Recent card i picked up, might possibly not be on the list yet/documented properly as i couldn't really find alot of info about it. Kind of an oddball though.

Edit;
Added the enclosure pictures/other end of the connector aswell, picture reference from here;
Re: External Creative 2x CD-Rom drive?

I've never seen this one before! An oddball indeed! Nice connector. Will add this card too of course. Thanks for the submission!

S95Sedan wrote on 2022-10-04, 14:16:

Recent card i picked up, might possibly not be on the list yet/documented properly as i couldn't really find alot of info about […]
Show full quote

Recent card i picked up, might possibly not be on the list yet/documented properly as i couldn't really find alot of info about it. Kind of an oddball though.

Edit;
Added the enclosure pictures/other end of the connector aswell, picture reference from here;
Re: External Creative 2x CD-Rom drive?

WOW! That is one cool piece of kit! Never seen or heard of that one before and I've collected a lot of SoundBlaster stuff.

What is the best sound blaster 16?

harris286 wrote on 2022-11-20, 17:25:

What is the best sound blaster 16?

It is the MK8330.

Impossible to answer this question because it's subjective, so you'll get silly answers like the one above.

P.S it's the CT2230

This sheet should be updated that Vibra16-Pro CT2502 chip doesn't have hissing/ringing and clipping bug like other CT2501, CT2504, CT2505 and CT25011 Vibra chips. People on vogons mentioned this several times and also I check it and didn't notice this problem on CT2502. Someone should check also ViBRA16CL CT2508 chip.

HiroSK wrote on 2023-04-20, 15:44:

This sheet should be updated that Vibra16-Pro CT2502 chip doesn't have hissing/ringing and clipping bug like other CT2501, CT2504, CT2505 and CT25011 Vibra chips. People on vogons mentioned this several times and also I check it and didn't notice this problem on CT2502. Someone should check also ViBRA16CL CT2508 chip.

What games did you use to test the ringing.

I tested my fifth Vibra16-Pro CT2502 based card, CT3600 this time, and the first one with CT2502 made in France. Like the others (CT2502 made in Korea or Malta), it does not have the clipping bug (tested with Skyroads like I did with the other cards). I too would like to know if there is any good way to test the ringing bug. I don't think I've heard it on any of the tested CT2502 cards so far.

aitotat wrote on 2023-04-22, 09:00:

I tested my fifth Vibra16-Pro CT2502 based card, CT3600 this time, and the first one with CT2502 made in France. Like the others (CT2502 made in Korea or Malta), it does not have the clipping bug (tested with Skyroads like I did with the other cards). I too would like to know if there is any good way to test the ringing bug. I don't think I've heard it on any of the tested CT2502 cards so far.

James-F has Tyrian and Wolf3D as examples of the ringing bug in his SB16 sample collection. The archive containing the sample files is linked in this post of his.

The ringing effect is a bit difficult to hear without either dialing up the speaker volume or using headphones.

Request: Add column for Floating Op-Amps too.
I know the CT1600 SB Pro 2.0, CT2230, CT3900, first revision of the CT2760, most of the CT174x series and almost every ViBRA 16 card has at least one floating section of a quad amp.
Somewhere I posted some examples of cards with floating amps.

As a personal aside, I still don't really think the CT1701/1703 are solely to blame for self-noise. I maintain it's the dern floating op-amps, which are often located close to the CT1701/1703 chip, so they likely share the same power planes. The op-amps themselves are also suspect too, especially if they are MC/KA3403 (these are practically LM324s with better crossover distortion figures), which are very noisy amps. The TL074 amps have a different noise issue: they have relatively high voltage noise (on the order of 22uV/Hz) yet very low current noise, but they also seem to be more vulnerable to EMI/RFI than your typical bipolar amp.

And before anyone tries to 'correct' me in good faith, has anybody noticed that the cards that use the 'better' mixer/dac chips tend to have fewer op-amps in general (and therefore fewer things to pick up noise)? They may even have some floating sections, but since there is much less stuff to go through, the resulting noise, assuming everything else being equal, will still be inherently lower than the cards with the more complex layouts, even if one were to transplant a "noisy" mixer/dac onto them.

