Sound Blaster 2.0 - No Digitized Sound

[Hambone]

In the midst of building a 386 system from some extra parts, I ran into a wall with sound. I'm having an issue with my Sound Blaster 2.0 (CT-1350B) not correctly outputting the 8-bit digitized audio. The FM synthesis works beautifully, but the digital audio is almost non-existent; quiet and so full of static that nothing is really heard even at full volume.
I have tried different IRQ's and both 220 and 240 I/O's. I have also tried different slots in my system. I remember having this issue with the same card when installed into an IBM 5170, and eventually I got it working, but I cannot remember what the fix was.
I've tried the v1.0 drivers in addition to the latest one's from 1993.
Last two things. I also disabled the joystick port in case that was conflicting. And i've tried the DMACTL jumper at the high and low settings with no noticeable difference.


I apologize if this topic has been covered. Believe me when I tell you that I used the search function on the forums, as well as google. A few hours burned in total.

 

[Trixter]

You need to clarify if you can hear any digitized sound at all, or none. If you can hear something, but it's faint or weak, then you may need to replace the capacitors on your Sound Blaster. If you hear nothing at all, set the board to port 220, IRQ 7, DMA 1, and ensure you're not relying on any parallel-port devices (no printers, no zip disks, no xircom network adapters, etc.). Then configure your software for those settings and try again. Those are the most compatible settings.

You may also need to set a BLASTER environment variable. For your card, I believe:

SET BLASTER=A220 I7 D1 T4

...will work in your autoexec.bat.

What software are you trying?

 

[Hambone]

You need to clarify if you can hear any digitized sound at all, or none. If you can hear something, but it's faint or weak, then you may need to replace the capacitors on your Sound Blaster. If you hear nothing at all, set the board to port 220, IRQ 7, DMA 1, and ensure you're not relying on any parallel-port devices (no printers, no zip disks, no xircom network adapters, etc.). Then configure your software for those settings and try again. Those are the most compatible settings.

The sound I hear with FM audio is crystal clear, perfect. It even works within games and Windows 3.11. The digital audio is BARELY audiable at full volume (both the card's potentiometer and external speakers are volume up), and when i say audible, i mean, if I didn't know the SB's test program was saying "Hello there, hello there," I wouldn't have been able to understand it. LOT's of static, very little volume.

You may also need to set a BLASTER environment variable. For your card, I believe:

Yes, I've set the environment each time I tried a different IRQ, I/O, or drivers. I've relied on the installers to set variables in my autoexec.bat & config.sys files. Attached to this post are my autoexec.bat and config.sys files as they stand.

What software are you trying?

Everything. From the test-sbc.exe and diagnose.exe programs included with the SB drivers, to Windows 3.11 sounds, the Vector Demo, Epic Pinball, Wolfenstein 3D, etc.

It baffles me that the FM synth works PERFECTLY, yet the digital audio doesn't. This makes me think there's a soft issue. I'm no expert, but a good look at the board's components and circuits doesn't yield any red flag, obvious problems; no bridged solder, no bulging caps, no corrosion.

As stated in the original post, I had this issue in a PC/AT, and got it solved, but that was a few years ago. Also, I had this card working in a PS/2 with zero issues. Sadly I don't have that machine anymore to test with.
I do have a Gateway Pentium II machine with ISA slots, but that'd be a day's worth of setting it up just to test the SB.

Last thing I gotta say is: Holy B@LL5! Is that THEE Trixter? The one from the demo scene?

 

[Trixter]

It baffles me that the FM synth works PERFECTLY, yet the digital audio doesn't. This makes me think there's a soft issue. I'm no expert, but a good look at the board's components and circuits doesn't yield any red flag, obvious problems; no bridged solder, no bulging caps, no corrosion.

I don't think it's a software issue; those manifest as no sound at all, or looping/skippng sound. The fact that you hear something, but it's scratchy, leads me to think it's a hardware defect.

The Sound Blaster 2 requires -5V, and some PSUs don't provide that. Since you're using an older PSU, it is supposed to provide that, but it's possible your PSU isn't providing -5V any more. (More info on this: https://www.philscomputerlab.com/voltage-blaster--5v.html ). I would use a multimeter to ensure your -5V is working from the PSU, and is also present on the ISA bus.

