The I/O base address on that card is jumper selectable using the two IOS jumpers next to the big CT2504 chips. Make sure IOS0 is closed, because with IOS0 open, the card is at address 260 or 280. Both of these addresses are uncommon for sound blaster cards. You get the default soundblaster address at 220 with IOS0 and IOS1 both closed, and the alternate soundblaster address (240) with IOS0 closed and IOS1 open. Most SB software should recognize the card at either 220 or 240.

Hi @heretyk.
Apparently, your Adlib Gold problem is referenced in the Microsoft Q132114 article. You are trying to detect a card with the driver removed without having restarted Windows. To resolve it, you have to remove manually the driver in the device manager of the card, the game port and the MPU-401 device. Restart windows. and run the Add New Hardware Wizard again. Now it should be detected correctly.
But if the problem persist, has the card is configurable by jumpers, make sure everything is OK with them and note the value for the base address, IRQs and DMAs. If there is no conflict Windows should detect it, so the automatic settings check box should be checked.

Edit: I just check the CT2800, it seems the only thing the jumpers can do is to configure the IDE port and IRQ 10, 11, 12 & 15 for it. The card is PnP, try to remove the IRQ jumper (put it half connected).

Datadrainer wrote on 2021-11-27, 20:08:

Edit: I just check the CT2800, it seems the only thing the jumpers can do is to configure the IDE port and IRQ 10, 11, 12 & 15. The card is PnP, try to remove the IRQ jumper (put it half connected).

That's not true. The OP is right: the CT2800 is a jumper-configured SB16-compatible card. IRQ and DMA settings are always configured using DIAGNOSE.EXE on non-PnP SB16 cards, but the base address is set by jumper IOS0 and IOS1 next to the main chip on the card. See https://wiki.preterhuman.net/Creative_Labs_So … ibra_16S_CT2800 for reference.

mkarcher wrote on 2021-11-27, 20:26:

Datadrainer wrote on 2021-11-27, 20:08:

Edit: I just check the CT2800, it seems the only thing the jumpers can do is to configure the IDE port and IRQ 10, 11, 12 & 15. The card is PnP, try to remove the IRQ jumper (put it half connected).

That's not true. The OP is right: the CT2800 is a jumper-configured SB16-compatible card. IRQ and DMA settings are always configured using DIAGNOSE.EXE on non-PnP SB16 cards, but the base address is set by jumper IOS0 and IOS1 next to the main chip on the card. See https://wiki.preterhuman.net/Creative_Labs_So … ibra_16S_CT2800 for reference.

Indeed, I missed the two jumpers. I know them, I have some Vibras and 16 that are software configurable, but I rarely use them. I agree the jumpers must be set for a 220 or 240 base address.

To answer the OPs question I would not throw away a card if you at least see it during post. (You can see it right after the HDDs are recognized and yeah you can use pause to pause it.)

And throwing in a picture of the card is always a good idea. At least it makes the gentlemen here go "hummmm" and if they do it in a synchronized way it might solve the problem.

Joakim wrote on 2021-11-27, 21:49:

To answer the OPs question I would not throw away a card if you at least see it during post. (You can see it right after the HDDs are recognized and yeah you can use pause to pause it.)

You only see PnP cards during POST. The CT2800 is not a PnP card and will not be displayed during POST even if it is working fine.

So to help @heretyk and to resume what needs to be done:
1) Check the base address is set to 220 or 240 with the jumpers. If not, set it correctly. Disable the IDE port if not required so its IRQ can be used by anything else.
2) Start Windows
3) Remove the drivers as I have mentioned earlier
4) Make a hard reboot
5) Install the driver again with "Add New Hardware Wizard". Use the latest official driver for the card, but the one with Windows should work fine. If using an external driver, you can have a problem with the driver of the game port, just force the installation.

Then it should be detected correctly and after a new reboot, should work fine if the resources can be acquired.

Thanks for all advice. I cleaned connectors and checked various jumper settings but again this card fails to work properly under windows or DOS in both of my computers.

I can only guess that sound and music generators reside in separate chipsets so this card still works partially even though one chipset is faulty.

Indeed. From your photo everything looks good on the card itself. Every components is here, there is no cut traces or bulging caps. The base address jumpers are directly connected to the TCQ chip and I don't see any visual problem here. I'm maybe saying something wrong but the OPL3 is using address 388h inAdlib Gold, and 220h or 240h and 388h for Sound Blaster cards. You say music is played. So it means 388h can be used by the program. And as the card is detected as a Adlib Gold, I think the problem is still related to the base address. Which is also surely used to detect and initialize the card. Is there something bad in the TCQ? Maybe...
Check the traces on the back to see if everything is okay.
The card can be sold for parts at a low price instead of being trashed. If someone is interested, you can have a part of what you paid for back.

I had a card like this, same behavior. It ended up to be the 46.5xxxMhz Osc. I exchanged it and it worked. You cannot get that crystal I think, It was only used by Creative. I got mine from a totally wrecked card.

Marmes wrote on 2021-11-28, 15:27:

I had a card like this, same behavior. It ended up to be the 46.5xxxMhz Osc. I exchanged it and it worked. You cannot get that crystal I think, It was only used by Creative. I got mine from a totally wrecked card.

Interesting. I would have suspected the 24MHz oscillator first. The 46.5xxx is for the 16-bit Sigma/Delta ADC/DAC, I think. The 24MHz oscillator is likely for the 12MHz clock of the 8051 "DSP" that controls the sound playback/record logic. Detecting a soundblaster card basically tests that the DSP responds as expected, so I would have guessed the 24MHz to be enough for successful detection.

Marmes wrote on 2021-11-28, 15:27:

I would have suspected the 24MHz oscillator first.

If I had slight chance that replacing 24MHz crystal will solve the problem I would take a risk. These crystals cost nothing and are widely available.

mkarcher wrote on 2021-11-28, 19:13:

Marmes wrote on 2021-11-28, 15:27:

I had a card like this, same behavior. It ended up to be the 46.5xxxMhz Osc. I exchanged it and it worked. You cannot get that crystal I think, It was only used by Creative. I got mine from a totally wrecked card.

Interesting. I would have suspected the 24MHz oscillator first. The 46.5xxx is for the 16-bit Sigma/Delta ADC/DAC, I think. The 24MHz oscillator is likely for the 12MHz clock of the 8051 "DSP" that controls the sound playback/record logic. Detecting a soundblaster card basically tests that the DSP responds as expected, so I would have guessed the 24MHz to be enough for successful detection.

Thank you very much!
I've got ct2800 with exactly the same problem. Got it working by changing the 24MHz oscillator. Took one off from old vlb io card.