Reset issues on pinball machines

Often people will contact me about reset issues with Williams pinball machines, primarily associated with the WPC era from the early 1990’s. I presume they do a little searching around the internet and come to the conclusion this is a real common problem, solved by replacing BR2 (bridge rectifier) and C5 (filter capacitor).

What happens is that many people will attempt to shotgun these parts (shotgun means to replace without knowing if they are in fact defective).  Some of these people will have limited de-soldering experience, and end up damaging their Power/Driver board.

In my professional experience (30+ years), my opinion is that there is no common Williams reset problem.  Reset issues can occur in all solid state pinball machines and all brands, and it can be caused by many different things, most of them related to the power chain. When the voltage drops below a threshold, the circuitry is designed to reset the pinball machine.

I’m all for people repairing their own machines, and I’m happy to help and teach them.  But shotgunning parts on a printed circuit board is usually not good for the board.  The heat and physical stress from de-soldering a part will usually lift the copper pads or traces from the fiberglass, or pull out the metal plating that is inside the hole that the component pin is going through.

If you have a reset problem, get the correct diagnosis before swapping out parts.

With my oscilloscope, I can check BR2/C5 in about 60 seconds.  It’s immediately apparent when the bridge rectifier is defective; the pulsating DC will only have every other pulse showing.

Here are some reset issues I’ve worked on, and what the problem ended up being:

  • Independence Day (Sega):  This is one of the very few that ended up being the bridge rectifier.  In the case of Sega, it’s called BRDG21.
  • White Water (Williams): The owner was having a reset issue and had read about BR2 being the culprit.  He wanted me to verify that BR2 was defective. It turned out that it had already been replaced, along with C5. While I was at his home, the basement lights dimmed when the furnace blower switched on.  I suspected an issue with the house wiring (it was an old house).  He turned on other appliances and the line voltage reading dipped down to about 105 volts.  I suggested he try a uninterruptable power supply (UPS), normally used for computers, to handle these brownouts.
  • Black Knight (Williams): The basement of this home was wired with ground fault interrupters (GFI) after a flood.  Most pinball machines won’t work with a GFI.
  • Star Trek: The Next Generation (Williams): Reset issues on this machine were only in the first few minutes of power being turned on.  I replaced the inrush current limiter (varistor in the power switch box).
  • Starship Troopers (Sega): Resets were pretty random, but grouped together.  I traced the problem back to the F23 fuse.  It was loose in the fuse clips and running very hot.  I tightened the fuse clips and reinstalled the fuse.
  • Twilight Zone (Bally/Williams): The reset problem was caused by the power connector, where the 5 VDC leaves the Power/Driver board at J114.  The insulation displacement connectors (IDC) have metal forks which pierce the insulation of the wire to make a connection with the copper inside.  These connectors are problematic due to the wire working loose due to vibration and movement.  I re-seated the wires into the connector and the problem was solved.
  • Doctor Who (Bally/Midway): same problem as Twilight Zone above, except at connector J101, where the low voltage AC power enters the power driver board before going to BR2.

This is just a sampling, but failures of BR2 are not as common as some people think. Also, I have yet to see a case where C5 was weak or needed to be replaced.

Spy Hunter Pinball Machine (Bally, 1984)

Symptoms:  Not all sounds are being played.
Location: Broomfield, CO

The owner had already replaced the troublesome capacitors, but the Cheap Squeak sound board was not producing all of the sounds.  I bench tested the sound board and everything seemed to be fine.

It turns out there is a setting that enables the full set of sounds.  What makes it worse is that if the MPU board battery goes dead, the default sound setting is “chime only”.  To re-enable the full sound setting, follow these steps:

  1. Open the coin door and press the small test pushbutton switch near the top center of the door.  This will begin the various steps of the self test.  The first test is the display test and all of the displays will start counting up.
  2. Continue pressing and releasing the test switch, no faster than once per second, until you see the Ball Count display (right side middle of the back box) incrementing 1, 2, 3, with every press of the switch.  You will have to step through the other self-tests before the value will start incrementing.
  3. Continue pressing and releasing the test switch until it has counted up and the display is showing “18”. The other score displays will probably be displaying “00”.
  4. Press the replay/start button on the front of the coin door until the other displays are showing “03”.  They will increment starting at 00 with each press of the start button.
  5. Turn the power off, and wait about 5 seconds and power the machine back up.  Play a game and the sounds should all be working.  When a game is started, the background music will start.

On this particular pinball machine, pressing the replay/start button in Step 4 did not advance the value shown in the score displays.  This was because the switch was grounding out against the metal support behind the switch. There is normally a thin, stiff, piece of cardboard (called “fish paper”) to insulate the switch, but it was missing.  The metal support was taped to insulate it, and the switch began to work properly.

