adjustment range is not centered.

With this in mind, I once had a Roadshow where I could not get the eddy board's LED to turn off, no matter where the adjustment pot was moved. Normally Roadshow uses 2k ohm R1 resistors for all three eddy boards - but in this case I had to replace the R1 resistor with a jumper wire (0 ohms). This put the adjustment pot about dead center, and the eddy boards worked great (with the 2k ohm R1 resistor, the eddy boards would not adjust, and hence would not work.)

Left: the actual sensor that senses the ball. This is a smaller sensor as used on the outlanes of many games. Right: another type of eddy sensor that senses the ball. This sensor is used in Theatre of Magic and covers a wider area.

Second Generation (Auto Adjust) Eddy Sensors.
Games made in 1996-1998 (like Sacred Stiff, Cirqus Voltaire, Monster Bash) use a second generation Eddy sensor. Instead of having a potentiometer under the playfield to adjust sensitivity, these are "auto-adjust" Eddy sensors. This style of Eddy sensor is better, as they do not go out of adjustment. But they also use more logic parts, meaning more electronic parts to potentially fail. You can buy replacements for these boards at PinBits. The new auto-adjust boards are plug compatible with the older manual adjust eddy boards (generally speaking), but some resistor values may need to be changed (again R1).

Twilight Zone Eddy Sensors.
The eddy sensor that causes the most trouble in Twilight Zone is the sensor by the ball trough (switch# 26). Note eddy sensors were used as early as Twlight Zone. The eddy sensors in TZ are different than the later sensors, and do NOT have an adjustment pot and they are not auto-adusting (they also are called a different name, like the "Trough Proximity" board). On the ball trough sensor, it is actually two boards: the sensor board, and the driver board (the driver board is the one with the two molex connectors; a picture of the two boards is here). The only adjustment you have on the TZ eddy is moving the sensor board closer to the ball. This can sometimes fix many problems.

Another common TZ problem are the molex connectors on the driver board. Just taking the two pin molex connector off and putting it back on its header pins will usually the problem. If not, this small board often needs to have its molex header pins resoldered. The solder joints on the board's header pins can crack. Also, it is possible for the TDA0161 (Williams part number 5370-13452-00) chip to die on this board. If you don't want to replace just this chip, the whole proximity driver board is available for under $15.

Modifying your Twilight Zone Eddy Sensor.
Ray Johnson ( http://www.aros.net/~rayj/action/tech/tz_prox.htm) came up with this cool modification. It adds a small PC-board trimmer pot to the sensor PC board. This allows you to always be able to adjust the sensitivity of the sensor. Here are the steps:

1. Buy a small PC-mount trimmer pot. Get the lowest resistance rating you can find (something around 100 ohms would be ideal, but the most common "low rated" pots are about 1k ohms). Some of these small pots can be very, very touchy, so it's best to get one that has a low resistance rating (like 100 ohms), which allows you a good accurate adjustment. The average amount of resistance you'll want from the pot is around 20 to 30 ohms, so check your pot with your meter first to make sure it will let you adjust it easily to this value.
2. With the power off, remove the sensor board from the game. Two hex-head screws hold it to the underside of the playfield.
3. On the component side of the board, cut the trace between the connector pin and the sensor. This is the only trace on this side of the board, so you can't miss it. Use a sharp knife, or X-Acto blade, to slice through the trace. Use multimeter to make sure there is no continuity after you've made the cut.
4. Scrape some of the insulation off the trace that leads to the sensor (see image above). Remove enough to adequately solder a jumper wire onto the bare metal of the trace. Click here for a picture of this step and the prior step.
5. On the solder side of the board, use a small marker to mark the position of the three legs of the trimmer pot onto the PC board. Drill three holes in the board through which you will mount the pot. Use a very small drill bit (1/16" or smaller). Click here for a picture.
6. Install legs of pot through the holes you drilled in the PC board. Bend the legs on the other side of the board to hold the pot on the board. Click here for a picture.
7. On the component side of the board, connect two small jumper wires to the pot. The first jumper wire will come from the back side of the pin on the connector (the one with the trace going away from it). The other jumper wire will come from the other side of the trace that you cut, where the insulation was scraped away. Connect the other ends of the jumper wires to the pot. One goes to the middle leg, and the other goes to either side leg (doesn't matter which side leg). Click here for a picture.

The modification is now done. Install the sensor board and the cable that goes between it and the driver board. With the sensor board installed, the pot should be easily accessible with a small screwdriver. Now power on the game. With NO balls in the ball trough, adjust the installed pot just as described above (for the newer Eddy sensors):

• Turn the potentiometer until the LED just turns on.
• Now turn the potentiometer back until the LED just turns off.
Test your work by putting a single steel ball in the ball trough. The LED on the Proximity driver board should come on. Move the ball away from the sensor and the LED on the driver board should turn off.

Magnetic Reed Switches (beyond Eddy sensors).
Starting with SafeCracker and NBA Fastbreak, Williams started using a different ball sensor switch instead of Eddy sensors. This change came about because the Eddy sensor had reliability problems. Even the later self-adjusting Eddy sensors were not as reliable as needed.

Instead, Williams changed to a Magnetic Reed Switch (MRS) with Safecracker and NBA Fastbreak. This style of switch is contained in a black epoxy package, about 2" long, and 1/2" wide. Like an Eddy sensor, it can sense when a pinball is near the switch. Games which used this reed switch include NBA Fastbreak, Safecracker No Good Goofers, Cirqus Voltaire, Cactus Canyon and Star Wars Episode I. I believe these are the only games that used the reed switch.

MRS switches uninstalled, Williams part number 20-10293 (the "9937" is a manufacturer date code).

The advantage to the MRS is great; there is NO additional circuitry needed for the switch (unlike Eddy sensors). And MRS switches generally do not break, fail or wear out. This makes a MRS more reliable and cheaper. A MRS plugs directly into the switch matrix, just like a micro switch. It doesn't use an additional circuit board, or even a diode! Williams used the MRS under plastic ramps and under playfields.

a MRS switch under a Cactus Canyon ramp.

There are some drawbacks to a MRS though. First, it does not read a really fast moving pinball as predictably as an Eddy switch. For this reason, often Williams puts two MRS switches in parallel to compensate for this. Also the ball must roll directly over the MRS switch. Because the switch is only 1/2" wide, again two switches are often used in parallel to make sure the pinball is "seen" by the MRS. Finally, a MRS must be very close to the ball. If mounted under the playfield, they can only sense the ball through the thickness of a playfield insert or a plastic ramp, and not through wood (which apparently is too dense). The mounting for the MRS under the playfield is often two rubber grommets. If a grommet falls off, this will not allow the MRS to be snug against the playfield, making ball detection difficult.

The ball trough is the area where the balls drain and collect when a game is over. Up to 1993, Williams used a conventional ball trough design. This old style ball trough used mechanical switches to sense the ball's presence. It also used two coils to move the balls; one to kick the ball from the outhole to the trough, and another coil to kick the ball from the trough to the shooter lane.

Starting in 1993 with Indiana Jones, a new ball trough design was used that instead relied on gravity to feed the balls into the trough. This saved one coil (the outhole coil was no longer needed). The new design also used opto switches instead of mechanical switches. This allowed one ball trough design to be used in all Williams games, regardless of the number of balls used in the game. The ball trough could now comfortably hold from one to six balls (depending on the game; most used four to six balls).

The two opto boards used on either side of the ball trough to sense the balls. Note the large blue resistors used on the top board. Often these resistors can vibrate and break. This will give the opto board false ball senses or no ball senses.

Ball Trough problems (Random Multi-Ball, Drained Ball not Sensed, Game won't Start).
When the opto ball trough was first used on Indiana Jones, Star Trek Next Generation, Judge Dredd, Popeye, and Demoman, William bolted the opto boards right to the side of the trough. The vibrations from the trough often caused the leads on the large blue two watt resistors and the infra-red LED's on the opto transmitter board to break. This would cause the game to start random multi-ball at just about anytime during the game. Often the game would never end (because the trough would not reconize when all the balls had drained).

