there is a bit of corrosion in the trace in the gap between the 8bit isa pins, but to the left (its probably not completely broken). How does the back look?
you could try something like doom and select sound blaster option for the music. This should be using the sb io port (i.e 220) instead of 388. The adlib option should use 388.
are you running any kind of mixer from the driver package? maybe the FM volume is off/low.
you might try unisound, that specific part number isnt mentioned but maybe it works. If it does unisound can set volumes as well.

Pickle wrote on 2025-09-25, 12:35:

there is a bit of corrosion in the trace in the gap between the 8bit isa pins, but to the left (its probably not completely broken). How does the back look?

Good eye, didn't see that. Scraped it to check, the copper trace underneath was pristine surprisingly. I looked through it with a magnifying glass, but I couldn't find anything wrong either. The other side was clear, too.

Pickle wrote on 2025-09-25, 12:35:

you could try something like doom and select sound blaster option for the music. This should be using the sb io port (i.e 220) instead of 388. The adlib option should use 388.

Haven't tried Doom yet, I have tried Adlib Tracker II, however, but it won't detect OPL3 on its own. Forcing it to 388 makes it start, but no sound either. Messing with OPL_latency in the settings did nothing.

Pickle wrote on 2025-09-25, 12:35:

are you running any kind of mixer from the driver package? maybe the FM volume is off/low.

No, I get the suspicion a mixer won't do anything since the chip isn't detected, but I'll check anyway.

Pickle wrote on 2025-09-25, 12:35:

you might try unisound, that specific part number isnt mentioned but maybe it works. If it does unisound can set volumes as well.

Didn't see it mentioned there, either. Guess I'll try tomorrow, since it's getting late here.

I have a card based on the 82C928-002 and also struggle with it. in a few games. Mostly I remember Doom would either do sound OR midi but never both.
In the end I found unisound worked better then the official drivers.

UNISOUND - Universal ISA PnP Sound Card Driver for DOS v0.82a

This was on a dos/Win3x machine, for which I could never find WIn3x drivers, but get basic sound from the drivers loaded in dos

Welp, UNISOUND turned out to be a total no-go. Doesn't detect anything whatsoever, no matter if I use that /NOPNP switch, set PNP OS option in the BIOS on or off, nothing changes. Even tried UNISOUND after having initialised the card with the JKnipperts driver (fixed the SB mode issue with that driver BTW), it still didn't see anything at all.

I've copied DOOM into the HDD, I'll try it tomorrow, along with Tyrian, since I love that game, and it has WSS support as well.

Two things I've found odd so far, PhilsComputerLab got a card with the same 82C929A chip, but a (probably) slightly newer AD1846JP codec, working fine. The other odd thing, my card seems to lack a 16 KHz crystal, while most I see online (including that card) have both 24.5KHz and 16KHz. Is that relevant in any way to FM synth working? Is there any real difference between AD1845 and AD1846?

If you have a multimeter, you could check if power is reaching the proper pin (also check its ground connection). You could also check whether there is continuity between the Opti chip and the FM chip on the expected pins

It is unfortunately possible that the FM chip has simply failed.

darry wrote on 2025-09-27, 00:28:

If you have a multimeter, you could check if power is reaching the proper pin (also check its ground connection). You could also check whether there is continuity between the Opti chip and the FM chip on the expected pins

It is unfortunately possible that the FM chip has simply failed.

This seems more and more likely, even though I had never even heard of that before. Checking it with the multimeter showed everything fine, but I know it doesn't work. Is it possible for the Opti chip to have failed on the FM part and for that to be the reason it can't communicate with the OPL3 chip? It might also be those tiny SMD capacitors, but that would result in the whole card being a dud. If only I had a hot air station, I would swap its OPL3 and DAC chips with my crappy ALS100 card, and see if either has FM. If it's the FM chip, it would result in the ALS100 not having FM, while the Opti would.
Another thing I noticed, Win98 didn't even use the 388 I/O with the "Basic Configuration 0000" in the Resources tab, but selecting "Basic Configuration 0001" did. The result, however, was Windows refusing to see the card, despite every other resource being exactly the same.

Maybe I'll end up using that ALS100 card, but I would prefer to get this working fully. I don't like that one because the sound itself is too muffled and lifeless, with plenty of noise on top of it. The good ALS100 is in a different system.

