So Flash Gordon I'm working on has speech but no sound. Caps are fresh, and I've replaced the pots.

The J1 connector and wires were heavily hacked and at one point I was able to get game sounds working by wiggling the wires connected to J1 a bit, but never for long - usually just had static.

By "game sounds", I mean the ones triggered by in game switches and flashers. If I press the self-test button on the Squawk an Talk board all the speech works and also the sound effect at the end of the test, though I believe that sound is MPU generated and not part of the sound circuit that isn't working as noted below:

Source: http://www.pinrepair.com/bally/squawk.htm
"S&T Sound Generation.
Sound generation/special audio effects are handled in two ways: 1 - A General Instruments AY3-8912 PSG IC fitted in U12 under control of the PIA at U11 (which is in turn controlled by the S+T 680x) generates basic square wave tones, noise and sound shaping envelopes that control the output
amplitude to the amplifier circuits. These tones are limited in scope (how many different square waves can you have?). "

So I replaced the J1 connector and cleaned up the wires. Now I get nothing at all in terms of sound. On startup there is a clicking sound like someone tapping on a microphone. Though the self-test work fine still.

I swapped in a known working U11 PIA chip and no difference. I don't have a spare U12 though so will need to order one if I don't find another issue.

I've been testing the connections between J1 pins and their various landing pins on other boards to verify my connector wiring, and I've run into a few that are referenced in the schematic that I can't figure out where they are (shown with "?" next to them in the pic below).

SAndTJ1Schematic (resized).JPG

J1 P9 -> A3J3-5 ?
J1 P10 -> A3J3-12 ?
J1 P15 -> A3J3-23 and A2J3-19 ? (I discovered that A2J4-2 has continuity with J1 P15 though)

Does anyone know where these reside? I'd just like to know 100% that the connector is solid before I go and order a new U12 (AY3-8912 PSG IC). Other suggestions are also welcomed.

Good thinking, yes #18 is on 03.

Okay, there is no continuity between them, which is fine then.

All checkout.

Confirmed jumpers below U12 are not there, no continuity between the points either.
I didn't replace C36 myself but it is a modern looking cap.

I'm not precisely sure what you mean by "activity", but TP7 reads 5.09 -> 5.10 VDC, seems to waver between those two values but couldn't say that it's in response to a sound being played or not.

Yes, 5 flashes on boot on the S&T board.

Okay, I messed up earlier, was checking TP2, not TP7.

Using TP7 with the sound test (4th press on the service button on the coin door) I see no voltage change, steady at 0.044VDC. However, if I press the self-test button on the S&T board I do see the voltage changing on TP7 while the speach is playing, and even more dramatically during the sound effect at the end of the test.

So retested with the multimeter, self-test on the board TP7 ranges between 0.06 and 0.071VDC while the speech is playing, until the final sound effect plays (which I *can* hear), then the VDC jumps around to 0.2VDC TP7 while it plays.

What does TP7 do in game mode? (when the sound board is playing background sounds), and when you hit some switches? Sound test via the coin door just produces a filtered noise effect - they should have used a sound effect with tones and more substance.

Nada, TP7 sits at 0.066VDC during gameplay, including when targets are pressed. Same with self-test from coin door button pressed 4 times.

Can you post a clear high res picture of the sound board?

IMG_7837 (resized).JPG

Do you have an oscilloscope by any chance?

No scope. I do have a logic probe though.

Okay, something strange just happened, fired up the game and heard a siren sound during bootup (never played any sound during boot before). Started a game, sound playing! Not sure it's the right sounds though, if I hit targets different sounds play, but sometimes I get speech that repeats quickly, like "15, fif, fift, fifteen seconds".

Second boot, no startup sound. Start game, no speech at beginning. Hit targets maybe one in 4 or 5 switch hits will play a sound, doesn't seem to be same sound each time either.

Using the cabinet door sound test I heard the square wave sound for about 4 loops then it quits.

Third boot, no sound at startup or during game except it plays that square wave test sound maybe once in 10 or 15 switch hits. Self test via coin door plays nothing.

Self-test via S&B switch works fine still.

So it's very intermittent. Assuming the connector is fine now, maybe a socket issue?

Wiggling the harness doesn't seem to affect anything.

So now when I boot the game I get a startup sound, though not always the same one. When I start a game no sound or speech at all. If I hit targets/close switches the only ones that reliably play a sound are the bottom pop bumpers. They typically play the square wave test sound when hit and it will keep playing until I hit another switch, which will end the sound. Once in awhile a laser sound will play when the pop bumper is hit instead, but rarely.

Results:

U16 pin 13: Nothing (neither high nor low lit on logic probe)
U16 pin 11: High then short pulse to low as some/most coils fire, but not all.
U16 pin 9: High then short pulse to low as some/most coils fire, but not all.
U16 pin 7: High then short pulse to low as some/most coils fire, but not all.
U16 pin 5: Nothing (neither high nor low lit on logic probe)
U16 pin 3: Low
U16 pin 14: High with short pulse to low on some coils.
U16 pin 12: Nothing lit with short High pulse on some coils.
U16 pin 10: High
U16 pin 6: High
U16 pin 4: Nothing
U16 pin 2: High

Then go to sound test mode and recheck just the following two pins:
U16 pin 3: High with short pulse to low every second or so.
U16 pin 2: High with short pulse to low every second or so.

Note that if I just leave the sound test on I'm not hearing anything but every few mins it might play a sound for a couple of seconds and then stops again. When it does play, it seems to play for the full sound/1 second, and then when it quits there is a quiet "pop" sound instead of the next sound being played.

Results:

4.99VDC steady on u16 p1.

