Seeburg AY/AQ 100/160 Jukeboxes

Symptom:  After about 15-20 minutes, the mechanism no longer stops at selected records.

I’ve come across a couple of Seeburg jukeboxes of this vintage that have had this problem.  It has been a difficult problem to diagnose in the field, but I was fortunate that an owner allowed me to take the Tormat Selector Unit (TSU) back to my home office where I could bench-test it and really dig down to find out what the problem is.

The first thing I had to do was replace R513 (2.2 Meg) because the -7 volt test signal wasn’t working.  This test voltage is present on TP-C.  Momentarily feeding this voltage into the RCA jack where the Tormat plugs in (J510) will test the pulse amplifier.  The Tormat Pulse Amplifier (TPA) is the gold colored box mounted to the TSU, with a 12AX7 tube on it.

Tormat Pulse Amplifier (TPA)

Tormat Pulse Amplifier (TPA)

Fairly quickly, I was able to determine the problem was in the pulse amplifier.  I hooked my oscilloscope to the output at Pin 4 of the TPA socket (J509), accessible from the rear of the TSU.

I connected a jumper lead from TP-C to the shaft of a small Phillips screwdriver that would fit into the RCA jack.  Moving the screw driver in and out would send pulses through the amp.

Output of pulse amplifier when working correctly.  Pulse is about 75 milliseconds.

Output of pulse amplifier when working correctly. Pulse is about 75 milliseconds.  Note ground is at top of pulse and the output normally sits at about -76V.

When everything was cool, the pulse output was 75 milliseconds long.  As the TPA and TSU warmed up, the pulse got shorter and shorter until it was gone.

Output of pulse amplifier after it warmed up.

Output of pulse amplifier after it warmed up.  Pulse is now way too short to trip the mechanism solenoid.

After trying a new 12AX7 tube, I opened the TPA box and checked the resistors.  Some were marginally out of tolerance (all to the high side), but this was minor.  Using hot air, I heated the amp and measured the resistors again and there wasn’t a significant change.

I measured the capacitance of C506 (0.05uF, 400V), which is partially responsible for stretching the pulse to 75 milliseconds.  It was 40% higher than it’s rated value.  As I heated it up, it’s value more than doubled from it’s already high value.  This was the culprit.

The closest capacitor I had on-hand was 0.05uF at half the rated voltage.  Since I wasn’t seeing more than 195V in the circuit, I tried a 200 volt cap as a temporary fix.  And it worked!  The pulse output remained at 75 milliseconds throughout the hour I tested it.  I have ordered the 400V version of the cap as a permanent replacement.

New capacitor installed.

New capacitor installed.

I’m glad to have finally solved this mystery.



Star Trek: The Next Generation, Williams Pinball Machine (1993)

Location: Lone Tree, Colorado.
Symptoms: Kept losing track of balls.

I have worked on more ST:TNG pinball machines than any other model.  Which is kind of cool since it is one of my favorite games to play.

This machine needed cleaning and tuning up.  It also suffered from broken wires on one of the cannons, which is a problem I’ve seen with every ST:TNG I’ve worked on.  The rotation of the cannons causes the wires to flex.  Eventually after a thousand flexes, a wire will break. Somebody should supply replacement wiring harnesses — connectors on one end and bare wires on the other — to make replacement easier.

Usually what I do is identify which wire is broken and run a replacement beside the original harness.  So far, there has always been more than one wire broken.

The biggest problem with diagnosing these broken wires is that when the cannon is sitting in its normal home position, everything is fine.  Usually the wires open when the cannon rotates out to the playfield.  And the problem with the diagnostics is that you can’t test the solenoid, light and opto-sensor while the cannon is moving.

This machine had an interesting symptom where during game play, the ball would load in the the cannon, then it would swing out, but it wouldn’t shoot until it was back in the home position. This would fire the ball back down below the playfield on top of an existing ball.  There is a limit switch that is supposed to keep you from shooting the ball anywhere other than the open playfield. Apparently this limit switch is ignored if the solenoid wires break open when the cannon rotates out.

After I repaired the broken wire to the solenoid, I noticed the cannon was shooting during start-up. This symptom I had learned about on a previous repair.  One of the wires to the opto-sensor was broken.  The machine thinks there is a ball there and tries to get rid of it.

After fixing the cannon, the machine would still lose track of the balls under the playfield.  I discovered the ball diverters under the playfield were sticking.  I cleaned those, as well as the opto-sensors and it seems to have solved all of the problems.

Although the game is working fine, the right outlane switch is bad and will be replaced on a subsequent visit.