To fix this problem, Williams redesigned the attachment points for the two opto boards. Instead of being bolted directly to the trough, the mounting holes on the opto boards were enlarged (and one hole moved). Then rubber gromets where inserted into the holes, and short metal tube bushings where inserted through the rubber gromets. When the opto board bolts where tightened down, they tightened on the metal tubes. This allowed the opto boards to "float" on the rubber gromet, reducing vibration considerably.

Also be aware that on Star Trek Next Generation if fuse 103 on the Power Driver Board is blown (3A slow blow), the game will not start and will constantly throw out balls. Fuse 103 powers the solenoid which controls the upper diverter on the under-the-playfield diverter. Without a working diverter, the game can't load the balls where it wants, and the game will attempt to load and reload balls continually.

Also another tip concerning Indiana Jones: Check the front right switch on the bottom side of the mini playfield. Balls hit it underneath and mash the wires/diode/switch lugs together creating a short. Since this mini-PF switch is in the same row as the ball trough jam opto in the switch matrix. This can cause the game to continually kick out balls because the machine thinks the ball jam opto has a ball in front of it, and kicks out another.

The front right mini-playfield switch on Indy Jones. This switch's leads often get crushed by flying pinballs, shorting them together. This can cause all kinds of switch matrix problems including continual multi-ball and switch matrix confusion (multiple switch closures by a single switch closure).

Later Opto Board Design.
To make the opto boards more resistant to vibration, starting with World Cup Soccer 94, Williams moved all the electronics off the opto boards and onto a separate board. This meant that only the optics were on the trough opto boards, and no other components. No longer could the large blue two watt resistors crack from trough vibrations. Unfortunately, Indy Jones, Star Trek Next Generation, Judge Dredd, Popeye and Demoman all use the older ball trough opto boards with the easy-to-break blue resistors and bad mounting design.

Check the Shooter Lane switch.
Though usually not the problem with random multiball (a closed shooter lane switch does not get the ball to the shooter lane), it's a good idea on most WPC games to make sure this switch is in good condition and working. Use a ball to test the switch (in switch test T.1).

Ball Trough Divots (Indy Jones to Cactus Canyon).
Another problem with the new ball trough design is "divots". As the pinballs fall into the ball trough from the playfield, they eventually make divots into the metal. This can cause the balls to hang and not roll the length of the ball trough and down to the shooter lane upkicker coil. All sorts of weird game problems can occur from this. The most common is trying to start a game by pressing the start button, and the game responds with "pinballs missing", or a game that doesn't end when the ball drains. Random multi balls can be caused by this problem too.

At first look, where the balls fall from the playfield into the trough would seem to be the problem. But that really is not the big problem; where the balls rest in the trough "V" slot can develop very small divots or nicks in the metal. All these newer game use four to six balls, and often a pair of nicks in the metal can exist where each ball rests in the trough!

To fix this, a Dremel tool or a hand file can be used to grind the divots out of the metal. After the nicks are ground out smoothly, sand the sides of the "V" in the trough smooth with 220 or 320 sandpaper. If this doesn't work, order a new ball trough, part number A-16809-2. This newer design of the ball trough should last longer and divot less.

On the left blue circle is where the balls slam down into the trough. But the big problem is the two smaller blue circles, center and right. These very small nicks will stop the balls from rolling down the trough as a single ball is fed to the shooter lane. These causes all the balls to hang and not roll the length of the ball trough.

Buying a Ball Trough Mounting Upgrade Kit.
If you want to upgrade your Indiana Jones to Demo Man ball trough to the current board mounting design (which can help solve random multi-ball problems), order an upgrade kit, part# A-18244. This includes two new opto boards, and all the mounting hardware needed (the mounting hardware is absolutely necessary). At $50, this is an expensive kit!

Modifying the Existing Trough Boards Mounting Instead.
Modify the existing trough boards can be done for much less money. The parts can be ordered from Williams:

• (6) Metal bushings, 3/16" outside diameter and 3/16" long, Williams part# 02-4975, $0.28 each.
• (6) Rubber grommets 3/16" inside diameter and 1/4" to 7/16" outside, Williams part# 23-6626, $1.02 each.
• (6) Trough board mounting screws (same #6 size/thread as the originals, just 3/4" long).

These parts can be bought locally. Rubber grommets can be bought at any decent hardware store in the electrical department. The inside diameter grommet hole (the important part) is 3/16". The outside diameter can vary from 1/4" to 7/16". The metal 3/16" bushings can be bought at hobby shop that sells 3/16" brass or aluminum tubing (usally in 12" lengths), used for hobby applications. This tubing cuts easily with a Dremel cut-off tool, or for $5, most hobby shops also sell small tubing cutters (easier to use than the Dremel). Buy metal tubing which fits easily but snuggly inside the 3/16" rubber grommet (3/16" or even 5/32" outside diameter tubing). The longer 3/4" #6 trough board mounting screws are also required, and are a standard hardware store item.

The rubber grommets and metal tubing which goes inside the grommets. Three grommets/tubes are needed for each of the two optic boards.

After buying the above parts (either from Williams or elsewhere), drill or use a hand remer and make the trough optic board holes bigger, about 1/4" (or up to 7/16", depending on the outside size of your rubber grommets). Be careful you don't drill through any board traces (this should not be a problem). Also, drilling the center hole is optional; mounting with just two (outside) of the three holes per board works fine too. Insert the rubber grommets in the enlarged holes, and put the metal tubing inside the grommet. The metal bushing should be just very slightly shorter than the width of the rubber grommet, no longer. This allows the board bolt to tighten down on the bushing, but leaving a bit of "play" in the board (which the rubber grommet gives).

More Random Multiball: the Ball Trough Optic Resistors.
On Indy Jones, Star Trek Next Generation, Judge Dredd, Demo Man, and Popeye, the ball trough optic boards have several large blue resistors mounted to them. Since these boards get a fair amount of shock and vibration from balls, often these resistors can crack or break. If this happens, random (and continual) multiball can result. Check these large blue power resistors for breaks or cracks. Usually the resistor leads break right where they connect to the circuit board.

Do not try and repair the resistors; just replace them. They are 270 ohm 2 watt resistors (do not replace with a version less than 2 watts). These are available from Digikey, part number ALSR3J-270-ND, $1.37 each. NTE/ECG sell these too at many local electronic part houses for about 99 cents a pair.

Ball Trough Optos.
The ball trough optos also commonly break from ball vibration and wear. Every optic is a pair; a transmitter (which gives off infra-red light), and a receiver (or photo transistor, which sees the infra-red light). The receiver rarely goes bad. The transmitter optics are on the trough board closest to the coin door (lucky for us, as this board is easiest to access). The transmitter optic is available from Radio Shack, part number 276-143c, $1.69. This replacement optic transmitter is blue in color, and works fine as a replacement. Gregg Woodcock also sells yellow trough LED infrared transmitters at www.ClassicCoinOps.com/wmsoptos.htm, for a really nice price. In either case, this part should only be installed one way. Printed on the circuit board is a round circle with a flat side. The optic also has a flat side, which should match the circuit board.

The receiver optic is also available from Radio Shack, part number 276-145a, $0.99. This receiver is clear, unlike the Williams receiver. The flat edge of the receiver needs to be mounted closest to the top edge of the circuit board. That is, the flat edge goes in the hole furthest away from the hole that has the notch drawn on the circuit board. Digikey also sells a receiver, part number PN104-ND. When installing this photo transistor remove the center pin before installing. Just wiggled the center lead back and forth until it breaks off at the base. Install this part so the notch at the base lines up with the notch drawn on the circuit board.

The New Williams Ball Trough and the Blue Resistors.
If using the newer metal trough #A-16809-2, and using all three mounting holes, it will also be necessary to move one of the large blue resistors to the back of the board, and drill a new center position mounting hole in the opto board. Another option (and spending $50 is not an issue), order the upgrade kit from Williams, part# A-18244, and get the two new trough opto boards and the mounting hardware. Or use the existing trough boards with just the two outside mounting holes. If drilling the current trough boards is not an option, they can always be mounted with two of the three holes instead. This works fine too.

Bad Ball Trough Connectors.
Another ball trough problem can be related to the connectors used on the ball troughs. Again, due to vibration, the solder joints for the circuit board header pins can crack, causing intermittent connections. To fix this, reflow the solder on the connector pins on both trough boards.

