did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

[stanwebber]

i've got a pretty rare motherboard that i'm evaluating for repair. it's a biostar m7mia-raid which is the only ddr socket a motherboard ever produced, consumer or industrial, with an isa slot. its 20 years old now and finally gave up the ghost last month. i was saving a dd image of the hdd to a portable usb drive when i found it unresponsive with a black screen and a solid lit hdd led. after forcing it to power down it never posted again. the psu/cpu/case fans will spin and the onboard motherboard power led will light up, but no beeps or error codes even with the cmos reset and ram plus all expansion cards removed.

given the age, bad capacitors would seem to be the likely culprit, although their exterior appearance is absolutely pristine and the board never experienced any instability prior to dying so it's not a slam dunk. what concerns me more is that the motherboard was hooked up to a faulty power supply for more than a decade. the psu in question is an expensive clever power 500w unit that probably weighs more than 3 of those junk psu jobbies put together. it remains functional to this day and as far as i can tell with a multimeter on my testbed all the rails are still fairly close to spec under load.

the problem with the psu is that when the system is off it leaks voltage on the 12v and 5v rails. i measure around 1.8v on the 5v; 4v on the 12v wires when powered down. this only ever came to my attention when i happened to have the side panel off while swapping out components and noticed the case fans were still very slowly spinning (onboard motherboard led was also lit up when it should be dark). i'm assuming this could have been the condition of the pc for years (maybe even over a decade). after i caught the issue the motherboard persevered on for maybe another year with infrequent use before dying. every component from it, except ram, made it into a replacement kt133a sdram system and seems to be perfectly functional so nothing else got taken out with the motherboard.

my question is what kind of impact could this prolonged exposure to leaking voltage have had on system components? what electronics would be susceptible or immune? would there be any visible evidence to look for? hell, was it good, bad or indifferent to the capacitors?

[Topcat]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

Caps fail without visible signs. Before you determine that power supply still being good after all these years, you need to crack it open and look at the caps....if they appear ok, you then need to look at the outputs with a scope. A meter only tells you the voltages are present.....but does not tell you the condition of the voltages....

As for the motherboard, Biostar was no stranger to the cap plague. What brands are on this one?

[momaka]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

The board may still be salvageable / working... but like TC said, you can't really tell if the PSU is still good or not after all these years. Just about everything from that era was impacted by bad caps... and not only that! With PSUs in particular, it wasn't uncommon for some of them to use tan organic glue that goes conductive over time (especially with heat and moisture.) When that happens, this can cause all sorts of weird behavior - both for the PSU and the system.

So really, the best thing to do here would be to post some pictures of the PSUs internals and also of the motherboard. As the saying goes, a picture is worth a 1000 words... well, IMO, sometimes even more.

[Per Hansson]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

the problem with the psu is that when the system is off it leaks voltage on the 12v and 5v rails. i measure around 1.8v on the 5v; 4v on the 12v wires when powered down. this only ever came to my attention when i happened to have the side panel off while swapping out components and noticed the case fans were still very slowly spinning (onboard motherboard led was also lit up when it should be dark).

This in itself would be extremely unlikely to cause the board to die.
A low voltage does not really stress the board unless it is really poorly designed.
I looked for pictures of the board and I can't guess the models of the caps other than they look to be extremely cheap.
So based on that it is not impossible to be a capacitor problem.
But it is also not completely infeasible that the dd overwrote the BIOS if there is a bug in the system memory map.
So reading the BIOS in a programmer to check that it is ok would be a good first test.

[stanwebber]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

here's an image of the board and closeups of the largest capacitors as best i could manage with the available light. the northbridge heatsink isn't missing--i scavenged it for the replacement system. i don't ever plan on putting the psu back into continued service, but i still have it and can take some photos of the internals tomorrow for the curious.

[stj]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

"jackon" seriously?
those need to go for a start!

[Uranium-235]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

i've got a pretty rare motherboard that i'm evaluating for repair. it's a biostar m7mia-raid which is the only ddr socket a motherboard ever produced

No, it's not

[momaka]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

here's an image of the board and closeups of the largest capacitors as best i could manage with the available light.

Appreciate it. Cool board too!

Unfortunately, none of the cap brands look like anything I'd leave on there. Jackcon are fairly flakey. I can't tell the manufacturer of the green caps, but just by their sleeve color and vent, I can tell you they are not a known good Japanese brand.

In any case... before you start (or think about) changing the capacitors, let's do some simple checks first.
Start by resetting the CMOS first. Typically removing the CMOS battery will do it, but check your MB's manual - some boards may store CMOS settings on a flash, and the only way to reset such CMOS is through changing a jumper on the MB.
After doing this, return CMOS reset jumper back to normal position and install a good CR2032 battery. Then fire up the board with minimum required hardware (i.e. CPU, CPU heatsink & paste, PS-ON switch, speaker/buzzer, and 1 RAM stick - in either slot or whichever one the manual requires for 1-stick operation.) Do this with a known good PSU and not the PSU that you were using up to this point.