TechieDude wrote on 2025-09-27, 10:29:

This seems more and more likely, even though I had never even heard of that before. Checking it with the multimeter showed every […]
Show full quote

darry wrote on 2025-09-27, 00:28:

If you have a multimeter, you could check if power is reaching the proper pin (also check its ground connection). You could also check whether there is continuity between the Opti chip and the FM chip on the expected pins

It is unfortunately possible that the FM chip has simply failed.

This seems more and more likely, even though I had never even heard of that before. Checking it with the multimeter showed everything fine, but I know it doesn't work. Is it possible for the Opti chip to have failed on the FM part and for that to be the reason it can't communicate with the OPL3 chip? It might also be those tiny SMD capacitors, but that would resort in the whole card being a dud. If only I had a hot air station, I would swap its OPL3 and DAC chips with my crappy ALS100 card, and see if either has FM. If it's the FM chip, it would result in the ALS100 not having FM, while the Opti would.
Another thing I noticed, Win98 didn't even use the 388 I/O with the "Basic Configuration 0000" in the Resources tab, but selecting "Basic Configuration 0001" did. The result, however, was Windows refusing to see the card, despite every other resource being exactly the same.

Maybe I'll end up using that ALS100 card, but I would prefer to get this working fully. I don't like that one because the sound itself is too muffled and lifeless, with plenty of noise on top of it. The good ALS100 is in a different system.

I don't know enough about how an OPL3 physically interfaces with the ISA bus or with the Opti chip to provide any useful guidance beyond the basic stuff I already mentioned. Somebody here with much more extensive knowledge might be able to suggest some additional tests.

My uneducated hunch would be that an unbranded (and possibly marginally legal at the time) OPL3 clone chip made by an unknown foundry who knows where would be more likely to fail than the Opti chip, which was likely manufactured by a reputable foundry. Again, this is an uneducated hunch, backed only by the shared experiences of some people here and elsewhere having reported dead or damaged clone FM chips (and/or their clone DACs).

If I was in a similar situation, I would possibly try going through the data sheets for both the OPL3 and that Opti controller and testing for any unexpected logic levels (logic probe) or signals (oscilloscope) on the chips themselves and any clock sources (oscillators can fail). It would be an interesting learning experience for me, but possibly a useless one as, even if IDed a faulty component, my soldering skills would probably not be up to the task of unsoldering/soldering a surface mount chip.

You might try chip quik to remove without hot air. It’s a low melt temperature solder and just moving around the pins can keep it molten long enough to move it.

darry wrote on 2025-09-27, 11:12:

I don't know enough about how an OPL3 physically interfaces with the ISA bus or with the Opti chip to provide any useful guidance beyond the basic stuff I already mentioned. Somebody here with much more extensive knowledge might be able to suggest some additional tests.

My uneducated hunch would be that an unbranded (and possibly marginally legal at the time) OPL3 clone chip made by an unknown foundry who knows where would be more likely to fail than the Opti chip, which was likely manufactured by a reputable foundry. Again, this is an uneducated hunch, backed only by the shared experiences of some people here and elsewhere having reported dead or damaged clone FM chips (and/or their clone DACs).

If I was in a similar situation, I would possibly try going through the data sheets for both the OPL3 and that Opti controller and testing for any unexpected logic levels (logic probe) or signals (oscilloscope) on the chips themselves and any clock sources (oscillators can fail). It would be an interesting learning experience for me, but possibly a useless one as, even if IDed a faulty component, my soldering skills would probably not be up to the task of unsoldering/soldering a surface mount chip.

I didn't even know OPL3 chips could fail. I would expect them to be produced in the same foundry as the "real" ones, since they're identical. It would also make sense if said foundry produced various IC's for many different brands in China, Taiwan etc.
I don't think the oscillator has failed, since that would result in no sound at all from the card. I also don't have an oscilloscope to test that theory sadly. My multimeter does have a Hz function, which showed 14.3MHz, so the clock signal is OK, too.

Pickle wrote on 2025-09-27, 12:43:

You might try chip quik to remove without hot air. It’s a low melt temperature solder and just moving around the pins can keep it molten long enough to move it.

I've heard about it and seen it, but last I heard it's kinda expensive, and probably even more so now with the tariffs, since it's an American product, and I'm in Greece. I'll look for other low melt solder brands. Or maybe the time has come to get a hot air station