Testing the Ball Trough Optos.
After modifying the trough boards and grinding the divots out of the trough, I connect the transmitter and receiver boards to their connectors. Now I dim the lights to the room, turn the game on, and go to the first switch test T.1. Using a Radio Shack or MCM infrared detector card (or a digital video or digital still camera), check all the transmitter LED infrared optos to see if they are working.

After that is done, shine a small pocket flashlight or TV remote control into each of the receiver board detector optos. They should register in the T.1 switch test (room needs to be somewhat dim for this; ambient room light can also activate these). Turn the game off and assembly and install the trough board on the trough, and install the trough back in the game.

Now it's time for another test, one that is especially good to verify your work, or to test the trough if you have not modified it. With all the balls removed from the game, turn the game on and go to the first switch edge test T.1. Most switches should show with a dot, indicating the switch as open (a sqaure indicates a switch is closed). But on optic switches, a blocked opto is a dot, and an unblocked opto is a square (opposite of what one would expect). There should be a number of squared switches, indicating the opto trough switches (check your game manual for exact switch numbers). If your switch matrix has no squares (all dots), your playfield has lost the +12 volts powering the optic switches. Check fuses F115 and F116 (F101 and F109 on WPC-95) on the power driver board.

Now slowly roll a ball down the trough and watch it cause a square in the switch matrix to turn into a dot, as the ball rolls past each ball trough optic. When the ball is resting at ball trough optic one, physically push up on the ball lane shooter solenoid (that would kick that ball onto the playfield). This will cause that "trough jam" opto to turn to a dot. This opto only sees the ball as it gets kicked out, or if there are two balls jammed so they are sitting on top of each other at the right end of the trough.

Fill up the trough completely with balls, then remove the balls manually, one by one. Try this a few times to see if you can isolate any of the ball trough squares which are not turning to dots consistently.

Lastly, remove ALL balls from the trough and close the coin door. Press the flipper buttons to activate the flippers while still in switch edges test. Look for flickering square-to-dots on the ball trough column on the display. This tests flipper vibrations which can cause intermittent flickering on the opto switches. Now continue checking for bad optos by hitting the playfield with the meat of your fist near the flippers (it's not as bad as it sounds!) If any of the squares flicker to a dot, there is some vibration related problem (broken/cracked blue resistor or opto lead, or cracked header pin solder joints). If nothing has appears, leave the game in this test mode for 20 minutes (note some games will exit test mode automatically after 15 minutes) with no balls in the game. Be close by, within listening distance. If you hear the game "bong" that means a switch has opened/closed in the switch test. Go to the game and check the score display, as the last switch closed will be reported. See if this is a trough opto switch number. If so, it is a flakey opto or bad opto board resistor or bad connector. This "time test" allows the game to 'warm up' too, which often the other tests don't account for.

If all the trough switches change from squares to dots when the optos are blocked with a ball, and there is no flickering when the playfield is vibrated, and the game doesn't report any random switches in test mode for 20 minutes, the opto boards have test good. If there are still random multi ball problems, there is most likely a divot problem in the ball trough (see above).

Here a ball trough transmitter opto board is being tested outside of the game using an external 12 volt DC power supply. There are seven infrared LEDs here, but the one with the red arrow is not lighting. Check for a bad blue power resistor, broken traces, or even a bad opto itself. Note the digital camera this picture was taken with shows the infrared light quite well. Pic by Tx.

Testing the Ball Trough Transmitter Board outside of the Game.
The ball trough transmitter board needs +12 volts DC to operate, and nothing more. Because of this, the ball trough transmitter board can be tested outside of the game using an external 12 volt DC power supply. Also needed is some way to "see" the infrared light coming from the transmitter LEDs. A digital camera with a viewing screen works well, or a Radio Shack/MCM Electronics infrared card.

Dot matrix displays are one of the coolest features on a WPC game. They provide the score and graphic animations, and even video games within the pinball game. Note that the first three WPC games (Funhouse, Harley Davidson, the Machine) used the older style AlphaNumeric displays.

WPC Alpha Numeric Score Display Problems.
The first three WPC games that used AlphaNumeric displays have a common problem. The resistors R48 and R49 (39k ohm) on the AlphaNumeric Display board often fail and go open, or go out of spec. This can cause all the score displays in the game to work very weak, or not work at all. Before replacing a score display, replace BOTH of these 39k resistors with "flame proof" 1 or 2 watt versions. See the Williams System 11 repair guide at http://marvin3m.com/sys11/index3.htm for more information on repairing AlphaNumeric score displays. All the information there applies to these three WPC games (though the component label numbers will be different).

A dot matrix display on the way out. Notice the absence of some characters in the display (on the right side).

Dot Matrix Displays and "Outgassing".
The unfortunate part about dot matrix displays (DMD) is they wear out. Time will eventually kill these, and the display will "outgas" and fail. Because of the high voltage involved with score displays, the anode and/or cathode inside the diplay glass breaks down. This results in the "outgassing" of impurities that eventually change the internal gas properties, so the display won't glow (the gas must be very pure for the display to work). Often the gaps 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 will solve this problem, and is easy to get and replace (a 5 minute job). These do cost a bit of money though at about $115 each (complete). There is no way to fix an old "outgassed" display.

When a DMD starts to get blurry or displays gaps, the rumor is the power requirements for the display increases. It turns out this rumor is actually incorrect, at least as far as the High Voltage (-120 and +65 volts) is concerned. The HV (high voltage) power used by a display is directly proportional to the number of dots lit on the display. If a display is entirely outgassed and not lighting (even though the CPU is asking the display to lit), it will consume no more HV (high voltage) than a working display that is not lit. Kirb did some test of various displays and metered the results, proving this.

But what about the 5 volt consumption? Unfortunately we did not do enough testing of the 5 volts to draw any conclusions. But based on reports of outgassed displays causing game resets (stressing the 5 volt supply), it is reasonable to think that an outgassed DMD does consume more 5 volt power. Another interesting fact is that certain DMD makes consume more 5 volt power than others. The biggest 5 volt power hog is Dale/Visay, consuming nearly twice what other DMD displays use.

Regardless, I still encourage people to buy a new display if theirs is outgassed. The 5 volt power stress, particularly on games like Twilight Zone, can cause potential game reset problems.

Buy an entire DMD display glass and board, or just a new Glass?
A new dot matrix glass only can be purchased, which will also solve the "outgassed" problem. These are available for about $65, which is almost half the price of buying both the display and its attached circuit board. But trust me on this, don't be cheap; just spend the extra money and get both the display and its attached circuit board. Installing a new glass into the surrounding board is A LOT of work. And games produced in 1993 and later don't have "pin" style glasses, so these display glasses alone are NOT replacable. Even if a display has the "pin" style glass, it's just not worth the trouble to unsolder 132+32 pins, install the new glass, and resolder all those pins again. It's a solid two hours worth of eye straining work, and it's very easy to make a mistake. It's just not worth the trouble.

Are All Dot Matrix Displays the Same?
The short answer is "yes". But be aware DMDs come in different sizes. Williams always used the 128x32 column/row variety (DataEast for example used a 128x16 and a 192x64 display, in addition to 128x32). And yes a 128x32 dot matrix display from a Gottlieb, Sega, DataEast or Stern game will work in any DMD WPC/WPC-S/WPC-95 game or vice-versa (but note that DataEast/Sega/Stern have an additional controller board bolted to the back of their 128x32 DMD, which is not used on a Williams WPC game). Also it should be stated that some brands of dot matrix displays (like Babcock) require 12 volts to operate, and most others don't. I have seen problems where a DMD requiring 12 volts won't operate in a game, but one that does not require 12 volts will work.

Can the Dot Matrix Display Itself be Fixed?
This is a tricky question. Sometimes the display itself fails due to problems other than an "outgassed" score glass. The controller chips on the display glass' circuit board can die (they are static sensitive). This usually causes "garbage" to be displayed. Other problems I have seen includes delamination of the surface mounted parts on the score display glass' circuit board (often this is fixable). And the power .156" header pins on the display itself can have cracked solder joints, causing the display to not work (though sometimes these are nearly impossible to resolder, because the display glass is in the way!)

