VOGONS


First post, by artiemog

User metadata
Rank Newbie
Rank
Newbie

Hi all, looking for some assistance...My go-to old PC is one my dad built around 1998 with a Celeron 300A slot 1 CPU and an Abit BH6 motherboard. Unfortunately, it seems like at least one of these components recently took some damage.

On Saturday, I went to turn on the computer after not using it for a few weeks, but nothing seemed to happen when I pushed the power button. After checking everything, it appears that the power supply (a TOPOWER TOP-300SSA, date code 7.98) has failed. I took it out, attempted a paper clip start, and checked the voltage rails, and I got 0 V everywhere, so I hoped that meant everything else would be okay. I purchased and installed a new power supply, but unfortunately the machine still won't POST or give any error beeps. The fans all spin up and the case power light comes on, so the motherboard is apparently at least getting 5 and 12 volts, but the machine does not make any progress toward booting after that.

I took out all the RAM and PCI cards to try to get it to at least give me error beeps, but there was no change. I also attempted to reseat the CPU in the hope that maybe I had just bumped it or something when I was swapping the power supply, but it seems like part of the CPU card is missing somehow? There are no retention clip levers where they should be on the corners, nor is there any black plastic housing on the side opposite the fan; I can see the backside of the circuit board that the processor itself is on. I was unable to pull the card free because of this, but I did wiggle it and try to make sure it was at least seated snugly. No change, unfortunately.

I'm at a bit of a loss as to where to go from here. I'm handy with a soldering iron and comfortable trying to repair the motherboard, but I have no idea where to even begin troubleshooting, and I can't see any obvious signs of damage or even any unhappy capacitors (although I haven't taken the board out of the case yet to check the back side). I think I have another slot 1 CPU somewhere, so I'm going to look for it this week, but I have no idea if that one works or not either, and I'll still have to figure out how to get the original one out without (further) breaking anything. There's also the possibility that if the motherboard is damaged, it'll just kill the second CPU too. (And it's TBD whether or not anything else that was connected to the computer was killed as well)

If I at least knew that there had been overvoltage on one of the power rails, that would give me somewhere to start checking traces and components downstream of the motherboard power input, but to be honest I'm a bit flummoxed by this failure mode. All I can figure is that there was a brief spike at the moment the PSU died, and now it just can't even start up anymore so I can't repeat the condition.

I'd really like to get this machine working again, as it has a lot of both functional and sentimental value to me. I'm not ready to give up yet, but I'm not sure what would be best to try from here. Does anyone have any advice or suggestions?

Thanks!

Reply 1 of 8, by Deunan

User metadata
Rank Oldbie
Rank
Oldbie

These mobos did suffer from capacitor plague, did you replace the ones near CPU slot? If not take the mobo out, the CPU out of socket, and test the caps for shorts. Chances are one of the CPU power rail caps has shorted, took out a MOSFET, which in turn killed the PSU. It's a theory but it's not impossible.

If it's easier, and the nothing is smoking, you can probe the running mobo for voltages present. See if all the voltage regulators (esp. near CPU, but also RAM, etc) are doing what they should. If you do get correct voltages then you need either a POST card or another CPU to test with. Preferably both to be sure.

Reply 2 of 8, by PcBytes

User metadata
Rank l33t
Rank
l33t

Knowing it's an ABIT, replace caps first. I've seen them fail with zero outside signs.

"Enter at your own peril, past the bolted door..."
Main PC: i5 3470, GB B75M-D3H, 16GB RAM, 2x1TB
98SE : P3 650, Soyo SY-6BA+IV, 384MB RAM, 80GB

Reply 3 of 8, by artiemog

User metadata
Rank Newbie
Rank
Newbie

I'd heard about the capacitor issues with these boards but it felt like too easy of an answer for getting the board working again. Deunan's theory of something on the board shorting out the power supply is definitely interesting though. I'll check the caps and the voltage regulators and report back, and I'll look into getting a POST card. Thanks!

Any ideas on how to free the CPU module without those retaining clip levers being there?

Reply 4 of 8, by Grem Five

User metadata
Rank Member
Rank
Member

Pictures could help.

Reply 5 of 8, by artiemog

User metadata
Rank Newbie
Rank
Newbie

Well I went to take photos of the CPU but ended up being able to get it free fairly easily right afterward haha. All it took was a bit of gentle flexing on the edge of the slot.

I also took the time to tone out the power supply connector on the motherboard, and I did find some interesting things. There appears to be a dead short between the +5VSB pin and ground, and only 50Ω between the 5 V rail and ground. It looks like the -12 V pin is also bent a bit, which may have happened when I swapped the power supplies; the plug for the new one was noticably more difficult to connect snugly. I'll definitely repair this, but I doubt it was the only thing preventing the computer from booting. So I will pull the board probably this weekend and investigate further. Thanks to everyone for your help so far.

Reply 7 of 8, by bertrammatrix

User metadata
Rank Member
Rank
Member

If you suspect a short somewhere it'd most likely be a capacitor or the CPU voltage regulator.

I don't usually see dead power supplies from a board that has died (or when I have an oopsie) though, short circuit protection usually does it's job. Usually it's one thing or another, not both, but anything is possible.

Did the old PSU have a -12v rail? Does the new one have one? This is not a given - I'm not sure when exactly this requirement was dropped from motherboards, I do know that some wont work without it, and unfortunately some later PSU are missing it. Some without it I've opened up just to find the circuitry in place, but without an actual wire attached (must have saved a whole 1 cent there).

Reply 8 of 8, by momaka

User metadata
Rank Oldbie
Rank
Oldbie

Post pictures of the motherboard when you get a chance, as the other have suggested here.
If the motherboard really does have a true short-circuit between 5VSB and ground, it's possible the old PSU may have gone over-voltage on the 5VSB and killed something.

That said, do NOT toss the old Topower PSU in the trash just yet. If anything, we need to confirm if the PSU 5VSB circuit is indeed what killed the motherboard or not. Many old ATX PSUs used what's known as a 2-transistor self-oscillating circuit for the 5VSB, and some of these have been known for going out of control / over-voltage over time and killing things (e.g. the Bestec ATX-250-12E is one of the best-known examples in HP PCs).

So if you have a multimeter, take that old Topower PSU and measure the votage between the 5VSB pin on the ATX connector and ground. You should get a nice steady 5V +/-0.25V (e.g. 4.75-5.25V is acceptable... though I would expect much tighter regulation usually - something like 5 to 5.15V).
If you get nothing on the 5VSB pin (or very low voltage) or voltage that is too high and out of spec, then unplug the PSU, wait for like a minute, and them open it. After this, take good pictures of all sides under good light and post them here. That should tell a much fuller picture about what happened. If the PSU really went over-voltage on the 5VSB and killed something on the mobo, chances are the SIO/LPC probably bit the dust and you'll be unlikely to save it. But these are pretty rare cases.
IMO, also double-check all of your resistance measurements on the mobo and certainly do them with the CMOS battery removed. Moreover, if the new PSU can supply 5VSB voltage without the mobo tripping and killing it, then the short-circuit you measured is likely a fluke. So keep that in mind.
Bad caps could also be the issue, as the others have suggested. So we have to tackle this from multiple sides to try and figure out what happened... which is why I also think it's a good idea to see the original PSU inside.