Aladdin’s Castle, Bally Pinball Machine, 1976

Location: Longmont, CO.

Symptom: Pinball machine wouldn’t reset.

I visually checked the most likely relays and contacts: Game Over, Coin, and Credit.  After not finding any problems with the relays, I checked the switches on the score motor unit.  I noticed that the motor was hot and was obviously energized, but not turning.  The cams rotated freely but the electric motor itself was seized.

I disassembled the motor and managed to get the seized shaft out of the bearing. I cleaned both shaft and the bearing, but there was still some type of corrosion on the shaft that was preventing it from turning freely.  I would have preferred burnishing the shaft with some very fine 0000 steel wool, but I didn’t have any on-hand.  I used some 800 grit sandpaper and removed the corrosion (or whatever it was).  I lubricated the bearing and the shaft so that it slid and rotated freely.

After reinstalling the motor into the scoring unit, the machine reset correctly when pressing the start button.

The scoring motor and switch bank assembly is basically the heart beat that runs an electromechanical game.  Without it, nothing will work.

We (the owner and I) noticed that the Ball In Play lights were not resetting back to Ball 1 during the reset.  I checked the stepper in the backbox and saw that it was sticky, which is a common problem.  It didn’t seem sticky enough that it warranted rebuilding. The Ball In Play stepper unit is a step-up and reset type. The owner cleaned the old lubrication from the contacts and I added a turn on the spring.  The stepper seemed to function fine after that.  I informed the owner that if problems continued with the ball counter, that it would have to be cleaned and re-built.

We replaced some burned out lights on the playfield and talked about playfield care.  I mentioned that a lot of collectors have differing opinions, but that I liked Novus 2 for playfield cleaning and polishing.

As for me, I’m going to start carrying 0000 steel wool in my repair kits for cleaning up motor and stepper shafts.  It has worked well with jukebox restorations I’ve done.

United Shuffle Alley Bowling Machine, 1949

Location: Roxborough, CO

Symptoms: Various items not working properly

The owner had previously rebuilt the stepper relays and adjusted the relay contacts which had improved things considerably.  There were a few things that were still not working, including the “strike” and “spare” scoring, and the game reset.

After tracing two problems back to relay contacts that needed adjusting, I decided it would be more efficient to check and adjust all of the relay contacts rather than tracing each symptom. I found 3 more sets of contacts that needed adjusting.  After that everything was working pretty well.

It was an amazing machine with the sound of the electro-mechanical parts, buzzing, clicking and ratcheting the score, all of which was amplified by the wooden cabinet.  Some EM pinball machines are more complicated by comparison, but they don’t have solenoids and relay coils that are as large and noisy.  This thing is awesome!

Bobby Orr Power Play, Bally Pinball Machine

Location: Littleton, Colorado.

Symptoms: Two of the displays had digits that were blank (off). General illumination lights on the backbox were not working.

I ran the display self-test to determine the full extent of the display failures.  One digit on the Player 1 score and one digit on the Player 4 score were not working. They were different digits, so it wasn’t a problem with connections from the CPU board.

I started with the Player 4 display.  I disconnected and removed the display from the backbox panel and checked the digit driver transistors with a mulitmeter on the “diode” scale.  The transistors checked OK.  This was a little surprising to me.  So I connected an oscilloscope to the digit drive signal.  It was OK where it was coming into the display board.  There was no signal on the collector of level shifter transistor (Q6 in the diagram).  Surprisingly, the 100K resistor was bad and read “open”.

Schematic showing one of the digit drivers.

I had previously read about the power rating of these 100K resistors being too small as originally designed, but I wasn’t expecting to see a failure with no visual indication of the resistor being over-heated.  I didn’t have any spare 100K resistors with me, but fortunately Radio Shack is still good for some very common parts.

I disconnected and removed the Player 1 display and immediately checked the resistor for the failed digit and it too was bad.  With the resistors from the nearby Radio Shack, I was able to desolder and replace the resistors on-site.  I re-assembled the displays into the backbox and re-tested.  Everything worked fine.

As for the general illumination lights not working on the backbox, this is a common problem and is often related to a connector.  This was no exception.  I traced the non-working branch of lights back to the connector at the power supply board.  A portion of the connector on the PCB had been previously replaced, with a version with slightly shorter pins.  The lights started working with just touching the connector.  I checked the solder connections on the back side of the PCB and cleaned the connector contacts.