Example a ribbon cable problem on a WPC game (Demo Man). Can you tell it says, "Game Over"? Reseating the ribbon cables often fixes this. Click on the picture below for a larger version, and note the dark spots in the corners of this display - this is an indication the display is starting to outgas. Note it's not just the display ribbon cables, but also the other ribbon cables like the one between the CPU and driver boards.

Another example of DMD garbage that was fixed by reseating the ribbon cable between the driver and CPU boards. Picture by Wil.

Blank, Strange Garbage, or Diagonal Lines on the Dot Matrix Display
(Re-seating Ribbon Cable connectors).
This problem can be caused by a bad dot matrix ribbon cable. A blank display (assuming all the fuses are good and voltages are present) is usually a backwards installed ribbon cable from the dot matrix controller to the DMD itself. Garbage or diagonal lines is typically a problem with the large cable running from the CPU board to the fliptronics board to the sound board to the dot matrix controller board. The ribbon cable connectors are gold plated, and sometimes require a "reseating" (remove and re-install) of their connectors to "clean" them. Since these are gold plated connectors, reseating is an acceptable means of cleaning a gold plated connector. (All the non-ribbon cable connectors in the game are *not* gold, and if reseating "fixes" a problem, that means the connector board pins and housing pins need to be replaced! See Pinball Connector web page for more info on that.)

Example a dirty or removed ribbon cable from the dot matrix controller board to the dot matrix display itself. Reseating the ribbon cables often fixes this.

When reseating the ribbon cables, be careful not to re-insert the ribbon cable one pin off. This is very easy to do, making pins 1,2 hang off the side of the mail connector (or cable pins 1,2 connected to board pins 3,4). This will cause additional problems like garbage display (but luckly all are fixed with the proper reseating of the ribbon cable connector). Also note the red line on the ribbon cable - this indicates pin 1 of the cable, and it should align with the white arrow or "1 2" silkscreened on the circuit board. Luckily the only ribbon cable connector that can be easily installed "backwards" is the ribbon going from the dot matrix controller board to the display. If this cable is installed "backwards", usually the display is blank, showing nothing (like the display does not work).

Here's what happens if the sound board ribbon cable is connected one row of pins off-center.

Also be aware that an over-zealous previous owner may have runined a ribbon cable connector when it was reseated. It is very easy to rip the ribbon cable away from the connector, making the game do some very strange things (usually the diagonal lines are a symptom of this). So be careful when reseating ribbon cable connectors.

Finally, random vertical or diagnal lines could be caused by 12 volts not getting to the dot matrix display. This voltage comes directly from the driver board (see "Testing DMD voltages" below for diagnosing this problem further). Also some dot matrix displays (Babcock in particular) require 12 volts to operate, where other brands do not need 12 volts.

Missing Vertical or Horizontal Display Lines are Missing.
Another common problem is missing display lines in the DMD score display. This is very common with the "pin" style DMD display glass. This type of DMD glass has pins, bent at a right angle, that solder into the attached DMD circuit board. Often these pins break, due to vibration, right where they attach to the display glass' edge. Because of this problem, all the DMD manufacturers have changed to a very flat ribbon cable style of connection between the display glass and the attached circuit board. This largely solved the problem.

If missing some lines, and the score display glass is a "pin" style, often the pins can be reattached to the display glass using a conductive silver epoxy. This often works well, but is a difficult repair. It usually does not work if more than two horizontal and/or two vertical pins are broken.

Diagnosing Other Dot Matrix Problems.
If you are sure the display itself is working, there are some other things to check when a DMD doesn't work.

Make sure to check fuses F601 and F602 (all WPC games). F601 is used for +62 volts, and F602 is used for -113, -125 volts (or -103, -115). On WPC-S and before, these are 3/8 amp fast-blo 1.25" fuses (originally Williams used slow-blo fuses here, but about 1994 they changed to fast-blo, so either fast or slow-blo can be used). On WPC-95, these are T0.315 amp 5x20mm fuses.

The Dot Matrix Display circuit is the same in all WPC generations!
Even though there are three different WPC dot matrix controller boards, the DMD voltage circuit is nearly identical. Click here for the high voltage dot matrix display controller board schematics (showing part references for all generations of WPC dot matrix display controller boards).

It's easier to test voltages at the dot matrix display itself than at the controller board. Use the "key" pin for reference to figure out which is pin 1 and pin 8.

Testing DMD Voltages.
If the fuses are good on the dot matrix controller board (or audio/visual board for WPC-95), you should next check the power at the DMD itself. Voltages used are +62, +12, +5, -113 and -125 (or -103 and -115), or within +/- 10% of these values. Check these voltages at the dot matrix display with the display connected, or at connector J604 on the controller board. The pin out at the DMD is:
• Pin 1: -125 volts (-110 to -130 volts); Williams lowered this voltage to -115.
• Pin 2: -113 volts (-98 to -118 volts); Williams lowered this voltage to -103.
• Pin 3: Key
• Pin 4: Ground
• Pin 5: Ground
• Pin 6: +5 volts (4.9 to 5.2 volts)
• Pin 7: +12 volts (10 to 14 volts)
• Pin 8: +62 volts (58 to 68 volts)
All voltages should be pretty much right at the above specs, or within +/- 10%. In regards to the -113, -125 volts (or -103, -115), these two voltages need to be 12 volts apart - that's the important part. That is, if -98 and -110 volts are measured, those two voltages are fine. If -118 volts is measured, the other voltage should be -106 volts. If they are not within 12 volts of each other, the dot matrix controller's high voltage section probably needs to be rebuilt. If the -125 volts is missing, -113 volts will be missing too. If the +62 volts is above 70 volts, chances are good someone jacked up this voltage by changing the DMD controller 1N4759 zener diode to compensate for an outgassed dot matrix display (very common on games imported back to North America from other countries).

If any voltage is low, try disconnecting the power connector to the DMD, and re-measure the voltages. If they return to the correct voltages, the display is bad or the high voltage section on the dot matrix controller board is failing and can't handle the power draw of the display.

Remember the voltages created by the DMD controller card are -125, -113 (or -115, -103) and +62. The +5 and +12 volts come from the driver board. If the 5 volts is missing yet the game boots, there's a connector problem. If 12 volts is missing there's either a connector problem, or the dot matrix display itself is "sinking" the 12 volts (disconnect the DMD power connector and see if the 12 volts comes back up, if so the display is bad or maybe the driver board 12 volt section is failing). Or the 12 volt driver board section is failing. (Measure the 12 volts at the driver board, and then at the installed DMD, if the voltage is different there is a connector problem. If they are both the same voltage and are below 10 volts, there is a driver board 12 volt problem).

Lowering the -125 and -113 voltages to -115 and -103 volts.
At some point Williams lowered the -125 and -113 voltages to -115 and -103. This was done to increase the life of the score display. Just keep this in mind when measuring these voltages. The important part is these two voltage must be 12 volts apart.

Both the -125 and the -113 volts are the same voltage.
The dot matrix display will not work if both the -125 volts and -113 volts (or -115 and -103) measure as the same voltage. These two negative high voltages should be 12 volts apart. The difference in voltage occurs because of diode D6 (D3 on WPC-95), a 12 volt 1N4742 diode. The failure of this diode also kills transistor Q7 (known as Q7 in all WPC generations, a MJE15030). Also check resistor R8 (4.7k ohms 5 watts), if this is bad the two negative voltages will be the same. The -125 volts and -113 volts must be 12 volts apart, or the dot matrix display will not work!

The +62 volts drops to +12 volts under load.
When this happens, check transistor Q3 (all WPC generations). This transistor has probably shorted. Also check diode D3.

The +62 volts is not +62 volts.
On WPC-S and earlier games, the positive DC voltage trace that comes from a very small bridge rectifier BR1 is physically routed underneath resistor R9 (1.8k 5 watt resistor). Because of the heat generated by this 5 watt resistor, and the current drawn from the bridge rectifier, this circuit board trace can become burnt and break underneath resistor R9. Because the trace physically runs under this resistor, the broken trace can be hard to see. If the +62 volts is not +62 volts, check this trace. If the +62 volts is above 70 volts, chances are good someone jacked up this voltage by changing a DMD controller 1N4759 zener diode to compensate for an outgassed dot matrix display (very common on games imported back to North America from other countries).

