audio issues with these cards can have other causes as well, particularily when putting them into old VIA chipset machines and such. heck, keep in mind that some games are simply broken or need a very specific card to work.

i once removed the caps in that spot on a 2zs and they all measured well out of spec, but that being said, i did not bother removing any other capacitors. bad batches or not, it seems to me that using small capacitance GP caps in that area was a bad cost cutting measure in the first place as they let through excessive ripple. somehow the older ISA cards tend to still work even with these cheap caps though, perhaps the hardware there was just much less sensitive to this.

auron wrote on 2021-04-07, 00:54:

audio issues with these cards can have other causes as well, particularily when putting them into old VIA chipset machines and such. heck, keep in mind that some games are simply broken or need a very specific card to work.

i once removed the caps in that spot on a 2zs and they all measured well out of spec, but that being said, i did not bother removing any other capacitors. bad batches or not, it seems to me that using small capacitance GP caps in that area was a bad cost cutting measure in the first place as they let through excessive ripple. somehow the older ISA cards tend to still work even with these cheap caps though, perhaps the hardware there was just much less sensitive to this.

Wanted to resurrect this thread for a bit to add my two cents, Audigy 1/2/2ZS cards are the only ones I'd use in a machine that has an nForce chipset of any kind. I've had problems with X-Fi cards in such systems for ages. when I had a dual Opteron 285 machine as my daily some years back, it was the only card that'd work and not stutter all over the place or just lock up. Other issues on other machines with these cards have been with the audio not going crackly, but going robot-like, and I always chalked that down to resource issues. on the one machine I have these cards in, I don't have these issues, but like what keenmaster486 said earlier on in this thread, these do get used a lot at radio stations in aged machines running audio backend applications, and I do help out with an internet radio station that uses an Audigy 4 (Audigy SE?) alongside the onboard audio to do some external audio stuff with respect to the microphone due to the way the software being used is designed. Seeing as the 1/2/2ZS and even the X-Fi cards are probably having issues with these, it might be an idea for me to go and do these up for the other card and keep that problem out of sight and mind...

Should also point out that the card's main power cap to the left of the area of the replaced caps in the previous images in this thread, is seemingly something else to replace if these are getting overhauled as well. That's just a 16v 100uF cap but I've seen people replace them with up to a 470uF like is with the X-Fi cards, and I've also seen audiophiles go super nuts doing up X-Fi cards with 1500uF caps going in and Elna silmic caps in weird places on those cards as well. I'm not saying to take it that far as that's absolutely silly in my book, but if the voltage regs deserve lower ESR and a bit more capacitance, one surmises a 220uF wouldn't hurt to drop in at that one cap there I don't think. Creative always had a knack for getting away with whatever capacitors they had on hand, and I've seen G-Luxon, Wincap and Jamicon caps being used in spots. the Jamicon caps out of all of the cheap ones used seem to be the ones that held up the best and also seems to be the supplier Creative always went with back in the SB16 days from the ones I've had the pleasure of owning. I have one Audigy out of the three current that doesn't seem to work at all past POST, you drop it in and the thing won't let you get past mid-boot sequence, and this one has Wincap everywhere. The other two are Jamicon with a couple Wincap and G-Luxon mixed in randomly, but, only around the areas of the TAD/CD/AUX/SPDIF connectors.

Also wanted to generally ask if these caps getting replaced might also fix certain issues I've been seeing with my own advanced-age Audigy cards, such as in certain games having freezing happening where it'd just let out a high tone and then continue on like nothing happened. Sometimes I see the same thing happen playing music as well, but it's not as common (i.e. iTunes 11, playing audio of any kind, locally or off the owntone share.)
Before going and doing the search which led me to here, I was under the guise it might be a PCI bus bandwidth or resource issue, as changing the card out for a Turtle Beach Santa Cruz has resolved that one specific issue, but now I'm concerned I'm going to have the same issues on at least one more machine the card model is in, and thus, more questions than answers. As anyone that did these replacements have likely had their card(s) in use for a reasonable amount of time, I figured now is an okay time to also get some updates to see if the cards are still working properly long-term or if any other problems have arose.

for the old ISA cards they did use japanese caps quite a bit, got an SB16 CT2290 with mostly chemicon SMG and an AWE64 with nichicon VR on it, though especially towards the sblive era they were mixing in a lot of stuff like wincap as you said. with x-fi i've not seen any japanese caps at all, except the SMDs on the titanium HD.

replacing the caps mentioned in this thread is definitely worth a shot because as i mentioned, i verified the ones on the 2 ZS being bad with an ESR meter. as far as a complete recap goes, i wouldn't undertake that without a good desoldering station.

