Ideally I’d harvest one of these capacitors that are right next to the crystal on an old isa modem if that would work.

nope, those are in pico range for crystal, TOYOCOM is a generator. C29 from that modem looks like 100nF and should be fine.

rasz_pl wrote on 2024-12-23, 11:00:

nope, those are in pico range for crystal, TOYOCOM is a generator. C29 from that modem looks like 100nF and should be fine.

Ok, thanks. Did you assess the capacitance by the color or the location ?

Location. C67 68 go under crystal so those are tiny external load caps. C29 C65 are soldered to the big copper plane , most likely ground = decoupling for something.
I would guess that on your controller caps are called:
TC = tantalum, maybe ~10uf?
BC = bulk capacity, maybe 1uf? or anything above 100nF
C= everything else

You can just check where missing cap is wired to, if ground and +5V then its 100nF or maybe even 10nF and probably in parallel with BC27.
If between ground and output then its a load capacitor after all (rare) and you can use C67 C68 like you intended (they might be too big) or just leave it missing, its not that critical (mostly EMC I think? controls rise time speed).

Thanks for the details. I really appreciated it. I lost patience yesterday before you first response and used the C68 capacitor because it seemed low risk and the iron was still hot after doing the tantalum cap. The card is working ok so far. Just needs a BIOS upgrade.

The repair that I did on the Rev B1 board went great.

Now I got a Rev B2 board thanks to Beatoven. The board looks identical to the Rev B1.
https://theretroweb.com/expansioncards/s/prom … se-eide4030plus

I added jumpers, replaced the sketchy 74F245 near the PDC20230C chip with a 74LS245 , replaced TC8 and TC9 that were both shorting. I had them in the same orientation, with + to +12V for TC8 and + to ground on TC9.

Last time it looked like C6 was missing, so I replaced it. This time C6 didn't look like it was missing, so I din't put anything there.

I applied power and BLAM!!! TC8 blew molten metal about 8 feet, TC9 shorted again.

I cleaned it up and checked the traces and it all looks good. I don't see any other shorts.

Anything else I should check before turning any more tantalum capacitors into fireworks?

Reviewing a picture of a newer Rev B2 card I can see--

• C6 is not populated on the Rev B2
• TC8 and TC9 are replaced by electrolytic capacitors
• Newer 1996 BIOS

Does anyone have a Rev B2 with the newer BIOS or the specs for the Electrolytic Capacitors?

So I recapped, cleaned it up, plugged it in and it worked!

Well for about 60 seconds and then the TC8 tanty once again popped, sending glowing sparks of metal across the room.

The Blue cap is what was replaced. The Yellow is the replacement. They look the same to be.

The 74LS245 that I replaced doesn't seem to have any shorts and isn't even connected to 12v or -12v.

At rest, Ground is not connected to +12v or -12v

The answer has to be that one of the IC's on board is pulling a lot of power on +12v , right?

Edit - I found something -- +12v doesn't go to most grounds, but it does go to B31 , and B31 goes to ....
scratch that, +12V goes to B31 on both cards--
Scratch that, B10 goes to B31, not B09 which is +12v. This one is driving me crazy--

Traced out more of the card, hopefully correctly this time--

The +12v and -12v only go to the W83758 chip.

Looks like the W83758 chip does serial and game ports and the third AT IDE header.

I cleaned out some dirt from the serial port. Looks like something bad might have happened there at some point. I don't see any shorts there now.

I guess I could sacrifice another set of caps to see if cleaning the 9 pin serial port fixed things but I suspect I need to remove the W83758 chip because something probably blasted it.

Suggestions?

Are you sure you don't have any shorts? I know you mentioned that a few posts back, but... exploding caps?

Do you have a thermal camera?

CkRtech wrote on 2025-08-01, 22:28:

Are you sure you don't have any shorts? I know you mentioned that a few posts back, but... exploding caps?

Do you have a thermal camera?

I’ve checked for shorts every way I know how. At rest there’s no short. I don’t have any thermal imaging.

When I got the board, both TC8 and TC9 were shorting. I replaced them and after appling power TC8 popped open and TC9 was failed shorting and the power supply shut itself off. I replaced them again and the board worked for a minute before TC8 popped open. TC9 didn’t fail that time but I shut the power off myself.

Tantalum caps can pop from a voltage spike, a polarity reversal, or ripple current, says the internet.

The board was pulled from a system where someone had swapped the P8 & P9 connector. The W83758 chip could be damaged. I could try lifting the +12v and the -12v from the W83758 chip to take it out of the equation, but I’m worried that I might not be able to reverse that.

I’ll try some 10uf 25v caps before I start breaking pins. Looks like the later rev of the board has those.

I got some low ESR 10uF 50v caps from the Panasonic store on Amazon.

They fit. The card didn't blow up yet. It booted Win98 in compatibility mode and only the 286 was getting warm.

So maybe I just had some marginal tantalum caps in my tool box? Maybe?

I'd expected the controller firmware to prompt me with a RAM count on the board during post, so it looks like there's farther to go on this one.

At least it stopped going off like a fire cracker.

Edit -- Reseated the BIOS ROM and it is prompting correctly and counting RAM. Yeah! Too bad it wants to do LBA geometry instead on the ECHS geometry that the motherboard was doing. Always one more thing---

Glad you were able to get the card working! I wonder if the exploding tantalums were a separate, possibly unrelated, issue? Unless it was done a cost saving measure, the switch to electrolytics on later versions could mean this was a known problem.

Thanks for doing the ROM dumps - I was able flash a blank chip and upgrade the BIOS on my older version of this card.

When P8 and P9 were reversed, it would have sent -5v on +12v which would reverse the polarity & damage TC8 and +5v on -12v which would do the same to TC9 . So that was an issue. But it seems that there was also another issue that was causing them to blow afterward on this particular card. Because the other card that I found also had a damaged TC8, seems like the 16v tandalums might not have been quite enough to be power decoupling capacitorson the 12v lines on the EIDE4030 if there was a power spike or “oscillation” (which is just something I read, not something I fully understand here).

Also, if anyove has the 1.03 bios or newer for the card, please upload it.