The -125 volts is too High.
Another problem is the -125 volts (or -115) is too high, reading instead -140 volts. The usual cause of this problem is a broken trace on the circuit board. These traces are fragile, and the high voltage section of the dot matrix controller can get very hot, and burn them. Use your DMM set to continuity and check all traces.

Negative High Voltage Low, DMD barely lights.
Negative high voltage reads -102 and -93 volts, and the display barely lights. DMD high voltage controller section was just rebuilt, so that was ruled out. Checked resistor R6 or R26 on WPC95 (47K ohms) and it was open. Also checked resistor R4 or R30 on WPC95 (120 ohms) and it read 1k ohms (had to unsolder and lift one leg to test them). After resistors replaced, high voltage went up to -112 and -100 volts, and the DMD was nice and bright.

Rebuilding the Dot Matrix High Voltage (HV) Section.
If the fuses are good, and the display itself is good (tested in another game), it is time to rebuild the high voltage section of the Dot matrix controller board. But before doing that, raise the playfield and inspect all the connections from the transformer in the bottom of the cabinet. Though a rare problem, one of the connectors may have come apart or became oxidized.

After all else is checked, the best idea is to just replace everything in the high voltage section (parts also listed at dmdhv.htm). Note all these parts are also available in kit form from Great Plains Electronics for around $6 per kit. This is a *very* economical way to rebuild the dot matrix high voltage section. The parts to replace includes:

• Q6 (MJE15031 or NTE55): Controls the -125 volts (and supplies voltage to the -113 volts).
• Q7 (MJE15030 or NTE54/BUV27/BUV28): Controls the -113 volts.
• Q3 (Q1 on WPC-95, MJE15030 or NTE54/BUV27/BUV28): Part of the +62 volt section.
• Q4,Q5 (MPSD52 or 2N5401/NTE288): Part of the -125 (or -115) volt section.
• Q2,Q10 (Q2,Q3 on WPC-95, MPSD02 or 2N5551/NTE194): Part of the +62 volt section.
• D4,D5 (D1,D18 on WPC-95, 1N4758 or NTE5090, 56 volts): Part of the -125 (or -115) volt section.
• D6 (D3 on WPC-95, 1N4742 or NTE142, 12 volts): Part of the -113 (or -103) volt section.
• D3 (D2 on WPC-95, 1N4759 or NTE149, 62 volts): Part of the +62 volt section.
• Q1 (2N3904, WPC-S and prior only).
• R4,R5 (120 ohm 1/2 watt). Usually Ok, but replace if they look burned.

Check/Replace the Resistors too.
Also check the resistor values. Resistors either work or do not work, and are easily tested (unlike the above transistors). All resistors should be within 10% of spec. Replace any resistors that are out of tolerence or that appear burnt. The 5 watt resistors take the most abuse; if these are working yet cracked, replace them! Always mount resistors slightly above the board to allow air flow below them. On all these resistor, replace if they look at all damaged, even if they measure OK.

• 1.8k ohms, 5 watts: R9 on WPC-S and prior (R44 on WPC-95).
• 4.7k ohms, 5 watts: R8 on WPC-S and prior (R43 on WPC-95).
• 120 ohm, 5 watts: R11 on WPC-S and prior (R28 on WPC-95).
• 120 ohm 1/2 watt resistors at R4, R5 WPC-S and prior (R30, R31 on WPC-95).
• 47k ohms 1/2 watt at R3, R6, R12, R13 on WPC-S and prior (R25, R26 R27, R29).

An Alternative to Rebuilding the HV Section.
If the inexpensive HV rebuild kit from Ed at www.greatplainselectronics.com is beyond one's technical skills, there is an alternative to rebuilding the high voltage section. That is to purchase a pre-fabricated board which essentially does the same thing. The DMD-HVP (dot matrix display-high voltage power) board is available from www.pinball-parts.com for about $60. This plugs into and overlays the existing DMD controller board, replacing the original high voltage section on the original DMD controller board. Installs in about five minutes with no soldering. If the original high voltage section is blown on the original DMD controller board, it does not matter (as this completely replaces it). A good alternative for those that have more money than time, or limited soldering skills. Only works on pre-WPC95 games though.

I have some minor critisms with the DMD HV board though. For example, they use the smaller WPC-95 style fuses. Now this would be Ok if the board worked on WPC-95 games. But since it does not, it puts a mix of fuse sizes into a WPC game that otherwise don't use this smaller fuse size. This is bad for the end consumer that may have a supply of stock WPC HV fuses, which now won't work in their game! Also, I feel there should be LEDs for each of the high voltages to show at a glace that -125 volts, -113 volts, +62 volts (and perhaps the +12 volts and +5 volts) were working on the board.

DMD Components by Voltage.
Here are the same list of components, organized by voltage. If only a particular voltage is missing from your DMD, only these selective components can be replaced (not recommended):

• -125 volts: MJE15031 transistor Q6 (all WPC versions). MPSD52 transistors Q4, Q5 (all WPC versions). 1N4758 diodes D4, D5 (D1 and D18 on WPC-95). All these components supply voltage to the -113 volt section too. Hence, replace the -113 volt components too.
• -113 volts: MJE15030 transistor Q7 (all WPC versions). 1N4742 diode D6 (D3 on WPC-95), which drops the -125 volts down to -113 volts.
• +62 volts: MJE15030 transistor Q3 (Q1 on WPC-95). MPSD02 transistors Q2, Q10 (Q2, Q3 on WPC-95). 1N4759 diode D3 (D2 on WPC-95).

The BIGGEST Tip when Fixing the High Voltage.
The single biggest tip when fixing the high voltage section on the DMD controller is this: REPLACE EVERYTHING. This is a high voltage section. This means if all parts were replaced except for ONE bad part, this bad part can cause all the others just replaced to immediately fail! It's just not worth the trouble. Rebuild the whole high voltage section, and replace everything. In the long run money and time will be saved.

Example of a "cloudy" dot matrix display.

Cloudy Dot Matrix Display.
Cloudy display problems are strange. The display can test perfectly in the internal "line" dot matrix test. But when large areas or inverted graphics are shown, the display is "cloudy". This is usually caused by heat related problems. Fixing this could be as simple as adding new white heat sink compound to the three heat sinked MJE transistors. Also make sure they are tight to their heat sink. Check the three large 5 watt resistors too. If they are more than 5% out of spec, replace them (see above). Lastly, cold solder joints in the high voltage section can also cause cloudiness. Try reflowing the solder joints on the 5 watt resistors, the high voltage diodes, and the high voltage MJE transistors. If none of this works, rebuilding the high voltage section should solve this problem (see above).

Wavy Hum-bar, bounce, or Horizontal Roll on the Dot Matrix Display.
The "wavy hum-bar", graphic "bounce, or horizontal roll seen on the dot matrix display's images can be bad DMD power filter capacitors. On WPC-95, these are caps C28, C42 on the audio visual board. On WPC-S and earlier, these are caps C4, C7 on the dot matrix controller board. These original capacitors were 150 mfd 160 volts. This value is somewhat hard to find, but can be replaced with the more common 220 mfd 160 volt electrolytic caps (remember going up in value on electrolytic capacitor's voltage and/or capacitance is Ok, but never go down). If 220 mfd caps are used instead of the 150 mfd, don't get ones that are too large (due to their weight, vibration can crack the capicator's solder pads, essentially removing those new capacitors from the circuit!)

Additionally, if there is still a "wavy hum-bar" or horizontal roll or a display "bounce", try replacing the smaller high voltage filter capacitors. On WPC-S and earlier, these are capacitors C6, C9 and C10 (.1 mfd 500 volts) on the dot matrix controller board. On WPC-95, these are caps C29-C31 (.01 mfd 200 volts). If these caps fail, hum bars or roll can occur. As the game warms up the wave, roll or bounce may change (get better or worse).

Crystallized Solder Joints.
If a DMD display is not displaying correctly, and the voltages seem Ok, also check this. It's common for the solder joints on the zener diodes in the power section to crystallize, causing heat damage, excessive resistance, and finally a lost of voltage regulation. This can then lead to a failed DMD and damaged power circuits. These diodes are D3, D4, D5, D6 (D1, D2, D3, D18 on WPC-95) on the dot matrix controller board.