As stated in the original post, I had this issue in a PC/AT, and got it solved, but that was a few years ago. Also, I had this card working in a PS/2 with zero issues. Sadly I don't have that machine anymore to test with.
I do have a Gateway Pentium II machine with ISA slots, but that'd be a day's worth of setting it up just to test the SB.

Might be worth it to verify the card is ok, as your anecdotal evidence suggests, to me anyway, your testing system may have a faulty -5V issue.

Last thing I gotta say is: Holy B@LL5! Is that THEE Trixter? The one from the demo scene?

 

[GiGaBiTe]

I would say to recap the card just as a precaution. The caps on that card are 30+ years old.

There is another member on this forum with one of those cards that I recapped. Most of the caps on it were VERY bad IIRC. They don't need to show outward signs of failure to be bad. The ones on the card I worked on went high ESR and/or electrically leaky and effectively acted more like resistors than capacitors.

His card worked, but it was noisy before recapping it. It didn't have static on it like you describe, so you may have more wrong with the card, like a bad IC somewhere. I'd start with the low hanging fruit first though, recapping is easy.

 

[Hambone]

Okay guys so to recap (heheh) I dropped the SB into a Pentium II Gateway w/Windows 98, a known good system. using the drivers built in, tested the card with windows sounds, and DOOM. Same results as in my 386 and AT. No digital sound, or rather, scratchy garbled junk thats quiet. The FM works in this system too, although it is a bit scratchy. (im trying not to over think it, i know older stuff isnt so "cut and clear" even when new)

I would say to recap the card just as a precaution. The caps on that card are 30+ years old.

There is another member on this forum with one of those cards that I recapped. Most of the caps on it were VERY bad IIRC. They don't need to show outward signs of failure to be bad. The ones on the card I worked on went high ESR and/or electrically leaky and effectively acted more like resistors than capacitors.

His card worked, but it was noisy before recapping it. It didn't have static on it like you describe, so you may have more wrong with the card, like a bad IC somewhere. I'd start with the low hanging fruit first though, recapping is easy.

Im a bit timid to solder on this board, seeing how valuable they are nowadays! I did recently replace some leaky VARTA batteries with pin headers to connect w/e batteries. So my confidence has increased!
Having a bit of a hobby with old stereo gear, yes, im familiar with leaky caps. Seems to be the bane of the electronic world.

 

[Hambone]

I would say to recap the card just as a precaution. The caps on that card are 30+ years old.

There is another member on this forum with one of those cards that I recapped. Most of the caps on it were VERY bad IIRC. They don't need to show outward signs of failure to be bad. The ones on the card I worked on went high ESR and/or electrically leaky and effectively acted more like resistors than capacitors.

His card worked, but it was noisy before recapping it. It didn't have static on it like you describe, so you may have more wrong with the card, like a bad IC somewhere. I'd start with the low hanging fruit first though, recapping is easy.

A package from Digi-Key arrived at my house today! I went through and recap'd the board very meticulously. Upon reinstalling the card in a 486 DOS machine, i still have the same problem The FM sound plays perfectly, but the digitized is still broke. Im just glad the thing still works, considering this was my first recap of.... anything!

One minor thing i overlooked, some of the caps were larger in dimension than the originals, however they were rated the same. These are the big green 100uF 16v caps.

 

[GiGaBiTe]

Capacitor size isn't terribly critical, as long as it's close enough in characteristics to the old caps, and it fits. The first is easy because 30+ years of capacitor technology improvements. The latter can be problematic for obsolete package types, but that's not an issue here.

Other low hanging fruit you can try is to test all of the other passive components on the card (resistors, diodes, tantalum/ceramic capacitors and transistors.) This is very tedious, but random passive components being faulty can cause issues.

You may also try to resolder the legs on the PLCC ASICs, just in case there might be a microscopic break that can't be seen. Testing continuity on the legs to traces they go to is also a good idea.

If you're fairly certain that you've tested what's possible to test, you may have a failed IC or ASIC. This isn't uncommon, lots of old chips are dying from electron migration, oxidation due to package failure, original manufacturing defects rearing their head after decades, etc.

But all of this assumes you can trust the system it's installed in. You said that you haven't tried it in another system in awhile, you need to do that. Put it in another ISA system and try running digital audio on it to see what happens. IIRC, these early SB cards don't like ISA bus speeds too high, or host CPUs too fast. You may have problems with the card on something faster than a Pentium.