 

Electromechanical Pinball Machines

I’ve been working on a lot of EM (electromechanical) pinball machines lately. I usually don’t write about them here because the repairs don’t make for interesting reading and wouldn’t be very helpful to owners of other pinball machines.

Electromechanical pinball machines are the ones with the score reels instead of digital readouts, and have chimes and bells instead of electronic sounds. These machines use mechanical motors and stepper relays instead of digital circuits. It’s pretty amazing what they were able to accomplish using mechanical devices before the digital age.

As pinball machines become more popular, many of these old machines are being pulled out of storage and need a little TLC to get them up and running again.

Some of the EM pinball machines I’ve worked on recently are:

Dealer’s Choice (Williams, 1974)
Straight Flush (Williams, 1970)
Aladdin’s Castle (Bally, 1975)
Bally Hoo (Bally, 1969)
Super Star (Williams, 1972)

There are two common problems that most EM machines have: Dirty or mis-adjusted contacts or gummed up stepper units or score reel mechanisms.

There are hundreds switch contact points in an EM machine. Usually about one percent of the contacts are dirty or out of adjustment. What happens is the contact will get some dust on it. When the contact opens or closes, it sparks, which turns the dust into carbon. Carbon acts as a resistance, reducing the amount of power flowing through the contact. If the carbon build-up is substantial enough, no current will flow through the connection when the contact closes. Sometimes the current will flow through the contact generating heat, which is my theory on why they get out of adjustment.

The second most common problem is gummed up stepper units or score mechanisms. Usually the manufacturer put a thin layer of grease on the disk contacts. Over time this grease gets dirty and also turns into a sticky gel. The solution for this is to clean/rebuild the steppers.

I often get asked about the value of an EM pinball machine. Looking at the Mr. Pinball 2014 price guide, most EM pinball machines I’ve encountered are worth about $500, give or take, in very good condition. The items that affect the value the most are cosmetic, such as the playfield paint and wear, the backglass paint (which often peels), and the cabinet paint. These are the same things that affect the value of a solid state game, except supply and demand plays a bigger role.

Most of the time, electromechanical pinball machines are worth fixing, especially if the cosmetics are good.

Indiana Jones: The Pinball Adventure (Williams, 1993)

Symptoms: Left flipper not working.
Location: Denver, Colorado.

In Williams pinball machines of this vintage, the flippers are driven by the solid-state Fliptronics II board, which is located in the backbox in the upper left. There are 4 fuses, one for each possible flipper in the game: upper right, upper left, lower right, lower left (F901 – F904 respectively, each of which are 3 amp, slow-blow). In the case of many Williams machines, there are less than 4 flippers. If your game only has two flippers, you have 2 spare fuses available if you’re in a pinch.

In the case of Indiana Jones, there are only 2 flippers. When I was first contacted by the customer, I told him to check F904 which is labeled for the lower left flipper.  He went ahead and replaced it, but it didn’t solve the problem. So, I paid the machine a visit.

I determined that the flippers in this machine are wired to the upper flipper outputs, therefore F902 was the culprit. I vaguely recall running into this somewhere else.  Maybe that’s why it’s called The Pinball Adventure.

So, if you have a Fliptronics machine and you have a non-working flipper, check both upper and lower fuses.  By the way, the fuses are arranged on the circuit board in the same way the flippers are arranged on the playfield: upper left fuse is the upper left flipper, etc.

High Speed, Pinball Machine (Williams, 1986)

Symptoms: Some game-play functions not working, display issues, flasher bulbs not working.
Location: Frisco, Colorado.

When entering the test menu, the machine displayed a number of switch error codes. These switch error codes were causing the game to play incorrectly, for example, not diverting the ball to the “hide-outs”.  Although the error codes indicated 4 switches, only one was not working.  Ramp switch #42 wasn’t working, which prevented the ramp ball diverter from actuating, which caused the machine to think the other switches were also bad because the ball never rolled over them.  I cleaned switch #42 (very stubborn) and got it working.

Some segments of the alpha-numeric displays were not lighting up.  I traced this to cracked pins on the display tubes themselves.  I was able to repair some pins by soldering, but there was one pin broken right at the glass and couldn’t be soldered.  The owner deferred further repairs to the display. The options would be to replace the glass display tube or convert the game to new LED displays, such as PinScore.

When testing the bulbs, I noticed that none of the flasher bulbs were working. In this particular version of High Speed, the flashers are connected in series with a non-standard 7 volt bulb, #63.  Most pinball machines use 13 volt #89 bulbs for flashers.