A bad 6264 RAM chip on the DMD controller board can cause this problem (verify it's not the Dot matrix display itself first though!)

DMD Columns Stuck "On".
If there is a column or two stuck on (as seen in the picture above), chances are good the 6264 dot matrix controller card RAM at U24 (WPC-S and prior) has failed. Of course this assumes that the dot matrix display itself is not the problem (try the display in another game to verify). If not the display itself, replace U24 (WPC-S and prior) with a new 6264 RAM chip, and this should fix the problem.

Missing Lines on a DMD Display.
The first generation of dot matrix displays used pins to connect the DMD glass to the DMD circuit board. Due to vibration, often these pins would break right where they meet the display glass. This would give the display a "missing" vertical or horizontal line (depending on which pin broke). And often more than one pin would break, making an otherwise good display nearly useless. This problem was solved with newer DMD score displays that used a short thin flexible ribbon cable instead of the pins.

On displays with broken pins, there isn't enough material to solder the pins back to the display glass. But another technique can be used instead. This involves "conductive epoxy", and essentially gluing the broken pin to the score glass. The conductive epoxy has silver powder in it, so it conducts well. And it's the only way to get a broken pin attached back to the score glass. Usually one or two broken pins can be repaired in this manner (trying to do much more than three seems to not work well!) Just be careful not to short two pins together with the epoxy. Success rate is certainly not 100%, but it usually works. The epoxy is expensive though, because of the silver powder in the glue.

I have also used conductive epoxy to fix the thin ribbon cable variety of DMD displays with missing lines, where the ribbon cable has ripped away from the display glass. The success rate is not as high, but it can work.

using conductive silver epoxy to fix a missing line on a dot matrix display, where the metal pin broke away from the edge of the display glass. Note this display uses both the ribbon cable (at the circuit board) and the metal pins (at the display glass). But the conductive epoxy can be used to repair either style (pins or ribbon), but the success ratio is higher on metal pins.

Problem: Dot Matrix Display Got Blurry. When I was playing my Twilight Zone, the dot matrix display started to become very blurry. Within 5 minutes the display became almost unreadable. The dots to the left and right of the active ones started to flicker.

Answer: the ASIC chip on the CPU board was not making good contact to its socket. The ASIC chip is the large square chip on the CPU board. After removing the chip and cleaning all of its pins, and reseating the chip in the socket, the problem went away. Another thing to try is reseating the board ribbon cables in their sockets.

Problem: Funhouse alphanumeric display, character 16 was mimicking every segment being displayed in the other 15 characters.

Answer: If this is happening in display one, replace chip U8 (6184 Anode Drive) on the WPC display driver board. If happening to display two, replace chip U5 (6184).

Problem: My Twilight Zone's dot matrix display shows random vertical lines. At first it was just occassionally during game play, but now they appear from the moment I power on the game. The problem has gotten worse, and now every time I turn on the machine, all four flippers energize.

Answer: the problem was a bad ribbon cable. There is a single ribbon cable that goes from the CPU board to the fliptronics board to the sound board to the dot matrix controller. If the ribbon cable was mis-installed by one pin, or the cable has torn at its connector, this problem can happen. The ribbon cable houses the address and data lines to the fliptronics, sound and dot matrix controller. Often the ribbon cable's connectors can just be dirty, so reseating the connectors sometimes fixes this problem. If the ribbon cable is damaged, mis-installed or the connectors are dirty, strange things like this can happen. Another potential cause could be the lack of 12 volts getting to the dot matrix display controller board.

CPU Board LED Flashes.
A simple diagnostic LED (Light Emitting Diode) flash pattern exists on all generations of WPC CPU boards. These flashes can signify a problem and what might be causing the trouble. They can be seen immediately when powering on the game. LED's exist on both the CPU and Driver boards, but only the CPU board's LED have a diagnostic flash pattern. On WPC-S and earlier CPU boards, the LED's are labeled D19 to D21. On the driver board and all WPC-95 boards, they are labeled "LEDx" (with "x" being the LED number).

CPU Board LED Flash Codes, all revisions.
WPC-S and prior uses a "Dx" designation for its CPU LEDs. WPC-95 uses a "LED20x" designation.

• D19/LED201 (blanking): at power-on should be ON for about 3 seconds (1 second on WPC-95), and then turn off and stay off. When D19/LED201 is on, the blanking circuit is disabled (and will not allow any coils to be energized).
• D20/LED203 (diagnostic): After D19/LED201 turns off, D20/LED203 should stay flashing permanently while the game is turned on. This indicates the CPU is "running".
• D21/LED202 (+5vdc): this LED should ALWAYS be on. It indicates the CPU has +5 volts DC power.
  Problem Power-On CPU D20/LED203 (diagnostic) Flash Codes. If D20 does not flash continually, here are the flash codes diagnostics:
  • blinks ONE time: U6/G11 CPU game ROM bad
  • blinks TWO times: U8 CMOS RAM chip bad
  • blinks THREE times: U9 WPC custom chip bad (pre WPC-S), or G10 Security PIC chip bad (WPC-S and later)

WPC-S and Prior Driver Board LEDs, Test Points (TP), and Fuses.
For reference, TP5 is ground.

• LED1/TP3: +12 volts DC switch matrix circuit. Should be always ON. If off, check fuse F115. This is often caused by a bad CPU board chip U20 (see the switch matrix section for more details). The AC Power originates at connector J101 pins 4,5 and 6,7. It then goes through fuse F114, bridge BR1, capacitors C6 and C7, LED6/TP8 (18 volts DC), diodes D1 and D2, voltage rectifier Q2, fuse F115, LED1/TP3 (12 volts DC), then to connector J114 pins 1,2. Also, just before diodes D1 and D2, the circuit splits to the LM339 chip U6, and LED2/LED3.
• LED4/TP2: +5 volts DC digital circuit. Should be always ON. If off, game will not boot. Check fuse F113 (or bridge BR2 and capacitor C5). Though not likely to fail, there is also a voltage regulator LM323 at Q1, a LM339 chip at U6 ("zero cross"), and two 1N4004 diodes at D3 and D38. The AC Power originates at connector J101 pins 1 and 2. It then goes through fuse F113, bridge BR2, capacitor C5, voltage rectifier Q1, LED4/TP2 (5 volts DC), then to connector J114 pins 3,4. Note after fuse F113, the AC power also continues to diodes D3 and D38, and to LM339 chip U6. Then this "zero cross" power merges back into the +5 volt line before hitting connector J114.
• LED5/TP7: +20 volts DC flashlamp circuit. Normally ON. Twilight Zone and later, this LED fades off when the coin door is opened. If off, check coin door and fuse F111 (or bridge BR4 and capacitor C11). The AC Power originates at connector J102 pins 1,2 and 3,4. It then goes through fuse F111, bridge BR4, capacitor C11, LED5/TP7 (20 volts DC), then to connector J107 pins 5,6 (and J106 and J108).
• LED6/TP8: +18 volts DC lamp matrix circuit. Normally ON. If off, check fuse F114 (or bridge BR1 and capacitors C6, C7). Though not likely to fail, there is also a voltage regulator LM7812 at Q2, a LM339 chip at U6, and two 1N4004 diodes at D1 and D2. The AC Power originates at connector J101 pins 4,5 and 6,7. It then goes through fuse F114, bridge BR1, capacitors C6 and C7, LED6/TP8 (18 volts DC), diodes D1 and D2, voltage rectifier Q2, fuse F115, LED1/TP3 (12 volts DC), then to connector J114 pins 1,2. Also, just before diodes D1 and D2, the circuit splits to the LM339 chip U6, and LED2/LED3.
• LED7/TP1: +12 volts DC power circuit (motors, relays, etc). Should always be ON. If off, check fuse F116 (or bridge BR5 and capacitor C30). The AC Power originates at connector J112 pins 1,2 and 3,5. It then goes through fuse F116, bridge BR5, capacitor C30, LED7/TP1 (12 volts DC), then to connector J118/J117/J116 pin 2.
• TP6 (no LED): +50 volts for the coil. The AC Power originates at connector J102 pins 5,6 and 8,9. It then goes through fuse F112, bridge BR3, capacitor C8, TP6 (50-70 volts DC), then fuses F103/F104/F105 (and F102/F102), then to connector J107, J106 J108, and J109.
• LED2 (no TP): This LED is not always installed. High/low line voltage sensor. Normally ON, but flickers with the playfield lamps.
• LED3 (no TP): This LED is not always installed. High/low line voltage sensor. Normally OFF, but flickers with the playfield lamps.

