I picked up a Medusa a few weeks back and when I got it home you couldn't hear the speech. I adjusted the volume pots on the sound board (both sound vol and speech vol) until I could hear both clearly. The pots didn't adjust the volumes smoothly but would more cut in and out as you turned them, but was able to find a sweet spot for both eventually.

At that point there was noticeable background hum coming through the speaker, and sometimes a sound like an 'electronic windstorm'. However, while playing the sound quality was crisp and the speech clear.

I also noticed that the board had original caps on it. So today I replaced all the caps on the Squawk & Talk board as outlined on the GPE kit list (https://www.greatplainselectronics.com/proddetail.asp?prod=BALLY-STE-KIT). I didn't replace c16 or c22 though as these appeared to be tantalum caps.

When I turned the game on I once again had trouble finding a sweet spot with onboard volume pots. The sound had shifted where the high frequencies are now somewhat muted, and the speech is fairly distorted. Plus the 'electronic windstorm' sounds were still present.

So, I figured the volume pots were shot, and removed them. Not having replacements handy I instead installed a 330 Ohm 1/2 watt resistor is each pot's place (I think the pots were 1K Ohm), figuring these would be a reasonable value until I get new pots. Unfortunately, the sound still has muted highs and the speech is distorted, but the 'electronic windstorm' has at least disappeared.

Finally, I read that the TMS5200 could cause sound quality problems due to corroded pins so I popped it out and cleaned the pins, inspected the socket and solder points - all seem fine. No improvement in sound quality.

I'm at a bit of a loss since the board sounded fine before I replaced the caps. Suggestions?

I double-checked all the caps that I replaced and they match what is specified on the GPE kit. I haven't had a chance to verify those values with what's in the schematic yet, though I would be surprised if the GPE kit spec was wrong at this point.

When I was working on the sound board I noticed that the speaker wiring was hacked and one of the wires was just loosely wrapped through the speaker connection. I fixed that up but now I'm wondering if maybe the speaker could be damaged, resulting in too much bass and not enough highs?

Which R# is that? Would a 1Kohm pot work in that location?

These are now 2.2uFs, as specified in the GPE kit. I assumed the 2.0uFs were not widely available? Could that be the cause of the "clipped" high end?

These are all correct and were replaced as part of the Cap refresh. Not sure how I can test them though?

Failing that we need to start considering replacing some of your ceramics.
Don't shotgun these yet (let's figure out the new problem first), but these are all part
of what shapes tone in this circuit:
C21 (100 pF), C32 (68 pF), C23 (0.47 uF), C30 (0.1 uF)

I guess I could just shotgun replace all these. The shipping cost is more than the components, and the thought of replacing each one and then retesting the board seems like a lot of extra work.

Your pots probably just needed a cleaning. If you can clean them carefully and get them working smoothly I'd be tempted to reinstall.

I had the same thought and cleaned and tested them on the bench and re-installed them. They seem to be working okay, though the actual impact of adjusting them is a bit strange. I have the sound volume at max as that seemed to provide the best high-end sounds, but the speech one goes to distortion when set higher than about 50%. I suppose the pre-amp is being overdriven at higher volume. Would there be a way to improve them via component replacement? Certainly the speech sounds very low freq. and like it is missing the higher freq. vs. the sounds themselves. All the sound on this game is pretty low freq. heavy though.

I also replaced the orig. speaker with a spare Infinity-brand car audio speaker I had pulled from a vehicle a few years back, new one is rated 4 Ohm and 20V, but has a much larger magnet.

In the spirit of experimentation I went ahead and swapped these out for 1.8uF caps. I also replaced the 1k ohm volume pots on the board with new ones. At first I thought things had improved somewhat but when I went to adjust the vol pot ALL the sound cut out.

Restarting the game fixed things but after a few mins sound would cut out again and I noticed the green LED would stay lit on the board when this happens. Could the onboard amp be being overdriven?

So I pulled the board and replaced the volume pot with a 380 ohm resistor. Game seems to work again but haven't ran it very long. Sound seems the same as before, the only thing that improves the highs in the speech is to crank the speech volume pot, but then it distorts like crazy.