The problem with bulbs connected in series is that when one bulb burns out, the other bulbs in the string stop working as well. This makes it difficult to determine which bulb is actually bad.  In High Speed, the bulbs are connected in pairs.  To make matters more unintuitive, the pairs are interleaved.  So for example, the 4 bulbs for the left center playfield are paired 1, 3, and 2, 4.

I had to take each bulb and test it with an ohm meter to determine the bad bulbs.  Once I got good bulbs paired with good bulbs, the flashers started working.  Since the whole theme of the game is associated with being chased by the police with flashing lights, I can’t imagine it was much fun to play without the flashers.

 

Super Nova Pinball Machine (GamePlan, 1980)

Symptoms: Display issues, switch matrix problems, roulette wheel problem, sound problems, lamp problems.
Location: Lyons Classic Pinball, Lyons, Colorado

I’ve been working on a couple of GamePlan pinball machines recently (Andromeda is the other one).  There is nothing particularly difficult about working on GamePlan machines with the exception of getting replacement mechanical parts.

This Super Nova pinball machine has an aftermarket market MPU board installed.  The aftermarket board uses an integrated RAM/Lithium battery component.  This is really a bad idea.  What happens when the battery dies?  It’s not replaceable without also replacing the RAM it’s attached to.  If there is anything I’ve learned in 30 years of electronics, that RAM/battery component will no longer be available when the battery needs to be replaced. This will render the board useless at some point in the future.

There was a display problem with most of the LED digits with segments that weren’t working.  This was related to the edge connector that connects to the MPU board.  Someone had added solder to the edge connector, presumably to get it to fit tighter, but it was uneven.  I took some 400 grit emery paper and sanded down the contacts so each had just a thin layer of solder.  This corrected the display issues.

This machine had several problems with the switch matrix.  You can download a .pdf of the Super Nova switch matrix here. The matrix consists of 40 switches arranged by 5 “strobes” (outputs) and 8 “lines” (inputs).

The first problem was with the Switch Catcher Unit (SCU-1) circuit board mounted on the underside of the playfield.  There are 4 switches connected to this board.  (The purpose of this board is to make sure the MPU doesn’t miss a quick hit to these switches; it lengthens the switch pulses.) When any of the switches made contact, it would momentarily short the 5 volt supply causing the machine to crash/reset. I traced this to a faulty 74279 chip.

pb-8980

Switch Catcher Unit (with faulty 74279 removed and a new socket installed)

Some of the solder joints on the backside of the connector were cracked and needed to be resoldered.

The second problem with the switch matrix was that every switch connected to Strobe #1 wasn’t working.  I traced this to a faulty connector at the MPU board.  It was just one pin that was bad.   None of the connectors used in this machine are standard Molex.  So rather than replacing only the bad crimp pin, I ended up replacing the entire connector.

Now that the displays and switches were working, I tested all of the solenoids.  The only thing that wasn’t working was the spinning roulette wheel (the Space Lab).  I traced this to a bad relay and a bad transistor driving the relay. The relay had been previously replaced with the wrong one. There was a 6V relay instead of a 28V relay. At some point the transistor that drives it shorted and burned up the coil.  After replacing both, the Space Lab wheel worked fine.

Next on the list was fixing the sound.  It was only making a single sound when playing the machine.  Usually when a machine has sound problems like this, I will check to see if there are any videos on YouTube so I can see how the sound is supposed to behave. The first thing I noticed is that there is no background sound during play.

There is a jumper on the sound board that enables or disables background sound. This was missing.  If background sound is desired, the jumper should be in the left position as shown in the photo below.

Sound Board with jumper for background sound circled.

Sound Board with jumper for background sound circled.  (Click for larger)

After installing the jumper, I was still missing many sounds. Using the oscilloscope, I check for pulses on the sound select inputs and there were none.  The problem was with the connectors at the bottom of the backbox.

Connectors at bottom of backbox.

Connectors at bottom of backbox.

At some point in the past, presumably with the original MPU board, the battery leaked and caused most of the connector pins to corrode.

Corroded connector pins.

Corroded connector pins.  The tip of the  male pin on the right had broken off.

All of the remaining problems with this machine were related to these backbox connectors. I don’t know the manufacturer of the original connectors, but I was able to find some Molex pins that worked (Male: 02-08-2004, Female: 02-08-1002). The old pins were removed from the housings using an extraction tool. Then the holes in the housings for the Molex male pins had to be enlarged using a 7/64″ drill.

Indiana Jones: The Pinball Adventure (Williams, 1993)

Symptom: When starting game, 6 balls would eject into shooter lane.
Location: Denver, Colorado

Obviously, there was a problem with the ball trough optical sensors.  Upon closer examination, there was no +12V supply voltage on either the IR emitter side of the trough or the receiver side.