WPC-95 Driver Board LEDs, Test Points (TP), and Fuses.
For reference, TP107 is ground.

• LED100/TP100: +12 volts DC regulated. Should be always ON. If off, check fuses F101 and F106 (or diodes D11-D14 and capacitors C11, C12). If fuse F101 has failed, this is often caused by a bad CPU board chip U20 (see the switch matrix section for more details). Though not likely to fail, there is also a voltage regulator LM7812 at Q2, and two 1N4004 diodes at D1 and D2. If fuse F101 has failed, suspect the voltage regulator Q2. The AC Power originates at connector J129 pins 6,7 and 4,5. It then goes through fuse F106, diodes D11-D14, capacitors C12,C11, LED102/TP102 (18 volts DC), diodes D1-D2, voltage rectifier Q2, fuse F101, LED100/TP100 (12 volts DC), then to connector J101 pins 1,2.
• LED101/TP101: +5 volts DC digital. Should be always ON. If off, game will not boot. Check fuse F105 (or diodes D7-D10 and capacitor C9). Though not likely to fail, there is also a voltage regulator LM317 at Q1, a LM339 chip at U1, and two 1N4004 diodes at D23 and D24. The AC Power originates at connector J129 pins 1 and 2. It then goes through fuse F105, diodes D7-D10, capacitor C9, voltage rectifier Q1, LED101/TP101 (5 volts DC), then to connectors J101 pins 3 and 4, J138 pin 4, J139 pin 4, J140 pin4, J141 pin 4.
• LED102/TP102: +18 volts DC lamps. Normally ON (can flicker with playfield lamps). If off, check fuse F106 (or diodes D11-D14 and capacitors C11, C12). The AC Power originates at connector J129 pins 6,7 and 4,5. It then goes through fuse F106, diodes D11-D14, capacitors C12,C11, LED102/TP102 (18 volts DC), diodes D1-D2, voltage rectifier Q2, fuse F101, LED100/TP100 (12 volts DC), then to connector J101 pins 1,2.
• LED103/TP103: +12 volts DC un-regulated. Should be always ON. If off, check fuse F109 (or diodes D3-D6 and capacitor C8). The AC Power originates at connector J127 pins 1,2 and 3,5. It then goes through fuse F109, diodes D3-D6, capacitors C8, LED103/TP103 (12 volts DC), then to connectors J138 pin 2, J139 pin 2, J140 pin 2, J141 pin 2.
• LED104/TP104: +20 volts DC flashlamps. Normally ON. This LED fades off when the coin door is opened. If off, check coin door and fuse F107 (or diodes D15-D18 and capacitor C10). The AC Power originates at connector J128 pins 1,2 and 3,4. It then goes through fuse F107, diodes D15-D18, capacitors C10, LED104/TP104 (20 volts DC), then to connectors J133 pin 5 and 6, J134 pin 5.
• LED105/TP105: +50 volts DC coils. Normally ON. This LED fades off when the coin door is opened. If off, check coin door and fuse F108 (or diodes D19-D22 and capacitor C22). The AC Power originates at connector J128 pins 8,9 and 5,6. It then goes through fuse F108, diodes D19-D22, capacitors C22, LED105/TP105 (50-70 volts DC), fuses F102, F103, F104, then to connectors J134 pins 1,2,3, J135 pins 1,2,3.

Sound Board Error Beeps pre WPC-DCS (WPC alpha-numeric, WPC dot-matrix and WPC fliptronics.

• 1 Beep: Sound board OK
• 2 Beeps: U9 sound ROM failure
• 3 Beeps: U18 sound ROM failure
• 4 Beeps: U15 sound ROM failure
• 5 Beeps: U14 sound ROM failure

Sound Board Error Beeps WPC-DCS and WPC-S.

• 1 Beep: Sound board OK
• 2 Beeps: U2 sound ROM failure
• 3 Beeps: U3 sound ROM failure
• 4 Beeps: U4 sound ROM failure
• 5 Beeps: U5 sound ROM failure
• 6 Beeps: U6 sound ROM failure
• 7 Beeps: U7 sound ROM failure
• 8 Beeps: U8 sound ROM failure
• 9 Beeps: U9 sound ROM failure

WPC-95 Audio/Video LED.

• LED501: +5 volts DC, normally FLASHING (but at a slower rate than CPU LED203).
  Problem Power-On Audio/Visual Board Beep Error Codes:
  • 1 Beep: Audio/Visual board OK
  • 2 Beeps: S2 sound ROM failure
  • 3 Beeps: S3 sound ROM failure
  • 4 Beeps: S4 sound ROM failure
  • 5 Beeps: S5 sound ROM failure
  • 6 Beeps: S6 sound ROM failure
  • 7 Beeps: S7 sound ROM failure
  • 10 Beeps: Audio/Visual board's Static RAM bad

Often when you buy a used WPC game, upon power up, you'll get an error message stating, "Factory Settings Restored". This message indicates that the CPU RAM chip at location U8 on the CPU board has forgotten the game's bookkeeping and options settings.

Most often, this error occurs because the three "AA" batteries on the CPU board have died. These batteries should be replaced every year with good quality alkaline batteries (batteries are cheap, battery damage is expensive). The three batteries must keep at least +4 volts of power to the U8 RAM chip for it to remember. When power goes below +4 volts, memory reset can occur (and you get the "Factory Settings Restored" error message).

A bad battery holder. At first glace, this holder looks fine. But the two battery contact points on the left have corroded and fallen off. The contact on the right is the only one intact. These contact points are actually rivets, but corrosion will cause the face of the rivet to break as it goes through the fiber insulator, and the face of the rivet that contacts the battery falls off.

Changing Batteries.
If your game is working, and it's time to replace the batteries, follow this procedure:
• Remove the backglass and gain access to the CPU board.
• Turn the game ON.
• Note the orientation of the installed batteries (All positive terminals up, or to the right on WPC-S).
• Remove the old batteries and discard.
• Check the battery holder's terminals for any corrosion (they can be clean with 220 grit sandpaper if any corrosion). If damaged, turn game off and replace battery holder.
• Using a Sharpie pen, write today's date on the new batteries.
• Install the new batteries.
• Turn the game off.

If you install new batteries with the game turned on, the machine will not forget the old option settings or bookkeeping totals.

More on Installing Batteries and Measuring their Voltage.
On all flavors of WPC (except WPC-S), the batteries install with the positive terminal (the terminal with the "tit") UP. On WPC-S, batteries install with the positive terminals to the right. To not lose the game's memory and firmware settings, new batteries can be installed with the game powered ON (assuming the old batteries are removed with the game on too). After the new batteries are installed, turn the game off. Now measure the voltage with a DMM to make sure then are connecting to the battery holder properly. Put the black lead of the DMM on the lower left battery holder solder point (or on WPC-S the upper left), and the red lead on the upper right battery holder solder point (or on WPC-S the lower right). About 4.5 to 4.8 volts DC should be seen.

The Battery Holder: a Weak Link.
If after replacing the batteries, you still get a "Factory Setting Restored" error when turning the game on, suspect the battery holder. Use your DMM and check the battery voltage at the CPU board. With the game off, put your DMM on DC volts and put the black lead on ground (the grounding strap or on one of the screws holding the CPU board in place, or the bottom left battery terminal). Put the red lead on each of the CPU board's POSITIVE battery terminal solder points (positive is the "up" side of each battery). Test each of the three batteries' positive leads individually, starting at the left. You should get about 1.5, 3.0, 4.5 volts at each battery (note the batteries are additive and the first battery in the chain will give you 1.5 volts, and the last battery will give you 4.5 volts). If you don't these positive voltages, suspect damaged battery holder terminals. These corrode quite often if new batteries aren't installed religiously. Replace the battery holder and re-test to ensure proper repair.