As I usually do when I work on a pinball machine, I ran a light test and replaced several bulbs that were burned out on the playfield.


Chicago Coin, Fighting Irish and Big Hit pinball machines

Location: Longmont, Colorado.

Symptoms: Neither machine working correctly.  Many solenoids powered-up continuously and getting hot, relays buzzing, etc.

Both of these pinball machines are electro-mechanical (EM) from the early 1950’s; Fighting Irish (1950) and Big Hit (1952).  Given the age of these machines, both are in relatively good condition.  They have been in the owner’s family for decades, but haven’t been used for about 10 years.

I inspected the machines and both had sticky, gummed-up, stepper relays.  The Fighting Irish also had some sticky latching relays. The original lubrication on these parts had become gummy over the past 60 years.  This is a common problem with mechanical parts in older pinball machines and jukeboxes.  There is really no easy way to fix this other than to disassemble, clean, re-lube and re-assemble the parts.  Spraying something like WD-40 into the parts, if it works at all, is only a temporary fix.  Days later it will again be sticky once the WD-40 solvent evaporates.

I usually use denatured alcohol to clean the parts, then depending on the situation, use SAE 20 electrical motor oil for lubrication of shafts and Teflon grease for gears and slides.  (Avoid using white lithium grease, as it separates and dries out quickly.)  This only applies to older jukeboxes and pinball machines.  Solid-state pinball machines require practically no lubrication.

I rebuilt the stepper relays and showed the owner how to do it also.  Between the two of us, we got all of the mechanical parts working freely. We had to adjust the solenoid position on one stepper used in Big Hit because there wasn’t enough travel for the “catch lever” to reliably engage the tooth of the main stepper wheel.

The flippers on Big Hit were also gummed up.  They were removed, cleaned and re-assembled.

We powered-up Big Hit and all electrical items seemed to be working fine.  We powered-up Fighting Irish and it still had some electrical problems that needed to be addressed.

I found that the main connection to the secondary of the transformer had broken off.  I re-soldered it.  That got most things working.  Then after scoring about 70,000 points, the machine’s score motor would start continuously running, racking up points and never stopping.

There aren’t any existing schematics available for either of these machines.  So I just watched various relays and determined which ones were used when the machine got stuck scoring.  I held some relay contacts with my fingers and determined that one was repeatedly engaged when the scoring problem started.  With a tone and probe, I was able to trace the power for that relay to another relay’s contacts, and then back to another relay, and then back to the original relay where I discovered a broken wire to one of the contacts.  I re-soldered the wire and all worked fine after that.  When using a tone and probe, the machine must be powered off (unless you’ve got one that can withstand the AC voltage spikes present in an EM game).

A tone and probe seems to be a good tool for tracing wires in a pinball machine, especially when one doesn’t have a schematic.  It was invaluable on this pinball machine repair.  I’m still learning how to use it effectively on EM pinball machines. Since I don’t have an EM machine of my own, I will have to wait until the next repair call to learn more.

The owner of these machines has a number of minor repair items to take care of, but the bulk of the electro-mechanical problems are fixed.


Wurlitzer 3500 “Zodiac” Jukebox

Location: Longmont, CO

Symptoms:  Would not return after playing record.  Sound playing out of only one speaker (out of 4 speakers total).

As with other Wurlitzers of this era, the trip switch fails intermittently. I haven’t found a source for replacement switches, yet.  But with past experience, opening the switch and spraying contact cleaner inside seems to solve the problem.  The switch is glued together and to open it requires breaking it open.  This is done by inserting an X-acto knife under the edge of the top cover at the opposite end from where the trip wire attaches.  Usually about a half inch of the cover breaks away cleanly, allowing it to be re-glued or taped together again.  The contacts are directly underneath where the cover is removed.  I spray the contact cleaner on the contacts and cycle the switch dozens of times.

When playing records, I noticed the pair of top speakers (tweeters) were not working, and one woofer in the bottom was not working, leaving only one speaker that was working. No wonder it didn’t sound very good.  Wurlitzer, instead of referring to the right and left channels, refers to them as “A” and “B”. I swapped the speaker leads and determined that channel “A” of the amplifier wasn’t working. At the same time, when moving the “A” speakers to the “B” channel, the tweeter of channel “A” worked fine.  So there were two problems with the sound, channel “A” of the amp wasn’t working and the tweeter of channel “B” wasn’t working.