To the backbox we go.  The +12V for misc playfield sensors comes off the lower left of the Power Driver board. There are several 4 pin connectors located there (J116, J117, and J118).  All of the connectors looked good.  I checked the +12V there and it was still missing.

Fuse F116 was blown, which supplies the 12 VAC to the rectifier, which generates the DC version of the voltage.

 

Circus Pinball Machine (Gottlieb, 1980)

Location: Loveland, Colorado.
Symptoms: Displays not working, playfield lighting blows fuse.

I was really impressed by how immaculate this machine was.  It looked like it had just been un-crated.  It definitely had low miles on it.  No doubt home-use only.

The interesting thing about the displays is that the credit/ball-in-play display was working, and the score displays were not working.  They were dark/off.  I spent a few minutes looking at the schematic, searching for what was in common with the score displays, and at the same time, not in common with the credit/ball-in-play display.  There was only one thing: the filament voltages for the display tubes.

The score displays run off a 5 VAC filament supply and the credit display uses a 3 VAC filament supply. So I started by measuring the filament voltage at the Player 1 display and sure enough, zero volts.  I lifted up the playfield and found the 5 VAC leaving the power transformer.  So somewhere in between a connection wasn’t being made.

I found the problem at a wire to wire backbox connector. In this case, there was a pin that wasn’t crimped correctly at the factory.

Crimp pin was installed incorrectly at the factory.  The wire was inserted too far and the crimp went around the insulation rather than the conductor.

Crimp pin was installed incorrectly at the factory. The wire was inserted too far and the crimp went around the insulation rather than the conductor.

If I held the wire a certain way, the displays lit up.  I replaced the pin with a new one and all was good.

The next issue was that the general illumination lighting on the playfield would randomly blow a fuse.  This is usually caused by a short at one of the sockets.  I checked each socket and found one that had been damaged (right near where the prop bar is used to prop up the playfield). Just the slightest vibration would cause it to short out.  I replaced that socket and one other socket that was marginal.

 

West Slope Lubrication Tour

Locations: Basalt, Colorado; Grand Junction, Colorado.

First stop was to work on a Wurlitzer jukebox model 1100 (1948). Like many 1940’s Wurlitzers I’ve worked on, the selector shaft and heart-shaped cam was not rotating due to lack of recent lubrication. This causes it to play one selection regardless of what was selected.  Unfortunately, it’s not an easy thing to get freed up, and takes a lot of exercising and time.

The jukebox needed some other adjustments, such as the turntable height and the clutch. I replaced the line cord because it was in very dangerous condition. The color cylinder plastic sheets were also replaced.  This was an original, un-restored 1100 in good condition.

Next stop was to work on two Williams electro-mechanical (EM) pinball machines, Grand Prix and Aztec (both 1976).  The main problem with these machines were the stepper units were not freely ratcheting up and down. This is the number one problem with EM pinball machines.  A quick disassembly, cleaning and lubrication fixes the problem.  Occasionally the spring tension needs to be adjusted.

The Aztec was missing some electrical parts, which can’t be obtained unless someone is parting one out.  But I got it working as best I could.

Next stop was an AMI jukebox, model A (1946), which is also known as the Mother of Plastic. The selection mechanism was frozen due to lack of recent lubrication.  Like the 1100 above, it took a while to get it unfrozen and moving freely.  The selection buttons needed some contact cleaning and lube (buttons would stick when pressed). The tonearm wire where it plugs into the amp needed to the resoldered.  All in all, a great sounding jukebox.

It was satisfying to breathe some new life into these old machines.

Alien Poker Pinball Machine (Williams, 1980)

Location: Boulder, CO
Symptoms: Incorrect sounds

The sounds on this pinball machine were working when the machine was first turned on.  But after a very short time, the sound would go wonky.

After some discussion with the owner, we determined that the sounds themselves were not defective, it was just the wrong sounds playing at the wrong time.  For example when hitting a particular target you’d hear the ball drain sound rather than a speech clip.

Pressing the self-test on the sound board yielded all of the correct sounds.  This indicated that the sound select signals from the MPU board were not correct.  With the oscilloscope, I looked for pulses on pins 10-14 on IC 10 on the sound board, while the owner triggered things on the playfield.  Some signals were missing.  When I checked the same signal at J3, they were present.

The culprit was cracked solder joints on the connector.  In the photo below you can see cracked solder around the top three and the bottom three pins on the connector.

Cracked solder joints on connector pins.

Cracked solder joints on connector pins.

At the owners request, I pulled all of the boards out of the backbox and touched up any solder joints that looked suspicious.  After that, the proper sounds sounds played at the proper times.