A battery gone bad on a WPC game. Note the white "fur" on the bottom of the battery, and how it has corroded the chip and socket below it. The battery holder, chip and socket must all be replaced. Also the board must be washed with a mixture of 50/50 water and white vinegar (a mild acid) to neutralize the alkaline battery, and then rised with water. After drying, the corroded areas are sanded clean to the bare copper traces, and the components replaced. If the board isn't washed with this vinegar solution, the corrosion will return.

The best battery holder to buy for any WPC game is the new black plastic battery holder used in WPC-S and later games. This is Williams part# A-15814. This design of battery holder is much better than the pre WPC-S design.

Remote Battery Holder.
Another excellent solution to potential battery corrosion problems is to install a remote battery holder. This way if the batteries do fail and leak, the damage is limited to a $1 battery holder. The cost of replacing an entire CPU board because batteries have leaked and corroded the board is too big of a risk for me personally. Though it doesn't happen a lot, if you have ever had to fix battery corrosion, it's a lesson you will not soon forget. Because of this I have installed remote battery holders in all my WPC games. They cost less than an original style WPC battery holder, and it's good insurance.

I personally use a four "AA" battery holder, using the fourth battery cell area for a back-up blocking diode and as the screw area (to do this I put a 1N4004 or 1N5817 diode with the band towards the red wire, where the fourth battery would be located). The four AA battery packs seem to be easier and cheaper to find, but of course only use three batteries! Install three "AA" batteries, and then solder the red positive wire of the remote holder to the CPU board's main positive battery holder trace, and the black lead of the remote holder to the CPU board's opposite main negative battery holder trace (see pictures below for installation in WPC-89, WPC-S, and WPC-95). On WPC-89 and WPC-95, the main positive battery terminal is at the upper right of the original battery holder, and the main negative terminal is at the lower left. On WPC-S the main positive battery terminal is at the lower right, and the main negative at the upper left.

Some people ask why I put the "blocking diode" in my 4 "AA" battery holder? Well it is not required, but I put the blocking diode in as a backup diode (which prevents the CPU board from trying to charge the AA batteries when the game is on). The other advantage to this is the added blocking diode slightly decreases the voltage from the batteries. This is like an advance alarm clock for me, where the game will tell me when the batteries are getting low (opposed to them totally dying and leaking, and then I find out I need to replace them!) Using a common 1N4004 diode will give the most voltage drop (about .4 volt). This decreases the battery voltage just enough that the game will give me a "factory settings restored" error just before the batteries are totally dead - which is exactly what I want!

Using an inexpensive four AA battery holder, a IN4004 or 1N5817 blocking diode, & three AA batteries as a remote battery holder for the CPU board. On WPC games the diode is not required, and a wire can be used instead. A 1N5817 diode is used instead of a 1N914 or 1N4001 because of the forward voltage drop is less with the 1N5817 diode. But I actually prefer a 1N4004 as an "auto alarm" when the batteries are getting low.

Remote battery holder installed in WPC-89 game.

Remote battery holder installed in WPC-S game.

Remote battery holder installed in WPC-95 game.

Is Power getting Past the Battery Holder? (bad diode D2 or RAM U8)
If the battery holder is OK, next check to see if power it getting past the battery holder. Find CPU board diode D2 (all WPC revisions); this is a small glass diode, right next to diode D1. On WPC-S and prior, look to the right of the big square chip U9. On WPC-95, look just below the battery holder. With your game off and new batteries installed, put your DDM on DC volts and put the black lead on the backbox ground strap. Then put the red lead on diode D2 on the CPU board. The banded side of the diode should show about .5 volts less than the non-banded side (which should be about 4.3 volts). If only one side of the diode shows voltage, or both sides show the same voltage, this diode is bad. Diode D2 is a 1N4148 or 1N914 diode.

Next test for voltage at the CPU U8 RAM chip (all WPC revisions). With the game off, you should get about 4.3 volts DC at pins 26, 27 or 28 of chip U8. If you don't, the battery voltage is not getting to the U8 RAM chip, and the game will boot up with the "Factory Settings Restored" error. Note pin 28 of the 28 pin U8 chip is in the same position as pin 1 of the chip, but on the opposite row of pins. Pin 1 is designated with an impressed "dot" right on the top of the chip.

There can still be problems even if a new batteries are installed and all the voltages check out. If the game is still giving "Factory Setting Restored" or "Set Time and Date" errors, there may be a bad CPU U8 RAM chip. This does happen where a bad U8 RAM will suck the life out of new batteries, causing them to go dead in one to four weeks. But make sure to double check that battery holder. Even minor corrosion can cause this problem. The voltages may all check out, but the corrosion may be enough to limit CURRENT, and cause this problem. The U8 RAM chip is a 6264-L or 2064 RAM chip.

Batteries Die Too Quick.
Batteries in a WPC game usually last for years. If the batteries in a game die quickly (a few days or a few weeks), the D1 diode is probably bad. If the D1 diode has failed, the batteries are trying to power up the entire CPU board (instead of just the U8 RAM). This will drain the batteries quickly. Find diode D1 (all WPC revisions); this is a small glass diode, right next to diode D2. On WPC-S and prior, look to the right of the big square chip U9. On WPC-95, look just below the battery holder.

Also check and test diode D2. With your game off and new batteries installed, put your DDM on DC volts and put the black lead on the backbox ground strap. Then put the red lead on diode D2 on the CPU board. The banded side of the diode should show about .5 volts less than the non-banded side (which should be about 4.3 volts). If only one side of the diode shows voltage, or both sides show the same voltage, this diode is bad. Diode D2 is a 1N4148 or 1N914 diode.

Batteries are HOT!
Another problem can occur where the batteries get hot. So hot, they can melt the covering off of them! If this is not fixed, the batteries will surely leak, or even explode. This happens when the game tries to charge the batteries, while the power is on. The problem is usually diode D2 (1N4148).

Does the Battery Power Anything Else?
Actually yes it does! Besides the RAM chip at U8, the battery also supplies voltage to the large square ASIC (Application Specific Integrated Circuit) 84 pin chip at U9 in the PLCC (Plastic Leaded Chip Carrier) socket. Because the 8-bit 6809 microprocessor (the brain behind WPC, on the CPU board) is such a bad time keeper, the time information for the WPC clock is generated in the U9 ASIC chip. A DMM can be used to measure the battery voltage on the ASIC chip at pins 1,22,43, and 64. If 4 volts DC is not seen at these pins, suspect the PLCC socket for the square U9 ASIC chip. These delicate sockets can corrode easily from leaking batteries.

The WPC ASIC chip Pinout.

My Game's Time Clock is Slow!
There is an internal time clock that keeps the time and date for the WPC system. Within the game's adjustments, you can turn the clock display on, so it shows the time and date on the dot matrix display. On Twilight Zone, this internal time clock is used during attack mode to set the playfield clock. If you notice the WPC time clock running slow (losing time), or the game just won't remember the time (boot up error of "Set Time and Date"), the batteries are getting weak and need replaced. If you still have this problem with new batteries, suspect the battery holder's terminals. They may be corroded enough to cause resistance, and lower the voltage at CPU chip U8.

All William's WPC pinball games are very durable commercial devices. They are well protected against voltage surges from lightning storms. There are several lines of defense against voltage surges:
• Excellent grounding
• MOV (metal oxide varistor)
• Line fuse
• Power transformer (all voltage goes through a transformer)
• Bridge Rectifiers
If the power line to your WPC game is struck by lightning, usually this will take out the line fuse and the MOV. Damage beyond this is extremely rare. To repair your game, you will have to replace both the line fuse and the MOV.

The MOV lives inside the "power box".

The MOV is the green disc soldered across the lugs of the radio frequency interference filter.

The MOV (metal oxide varistor) is designed to have high resistance. But when its rated voltage is exceeded, it internally shorts. This immediately blows the line fuse and halts the power to the game, saving everything but the line fuse and the MOV itself. Smaller voltage surges are absorbed by the MOV without total destruction (though lots of small surges can eventually destroy a MOV and make it short).

The MOV is located inside the cabinet's metal power box, next to the coin box. If you need to replace it, here are the values needed:
• North America (115 volt power): 150 volt or 130 volt MOV.
• Europe (220/240 volt power): 275 volt MOV.

The rating is the voltage at which the MOV will short. Lower voltage ratings will provide more protection. But remember the power s