Those 1117 type regulators actually have a minimum ESR requirement, but it's hard to violate that with only 47uF. Some manufacturers of those regulators even recommend 100uF caps. In that case, I'd suggest just a long-life series instead of low ESR, although even cheap 105°C caps like Rubycon PX would be fine. Lelon or Elite would probably be fine as well if you're really cheap (I actually use Lelon caps sometimes, but I have no experience with Elite).

auron wrote on 2022-12-28, 23:38:

for the old ISA cards they did use japanese caps quite a bit, got an SB16 CT2290 with mostly chemicon SMG and an AWE64 with nichicon VR on it, though especially towards the sblive era they were mixing in a lot of stuff like wincap as you said. with x-fi i've not seen any japanese caps at all, except the SMDs on the titanium HD.

replacing the caps mentioned in this thread is definitely worth a shot because as i mentioned, i verified the ones on the 2 ZS being bad with an ESR meter. as far as a complete recap goes, i wouldn't undertake that without a good desoldering station.

my CT2290 in front of me right now has a mixture of Nippon, Elgen, Jamicon and Wincap from factory. it also has two matching panasonic/matsushita 16v 470s I shoved on because one of the originals incurred a dent during the move and measured short. Also had to replace the tant at C55 above the ISA notch as that got knocked off.

so by at least February 1995 (so says the board datecode) they were mixing brands based on what they had, and I wouldn't be surprised if I went and grabbed my CT1740 or CT1750 and found they had the same thing going on further back.
That being said, what cap brands are on the board could possibly be a side effect of what factory they came out of for different regions as well, after their popularity exploded of course. but as far as with the Audigy cards, I'm fairly certain by that point they had a factory in Taiwan or mainland China manufacturing them, and it's almost certain that the caps were acquired from a few suppliers of decent-ish Jamicon/Wincap/etc. caps that worked and would probably work for a few years, well within the life cycle of your average prebuilt home PC of the era, given so many of them appeared in HP and Dell PCs at that time. All to pinch pennies, of course, but even at that point, Creative was really hurting to stay relevant in the modern market of the time when nearly all new motherboards had some sort of audio chipset on them standard, and very few were being picky about it. If it didn't have onboard sound, it was usually an OEM machine and usually a Dell or HP and required that silly white front-panel connector.
It's possible that due to needing to fulfill contracts for the OEMs, it made more sense to put on whatever caps as well to save a few cents. This is also likely why up until the 2010s when Creative started getting proper-serious again, the main cap supplier was almost always Jamicon with anything else to fill in the gaps.
of course this is all speculation and should be taken with the biggest grain of rock salt you can spare.

lti wrote on 2023-01-01, 00:39:

Those 1117 type regulators actually have a minimum ESR requirement, but it's hard to violate that with only 47uF. Some manufacturers of those regulators even recommend 100uF caps.

Yes, you are absolutly right. The 1117 has some plot of ESR versus capacitance.
The lower the ESR, the larger the capacitance should be.
By datasheet.

In practice, 1117 is often used even with MLCC capacitors in millions of units.

Audio circuitry has its own characteristics, and after 78/7905 you really should not use low-ECP capacitors, because this leads to an increase in noise.
Capacitor before of the reguliator (which are on the PSU bus), it is better to put something exclusively lowESR, so that the garbage from the entire PC does not go into the audio card.

PS. The oscilloscope says OK.

Just a thought - is it possible that these caps go bad because of heat from the close-by VRMs, instead of being a low capacitance issue to begin with?

Did this fix with 16v 47uF Nichicons, works perfect now. Thanks!

This is a pretty common problem with Audigy 2 and X-Fi cards, especially around the on board power regulation. I usually have the best luck finding the correct physical dimensions with polymer caps from Kemet on Mouser.

8bitbubsy wrote on 2024-01-19, 17:09:

Just a thought - is it possible that these caps go bad because of heat from the close-by VRMs, instead of being a low capacitance issue to begin with?

The problem is, at least on the X-Fi cards, that Creative used caps that can't handle the ripple from the buck converter output. It shortens their life. If you pull the datasheet on the TI buck converter chip, you'll see they say a low ESR cap is mandatory.

Here's the reference design for the buck regulator chip used on the X-Fi cards.

Creative did not follow this guidance.

Nice to see the guidance calls for resistors rated in kilowatts... Shouldn't be any issue with de-rated heating. 😂😂😂

I guess 'w' here means ohms!

Also was thinking I've seen high value but pretty slim and short poly caps: and then the guidance actually mentions them.

8bitbubsy wrote on 2024-01-19, 17:09:

Just a thought - is it possible that these caps go bad because of heat from the close-by VRMs, instead of being a low capacitance issue to begin with?

I think that's what is happening on the Audigy 2, along with the caps just being cheap ultra-miniature 85°C parts from poor manufacturers. It only used linear regulators, so it didn't have the same problem as the X-Fi.

I don't think the capacitance was too low. There were lots of mods back when the cards were new, and they didn't fix the crackle.

ChrisXF wrote on 2024-01-21, 02:55:

Nice to see the guidance calls for resistors rated in kilowatts... Shouldn't be any issue with de-rated heating. 😂😂😂

And "mF" is not for milifarads, but for microfarads (μF, uF), unless "m" is an typo instead of "n" for nanofarads (nF) 😀

Yeah maybe some dodgy type setting or less than fantastic OCR there!