Using CMOS mode makes a difference:

U16 pin 14: High to low
U16 pin 12: Low to high
U16 pin 10: High to low
U16 pin 6: Low to high
U16 pin 4: Low
U16 pin 2: High

So a complete comparison of U16 input pins to output shows:

p9: HtL -> p10: LtH
p7: HtL -> p6: LtH
p11: HtL -> p12: LtH
p14: LtH -> p15: HtL
p3: Low -> p2: High
p5: High -> p4: Low

So everything checks out. ?

Well dang, saw some inconsistencies between my two captures above so went and did another pass with CMOS=on.

These set is mismatched now...

p5: Nothing -> p4: Low (with some noise)

If I touch the probe on P5, no lights, then if I touch on P4 afterwards, it starts with flickering High then gradually fades to solid low over about 1-2 seconds. Strange.

It looks like U16 is a CMOS 4049 hex inverter, I just happen to have some replacements on hand.

Should I go ahead and swap it out, or is this possibly still wiring related?

With the MM P4 reads 5.0vdc, P5 =0.00vdc.
For comparison, p3: Low (0.18vdc) -> p2: High (2.4vdc)

So to me that means that P4 (input) is high, and P5 (output) is low, which is correct. For some reason my probe doesn't like 0.0vdc.

In any case, all input and outputs pan out, right?

Well I've tested all the J1 connector wires for continuity with success in every case. Suggestions on how to proceed?

Dang. Yes I did, retesting shows p14: HtL -> p15: LtH

You can ignore pin 5 of U16, actually nothing is connected to it so that inverter isn't used - was wired for future expansion.

Okay, good to know.

Did you test from wires at the connector to the first component on the S&T board or just connector back to MPU board connector? In other words is the board seeing continuity to the wires at the connector?

I tested from the connector to connector, not connector to S&T board. So the J1 connector from S&T shows continuity with the expected various other connectors at the other ends of the wires.

I'll review the schematics and see if I can figure out where to test the J1 connector on the S&T board as well.

What was the game doing when you measured this? 2.4V out is very low for a "high" state signal on a CMOS chip. If the levels on these pins were steady (i.e. not actively pulsing) then this is suspect.. A high logic level out on a CMOS chip (which this is) should be 2/3 and above of supply voltage i.e. 3.3V or higher.

Retested:

With solenoid test running: P2: 5.4 vdc steady, P3: 0.133 vdc.
With sound test running: P2: 0.38 vdc, P3: 4.15 vdc

BTW, you might want to pull out the AY-3-8912 and give the pins a gentle clean with a small brass wire brush. Its pins look rather tarnished. The sound command sent from the MPU board through the U16 inverter goes to input pins on the AY-3-8912 - maybe you've got a bad connection there.

Yeah, I noticed that earlier and have already cleaned the bottoms of those pins. They weren't nearly as bad as the exposed tops though.

Okay, so I realized that my manual doesn't include schematics for the S&T board. The pins on the J1 connector looked a little tired so I went ahead and repinned it. Then with the plug installed I tested for continuity from the back of the plug "socket" to a test point on the board and all checks out. No change at all to the sound behavior after this though.

U16 pin 14 HtL --> U12 pin 14 LtH (reversed)
U16 pin 12 LtH --> U12 pin 13 LtH
U16 pin 10 L --> U12 pin 12 L
U16 pin 6 HtL --> U12 pin 11 LtH (reversed)

There's a special protocol for the MPU board sending a sound command to the S&T board. The protocol is timing related by the MPU boards CPU.
We recently had someone running an aftermarket MPU board that was configured for Sterns MPU-200 CPU speed which caused problems with the S&T playing correct speech/sounds. Are you running a factory Bally MPU board or aftermarket? Do you have anything that can measure the MPU boards CPU frequency, i.e. a frequency counter? maybe your multi-meter has a Hz (frequency) feature?

Based on the behavior where it seems like once the sound chip receives a command to play a sound, it does so until finished, and the fact that the sounds will play when one switch is closed and stop when another is closed, this seems more like a signalling issue between the MPU and the sound chip (as you've indicated, above).

MPU board is a Bally AS-2518-35 that came from a donor and I re-jumpered for new ROMs.

I'd don't have a way to read frequencies, unfortunately.

Any idea where I could find the schematics for the S&T board?

Wow, this is fantastic!

After looking through the S&T schematic yesterday I could see the 4 sound select lines you are referencing and how I should be able to trace them back. This explanation makes that a whole lot easier for me. I really appreciate this, I know how much time it can take to put together in-depth responses.

At one point during my testing last week the game was playing sound intermittently (which I noted). To summarize:

- On startup it sometimes played a sound, though not always the same sound.
- If I left the sound test running every few mins it might play a sound for a couple of seconds and then stop again. When it does play, it seems to play for the full sound/1 second, and then when it quits there is a quiet "pop" sound instead of the next sound being played.
- Starting a game, no sound. Closing switches would sometimes play a sound, again, not always the same sound for the same switch. Lower PF pop bumpers seemed reliable to trigger a sound, sound would keep playing until I closed another switch, then it would stop (and not play the sound for the switch).

All of this was sporadic and I haven't gotten any sounds this way in the last day or two. On-board self-test still works fine.

It does seem to be that there is no problem downstream of the sound chip, it is in determining what sound to play, and when. Once a sound is initiated, it plays and sounds fine.

I will retest in a bit and do some further tracing and report back.

Thanks again!

Okay, I have exciting news!

With all of my apparent mixups when testing and reporting values on U16 I was becoming suspicious. I'm not perfect but I'm not *that* error prone.

So I retested U16 P10 and P12 and indeed P10 was low and not changing, though P9 was.

Decided that U16 had to be flaky and reporting different results at different times during my testing - replaced U16 and the game is working perfectly now!

I would like to extend my heartfelt appreciation for all the help provided.