I decided to tackle the amp first.  I swapped the input cables from the tonearm to make sure the problem was not the cartridge or the tonearm wiring.  With my oscilloscope, I traced the signal from the input to the output of channel “A”.  The signal was fine until it got to Q6 and Q8.  Beyond that point, it was dead.  I checked the bias voltage between the base and emitter of each transistor. Q8 was 0.05 volts, which is way below the 0.6V needed.  Q6 was 1.2V, which was double the 0.6V that it should be.

At this point, since I knew I had a bad transistor and that it would have to cross-referenced and a replacement ordered, I decided to take the amp back to my shop for final repair.

Before I removed the amp from the jukebox, I wanted to find out what was wrong with the tweeter on channel “B”.  After checking continuity of the speaker connections with the ohmmeter, and checking the coil resistance of the speaker, the only thing left was the 8 uF capacitor in series with speaker connection.  I bypassed the capacitor with a jumper wire and it started working.  The capacitor is used to block the bass frequencies from coming out of the tweeter.  Somehow the capacitor had failed in an open condition. I added this to the list of parts to order.

With the amplifier at my home shop, I unsoldered both Q6 and Q8 from the printed circuit board to isolate them so I could test them individually.  I checked the base-emitter junctions with the multimeter in “diode” mode.  The junctions on silicon transistors should look like a diode, 0.6V one way, open circuit the other way.  In the case of Q6, it was open both ways.  Definitely bad.  Q8 checked OK and didn’t have any shorts between any pins.

As usual, with Wurlitzer, it is difficult to cross reference transistors because they used their own part numbers on them.  According to the service manual, Q6 is a 130537-5. Sometimes the first place I’ll go looking for a replacement semiconductor is NTE.  In this case, I typed in the part number 130537-5 and got NTE289A.  The basic specs seemed applicable to the circuit, so I was confident it was a good cross reference.  Another source I’ll use sometimes is this page that shows some of the cross references for Wurlitzer transistors.

As I was working on this amp, I noticed some leakage on two of the capacitors.

Capacitors (C19) on both channels are leaking electrolyte.  Click for larger.

A lot of people advocate replacing all of the electrolytic capacitors in a solid state amp of this vintage.  While electrolytic caps do have limited life, I prefer to wait until there are visible or audible symptoms.   The gray caps in the center of the above photo look burnt, but that is just residue from burning dust on some power resistors not shown in the photo.  Also, jukeboxes have spent most of their lives in establishments that had a lot of cigarette smoke, which leaves a residue.

The two leaky capacitors were replaced.  Since the new caps were much smaller with shorter leads, I soldered them directly to the back of the circuit board.  With the transistor replaced, the amp works fine.


Wurlitzer 3110, “Americana” Jukebox

Location: Boulder, CO

Symptom: Sound in one channel stops working after 20 minutes of use.

I have worked on this jukebox in the past for other reasons; it gets a lot of daily use.  When I first arrived, I couldn’t find anything wrong. I could hear sound coming out of each speaker.  I checked all of the connections.

We were on the fourth or fifth record, and I was about ready to leave, when it stopped working. It started with static sounds, then after about a minute, the sound was totally gone.  After further checking of connections, I learned it was sensitive to vibration.  I could tap the final stage of the power amp with the handle of my screwdriver and the problem would change in severity, but wouldn’t begin working completely.  My initial thought was there was a bad connection in the socket for one of the power transistors.

I brought the amp back to my home shop and removed each power transistor and replaced the mica insulators and cleaned the pin contacts.  I also ordered and replaced the larger electrolytic capacitors in the power supply circuit and the final power amp circuit.

Electrolytic capacitors gradually lose their capacitance with age.  I assume this is because the electrolyte paste eventually dries out.  In some cases, especially in high voltage vacuum tube amps, the paste will leak out and the capacitor eventually shorts out.   This amp is over 50 years old, and although the capacitors seemed fine, it’s just a matter of time.

However, after rebuilding the amp, the original problem still existed.  I isolated the problem to Q15.  This transistor drives the transformer that phases the final output stage drivers.  As a double check, I swapped the channel “A” Q15 with the channel “B” Q15, and the problem moved to the other channel.  I don’t know why it’s sensitive to vibration, but there must be a problem with an internal connection inside the transistor case. In any case, it needed to be replaced.

