can you spot anything wrong with this serial/parallel card?
Posted by tutimes67@reddit | vintagecomputing | View on Reddit | 16 comments
when i plug this thing into one of the 8 bit ISA slots on my riser, it stops my pc from doing anything. i see the CPU speed display and LEDs light up for a split second, but after that nothing. it seems like the PSU cuts off the power instantly or smth.
also, can anyone identify this card? the sticker on the back is too worn.
GGigabiteM@reddit
I see rust and tantalum capacitors. I'd go with one or more of the tantalum capacitors being shorted, and rust as a secondary problem.
Probe the four orange tantalums on there with a multimeter.
rpocc@reddit
A little remark: that will work only with desoldered capacitors, because they all are essentially parallel, and only a very good milliohm meter can attempt to detect particular shorten part with no guarantee
GGigabiteM@reddit
They're not in parallel, they go across the different power rails. And you can most definitely tell if they're shorted while in circuit.
C41 is from +5v to Ground.
C44 is +12v to Ground
C43 is -12v to Ground
C42 is another +5v to ground, but a different +5v source pin.
On the +5v rail, you'd probably normally see a few hundred ohms to ground. With a shorted tantalum, it would be near zero. The other rails will have higher resistances, but the same applies, you'll see near zero with a shorted tantalum.
The axial leaded caps above the 74 series logic chips are probably ceramic dipped capacitors, these rarely fail.
rpocc@reddit
But there are also lots of ceramic caps at each IC, which can short as well. I had a new device I’ve assembled by myself and had to desolder like dozen of capacitors before I found a defective one.
50-50-bmg@reddit
TBH, I would not even probe them, I would replace them with aluminum electrolytics and 100nF parallel ceramics right away.
vinciblechunk@reddit
I third this, I had a 386 motherboard fail in exactly this way. I specifically suspect C41, 42, 43 or 44
LousyMeatStew@reddit
I second this. From what you're describing, sounds like a short caused by a failed cap.
As far as identifying it, it's a pretty run-of-the-mill multi interface card - looks to be a highly integrated one at that. Hard to tell if the main IC with the QC sticker on it has anything printed on it - whether it's worn or never had anything at all. Most likely blank from the factory, though. It's basically the retro equivalent of a glop top.
MajorTomIT@reddit
They missed final dot In Republic of China ROC
djnw@reddit
One of those resistors, R7 looks distinctly scorched on one ends
Low-Charge-8554@reddit
Also R14(?) looks a tiny bit scorched on end. Also - strange no FCC ID. Pretty generic Multi I/O card
cazzipropri@reddit
Salvaged from the Titanic?
tutimes67@reddit (OP)
probably sat in a box outside (with other rusting cards) for years at my local recycling center.
ive also found an audigy 2 in that same box that wotked, sadly its not the case for this one
rpocc@reddit
The card is very regular, and the first thing I would do is probing supply lines with ground because there can be a failure in one of bypass capacitors causing short-circuit. But burnt chips may do so as well and maybe for detecting the real source of failure you need to search the source of heat by sending a decent current on +12 and +5 supply busses for prolonged time. Phone repair servicemen are good at searching the shorted components by using some kind of vaporizer with rosin, covering entire circuit with a thin layer which changes it look when heated.
A finger can help too :)
Js987@reddit
It’s a fairly standard combo 2x serial, 1x parallel, and 1x game port card. These things were very common even to the end of the ISA era, I can remember having one in a mostly PCI system in the late 90s even. They were inexpensive and everywhere, you’d throw one in any system you needed an extra port in, disable or change the port number of the ones that conflicted via the jumpers, and be good to go.
I see a lot of corrosion, what looks like a burned resistor, and a lot of discoloration that suggests to me this card was exposed to unfavorable conditions. Unless you’re looking for troubleshooting and soldering practice this isn’t a board I’d expend the effort on repairing, as d-subs are a nightmare to get off if you’re not experienced with it, and it’s not the sort of rare item I’d expend my effort on even having replaced them before.
Sneftel@reddit
The cause is almost certainly a power rail being shorted to ground, likely under the corroded D-sub connectors, in a capacitor, or in an IC (in decreasing order of probability).
D-sub connectors are not easy to desolder, and you shouldn't attempt it unless you've already desoldered a similarly massive two-row through-hole component from a different board. The ICs will be easier to desolder but sourcing them will be more of a pain. The capacitors are trivial to replace.
If it were me, I would check the resistance across each cap and replace any that didn't show infinite resistance. If that didn't work, I would desolder the D-subs, clean up any rust underneath it, and run the card without connectors; if the problem was resolved I'd source and fit new ones. If the problem went on, and I really wanted to save the card, I would replace all the 74xx ICs. If that *still* didn't fix things, I don't think I'd bother trying to replace the controller IC, which would be a real pain to source.
dwnw@reddit
burned resistor, visible rust, tantalum capacitors