So I'll probably put the 2.2 uF caps back in and try another new 1k ohm vol pot tomorrow.

The sound isn't distorting, just the speech. They both have somewhat clipped high frequencies, but really noticeable on the speech.

Jumpers are set to N (remote speech pot) and DD (remote sound pot), and M and CC are not connected (as expected).

After reading more details on the Squak and Talk board (http://www.pinrepair.com/bally/squawk.htm) I see that it intentionally filters and drops off volume for higher frequencies. I think what I am experiencing may be considered normal for this game/board.

I do wonder if using the computer controlled volume instead of the onboard pots might provide a bit more control to try to get the maximum speech volume/clarity vs. distortion. Does Medusa have the audit settings to adjust the sound and speech volume? What audit #s are they (nothing in the manual)?

Finally, my sound board has started to lock up since I put the 1.8uF caps in yesterday to C36, C43 (could be a coincidence in terms of timing). I replaced those with the 2.2uF ones again, but now when the board is cold it boots fine, but will lockup/quit producing sound after a minute or so. Once it's warm it will not boot at all (no flicker on the LED, stays solid). Letting it cool for a few minutes it will boot again, but lockup after a min. or so, etc.).

I pulled the following values from the S&T test-points (all VDC):

TP1: 0.03
TP2: Couldn't locate this TP?
TP3: 13.9
TP4: -4.8
TP5: 3.0
TP6: 2.0
TP7: 0.3
TP8: 2.4
TP9: 1.2
TP10: 4.0
TP11: 1.3
TP12: 4.6

Both the SPB and the S&T have had their caps replaced in the last month or so.

Thanks for the help!

Okay I read 4.98v at pin 35 on U1.

The CR2-4 diodes look good and test at 14.72v coming in to cr2 and 12.5v coming out of cr4.

I'm going to reflow connector pins...

One other detail is that it might not be temp. related. Had the gane one for 15 mins taking readings, rebooted and it came up fine. Worked for about a minute, then s&t locked with the green led on again.

Other times the led stays lite the whole time at boot.

What would cause the led to lite again after successful boot?

I guess the 6808 might be starting to flake out. I have a spare 6800 chip, could I use it in this board?

Reflowed pins on all sockets. No loose chips in the sockets. Swapped U11 with one from another board. No change.

So in the game the S&T either locks on led lit (no flicker or flashes) at boot, or occasionally boots successfully. When it boots successfully it works for about a minute then locks up, led on.

I tested the board on the bench using a pc power supply, 12v to J1 p10. Boots fine everytime, getting 4 flashes. Is that normal if the 8912 isn't present (to only see four flashes)?

I don't have a speaker connected so can't tell if it's actually producing sounds though. Also, pressing self-test never causes the board to reboot at the end of the sequence like it normally would when in the game and working.

Put board back in game, boots the first time, but not getting sounds anymore, just speech and noises. After first boot locks up and the led stays lit upon subsequent reboots. Usually if I disconnect it and wait a few minutes it will boot again the next time (once).

Possible bad u11 causing no sounds? Could swap it back again.

Any ideas?

Okay, swapped out the 6808 ic and my s&t is back in business.

So I moved the jumpers from 'N' to 'M' and 'DD' to 'CC' so I could take the preamp volume pots out of the equation and use the digital vol. controls via audit settings #20 (sound) and #21 (speech) instead. It was a bit tedious to keep going back into these audit settings to make adjustments but I eventually settled on sound = 14 and speech = 6. These values make the sound nice and present and the speech as clear as I can get it before the compression 'hiss' starts to be audible. The speech really starts to distort above 8 or so as well.

The speech is still lacking in highs but I understand that this is an artifact of the sound board implementation intended to avoid harsh compression artifacts.

At some point I may play around with R9 which controls the speech timing to see if I can raise its frequencies slightly. I tested my R9 and it's reading 137 Ohms.

I'll post here if I do the R9 experiment in the future.

Thanks for the help guys!