One of the difficulties working on solid state Wurlitzer amplifiers is that Wurlitzer used their own part numbers for transistors, probably even specifying to the manufacturer to print their part number on the case.

Power amplifier stage for channel "B" showing transistors with Wurlitzer part numbers. Note the 6648 number, which is the date code: manufactured the 48 week of 1966.

It’s difficult to figure out what the transistor is and to find a suitable replacement.  Fortunately, with some help from Google and Bing, I found out this same transistor (Wurlitzer part number 125721) is also used in their electric organs.  Somebody figured out it matches an NTE121, Germanium PNP Transistor, Audio Frequency Power Amplifier.

I was surprised to find that somebody was still making germanium transistors.  (99.9% are made from silicon.)  Keep in mind that the forward bias voltage on a germanium transistor is 0.1V, compared to silicon forward bias of 0.6V.  This is something I had forgotten.

The replacement part was ordered and the amp is now working fabulously.


Dirty Harry Pinball Machine, Bally/Williams

Location: Denver, Colorado.

Symptoms: Dot matrix display (DMD) flickering and displaying random garbage.  Test Report listed Sound Error.

This machine looked to be in very good and clean condition.

I opened up the backbox to investigate the display problem. There is a wide gray ribbon cable that connects the CPU board to the Dot Matrix Controller Board.  All I had to do was touch the connector where it connects to the display board and the problem went away and never came back. The connector didn’t appear loose.  I disconnected connector from the board, squeezed it from front to back to make sure the connector was making a good connection with the ribbon cable.

This type of connector is referred to as an Insulation Displacement Connector (IDC).  There are small gold plated forks that pierce the insulation in the ribbon cable and make connections with the internal conductors.  Sometimes a little oxidation forms between the the fork and the conductor and squeezing will help restore the connection.  If it’s really bad, the connector can usually be carefully removed and re-crimped on the cable next to the original location.  (To do this correctly, you will need a padded vise to gently squeeze the connector onto the cable without breaking the connector.)

Once the display was working, re-powering the machine indicated there was a Test Report.  This will only display if the the computer has detected a problem.  Viewing the test report indicated a Sound Error.  I went to the test menu and performed a Sound Test.  All of the sounds were working fine.  I powered down the machine and reseated all of the sound EPROMS on the sound board.  Upon powering the machine back up, the Test Report didn’t show up, indicating the problem was no longer present.

The customer asked about routine maintenance and what should be done.  The most important routine maintenance is keeping the machine’s playfield clean.  Every time the pinball hits a pop bumper, slingshot or even the flipper, the friction of the ball accelerating or decelerating on the playfield will cause a microscopic amount of metal dust to come off the ball. After a while, the grayish metal dust accumulates and starts to wear into the playfield. It migrates down into the switch contacts and mechanical parts, wreaking havoc. It leads to all kinds of intermittent and permanent failures.

Likewise, since the microscopic metal dust is coming off the ball, the balls should be periodically replaced. The ball surface develops microscopic pits, which makes the ball rougher, which increases the friction mentioned above, which increases the wear on the playfield, etc.

One quick look at this Dirty Harry pinball machine told me the owner had been keeping it clean.


Judge Dredd follow up

Location: Denver, Colorado.

A few months later, I was called back to the Judge Dredd machine mentioned in this previous post.

The thermistor in the power line module had burned out.  I replaced it and tested the machine.  I noticed that the shorted switch-row problem had returned.  The nice thing was, it was a solid failure this time and I was able to track it down.

In my previous post, I mentioned the short would go away if I unplugged J212.  This was the case again, and it deceived me into thinking the short was somewhere along this stretch of wires that heads towards the coin door.   In the switch table in the service manual, Switch 24 is listed as Always Closed.

(The switches and lamps in a pinball machine of this vintage are multiplexed in a grid.  This reduces the complexity of circuitry and wiring.  Rather than having 64 separate wires to 64 switches, by multiplexing, 64 switches can be monitored with just 16 wires.)

Somewhere in the machine, there is a diode connecting Column 2 and Row 4.  It turns out it’s on Coin Door Interface Board, which is were the wires from J212 end up (D2).

What is deceiving is that when a column is shorted to ground, any switch that is closed will show that row to be shorted.  Since switch 24 bridges column 2 and row 4, every switch in row 4 was shown as closed, leading me to believe the problem was in row 4.  But when I closed other switches in column 2, the entire corresponding row would show as closed.  And as a verification, I open the diode at D2 and the remaining switches in row 4 began functioning normally.