Might as well add AWE64 Legacy to the list. For funsies.

Eep386 wrote on 2023-05-05, 17:45:

Request: Add column for Floating Op-Amps too.

This isn't meant to sound critical, but were you (or anyone else) able to get any pre/post-mod measurements to help better-define this issue and its implications, as applies to the affected SB cards?

A proper recording setup (one that isn't excessively noisy by itself) continues to elude me. I am envious of those who are able to do these professional-grade MDFourier charts and graphs. 🙁
All I have (functionally) remaining of my SB16 cards is my CT2760 AWE32, which has had two floating amps patched. Should be a simple matter to undo the patches though.

That said, I found my old Echo Digital MIA audio interface again. I'll see if I can find something to deal with the grounding loops... Scratch the MIA, Linux naturally doesn't want to work with it... at least I got some control recordings made (nothing hooked to the line-in interface on the Realtek HD audio onboard sound).

Mother of derp! I mixed up the original and control samples... ARRRGH! Fixed 😅

OK, first results are in!! 😁
There may be more results to follow as I do more recordings. For starters, though, we'll go with a recording of the card playing no sound.

Methodology:
- Audio directly recorded from the Line-Out jack of the CT2760, going into the Line-In jack on my onboard audio.
- I had originally fixed the floating amps on the card, then carefully removed the fixes afterward to reconstruct the 'before' configuration
- All output volume levels set to 50% via Windows Me's Volume Control
- Computer case was closed between recordings
- CD audio cable hooked up
- Only the Line-In, Microphone and PC Speaker outputs were muted within same volume control
- I used Audacity as I couldn't get MDFourier working on Linux for the life of me... forgive me, I tried
- I honestly don't know what dB level my equipment records at. Likely commodity -10dB, or whatever it's called, as I'm using just the line input.

My particular setup:
- Motherboard is a Tekram P5V30-B4, with Pentium FDIV 90MHz, 64MB RAM, 512K cache, 3dfx Banshee 16MB PCI, and 40GB Barracuda IV hard drive running on a Promise Ultra/66 card, running Windows Me
- Sound card is a Creative Sound Blaster AWE32 CT2760, first revision
- I have replaced some of the MC3403 op-amps with TL074 and TL084
- Some of the caps on the TEA2025 output were changed, but are likely not involved in this test anyway as I am using the Line Output

The control recordings:
These are provided to show how noisy my lame-ass equipment is (and to show what to logically subtract from the other graphs).

Log Spectrum:

Linear Spectrum:

Recording: (zipped because my version of Audacity refuses to export to flac for some ridiculous reason)

With floating amps:
This is (hopefully!) representative of how the card would sound before the floating amps were fixed.

Log Spectrum:

Linear Spectrum:

Recording:

With floating amps fixed:
This is how the card performs after the floating amps were fixed. Notice that the highest peak went from around -69dB to -75dB after the fix, and the overall noise spectrum got much less spiky as well (for a lack of better term). 1000Hz looks like it went up a bit though - probably due to the wackiness of my recording setup. I'll re-record this after re-fixing the card and see if that smooths that wonkiness out.

Log Spectrum:

Linear Spectrum:

Recording:

Fixed amps, take 2!
Alright, this time around I cleaned up my mod wiring a bit, using shorter wires and not tapping grounds from where I ought not (read: from oscillators 🙃 ...)

Linear Spectrum:

Log Spectrum:

Recording:

Ooh, looks like I gained about a 1-2 dB average improvement from cleaning up the wiring. Some things merely fluctuated a bit, as expected (I really need better cabling!)
Ergo: it's important to fix design mistakes, but also to make sure you don't add mistakes in the process 😅

On a separate note, I am really seeing (and hearing) the effect of the dynamic brickwall filter. This is what I am getting for a spectral graph on Jazz Jackrabbit, SB16 setting.

Eep386 wrote on 2023-05-06, 17:46:

OK, first results are in!! 😁

Thank-you for gathering and sharing that information. That's a considerable effort!

The difference doesn't seem to be as great as I was thinking, so while I don't think I'll be modifying any of my own cards, it's certainly still an enhancement/fix worthy of consideration. I'm curious to get Tiido's take on things.