Once I determined the problem was actually in column 2, and the symptom indicated it was shorted to ground, I unplugged J205 and J212 to isolate the CPU board from the rest of the machine.  I measured the outputs of each column driver and found the output of U20, pin 17 was stuck at ground.

I replaced the ULN2803 with a new part and everything in the pinball machine tested normal.


The Shadow Pinball Machine, Bally/Williams

Location: Erie, Colorado.

Symptoms:  Neither ball diverter was working, drop targets on mini-playfield were sticking.

The ball diverters at on the left and right ramps were simply loose on their shafts.  I tightened up the set screws on the left diverter.  The right diverter was binding a little after tightening the set screws and I determined the arm was slightly bent.  I straightened it and it worked fine.

Right ramp diverter, with rings on either side. (Photo from

I removed and cleaned the drop targets from the mini-playfield.  They had been lubricated with white lithium grease, and as usual, it had dried and was causing them to stick.

I looked at the Test Report and it reported that the switch on the main playfield drop target was not working.  It was simply a case of the drop target sticking and never engaging the switch.  I removed and cleaned the drop target.

There were a couple of broken wires that I re-soldered to their correct locations; one was a flasher and the other a GI bulb.  I replaced 24 bad bulbs. There are four rings with lighted jewels incorporated into the playfield (see photo above).  They are illuminated with a grain of rice lamp that is hard-wired.  Three of those lights on this machine are out.  Marco has these and will order.  I’m guessing they are glued in place.

The machine also needs a new set of pinballs.   Any pinballs that don’t look shiny and new should be replaced on any machine in order to keep the playfield in good condition.

All in all, just routine maintenance.

Judge Dredd Pinball Machine (Midway/Bally)

Location: Denver, CO (Arvada)

Symptoms: Flipper was stuck, multiple connection problems, neither slingshot working.

Stuck flipper: The flipper was stuck because the end of stroke (EOS) switch was broken exactly at the flipper’s maximum travel point.  The flipper wouldn’t return because it was getting hung up on the broken contact.  For a quick fix, I put a piece of shrink tubing over the end of the contact to lengthen it.  As the tubing cooled, I flattened it with my fingers, which basically made a plastic extension for the contact.  Eventually the switch will need to be replaced, but for now it is working fine.

Connection problems:  There were some connection problems causing balls to be continuously delivered to the shooting lane.  Taking a look at the switch test diagnostics, I determined a whole “row” of switches was shorted to ground (row 4).  After checking every switch location in the row on the cabinet and under the playfield, and not seeing any obvious shorts, I decided to look at the back box.  There I discovered if I pressed on the lower corner CPU, the short would go away.  My initial assumption was that it was something on the CPU board shorting out.  Yet, when I unplugged J212 the problem would disappear.


Row 4 is the end pin on the connector (pin 8). So why does lightly pressing on the CPU board cause something associated with J212 to short/unshort.  I looked all around and traced the wires back and couldn’t see anything.  Unfortunately, I moved something and the short went away and never came back while I was at the customer’s site.  After I left the short returned. On a subsequent visit, the problem had gone away again and couldn’t be found.

Slingshots: The owner didn’t realize that neither slingshot was working.  Both had broken mechanical links from the plungers, one had a broken arm.  Replacement parts were ordered and installed.

Doing one last check of the Test Report, I saw that the EOS switch on the left upper flipper was disconnected.  I found the broken wire and resoldered it to the switch.

The Test Report also revealed that one of the ball trough opto switches wasn’t working.  Using a cell phone camera, which can see the infrared light of the opto switches, I determined that one of the IR LED’s was not illuminated.  I checked continuity and didn’t find any problems with the circuit board.  The LED was burned out.  After a discussion with the owner who assured me that the game is working fine, we decided to leave it as is.  There are 7 opto switches monitoring the ball trough and it was of the ones in the middle.  My guess is since the CPU knows it’s bad, it can work around it.  If it causes a problem, there is a very nice after-market IR LED board available which has several design improvements.

And finally, the entire game had intermittent power.  I quickly determined that the line cord was not plugged securely into the rear of the game.

Other than just a few test games, I’ve never played Judge Dredd.  It seems like an interesting and challenging game.

This post has a follow-up here.