If the board still doesn't POST, try some different/various RAM sticks. Also a different video card too... or just remove and re-install the video card you had, just in case the contacts in the AGP slot have gotten oxidized. While at it, also release and reinstall the CPU a few times. Although rare, I've seen CPU pins catch slight corrosion on the surface that can make the PC not boot.

If the board still doesn't POST after doing this, try removing all of the RAM from the MB and fire up the system again. Does the board beep an error code? If no, then that means it's not even getting to the point loading the BIOS/CMOS... at which point, I would suggest you try a recap. And if the board does beep without RAM but won't boot with any in the slot, then again it would be a good idea to do a recap.

And of course, if the board POSTs after any of the above testing procedures, I would still advise a recap - at least for the caps around the CPU slot and the memory slots... well, just about everything in the upper half of the board above the AGP slot.

[stanwebber]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

No, it's not

WITH an isa slot.

[stanwebber]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

momaka, thanks for all the suggestions, but i was pretty exhaustive in my troubleshooting efforts. i stripped down the board to just cpu, ram & video--then just cpu only. at that point i wasn't trying to get it to post, but just wanted to hear some kind of beep. i even swapped out the pc speaker from a working pc just in case. i also swapped out the cpu and ram with spares. one notable thing to report is that with the cmos jumper in the reset position the psu would not power on just like if the motherboard were working. i don't know if that's a continuity thing or if there's some logic in the circuit.

about the only thing i haven't tried yet is to pull the bios chip and read it back in another board and compare it with what's on file; however, a failed bios chip only impacts you on the NEXT boot if the pc is already running. i'd have to concede my system locked up for some other component failure.

biostar is by no means a premier motherboard company. i'd put them in the middle of the pack of taiwanese manufacturers, maybe a bit towards the front.

[Uranium-235]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

WITH an isa slot.

now THAT, I withdraw my previous statement

[stj]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

try another cpu fan
the board may need to see the tacho output to boot up.

[stanwebber]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

removed the cover and the psu internals are incredibly busy...would take considerably more disassembly to get a good look at everything. i took some photos, but i'm really not motivated to go further.

the cpu fan is a possibility to check. the iwill motherboard i transferred it over to reads 0 rpm on the cpu fan header, but i ran across some contemporary reports on iwill motherboards not detecting lower rpm fans so i didn't give it any more thought since mine runs low.

[momaka]

Re: did bad capacitors kill my motherboard or prolonged exposure to faulty psu?

momaka, thanks for all the suggestions, but i was pretty exhaustive in my troubleshooting efforts. i stripped down the board to just cpu, ram & video--then just cpu only. at that point i wasn't trying to get it to post, but just wanted to hear some kind of beep. i even swapped out the pc speaker from a working pc just in case. i also swapped out the cpu and ram with spares.

Hey, good to hear you did all this.
I just wasn't sure what was tested and what wasn't, hence my suggestions.

Indeed if the motherboard went dead mid-way through it working otherwise, it sounds like the issue could be more like BIOS going corrupt or perhaps a component died somewhere. I say this because electrolytic capacitors usually "work better" when warmed up... so it's less likely that a cap failed midway through operation - though, not impossible either!

I guess if you do have a way to check/save/reflash the BIOS, then do that first. And if board is still dead, try measuring some of the important voltages, like CPU V_core, Northbridge Vcc, RAM Vdd, RAM Vtt, AGP Vcc, and etc. These 5 are usually the first to check, though. Most of the time, they will appear good, even if the caps are bad. So this is not a very exhaustive check... but sometimes it can show if there are problems. In any case, if you do have any caps on hand, it might be a good idea to replace some around the CPU and RAM + Northbridge area.

biostar is by no means a premier motherboard company. i'd put them in the middle of the pack of taiwanese manufacturers, maybe a bit towards the front.

Yeah, when they try, they can make some decent motherboards. I have a few from socket 754, 939, and AM2 era. I got all of them for cheap (really scrap value) to fix and they all came back to life after a recap. I like their BIOS - usually Phoenix and with lots of options unlocked.

try another cpu fan
the board may need to see the tacho output to boot up.

+1
Some socket A motherboards implemented this "protection" so that if the CPU fan isn't spinning or connected, the motherboard does not give power to the CPU and therefore does not POST. All of this because socket A CPUs don't have thermal protection (I'm sure everyone here has seen that Tom's Hardware video from back in the days) and will burn out without a heatsink within seconds of being powered up.

Of course, I'm assuming that even after swapping CPUs, you installed CPU heatsink onto the CPU and with thermal compound? I'm only mentioning this just in case, because I have seen people here before not know about that and burn out their CPU.

removed the cover and the psu internals are incredibly busy...would take considerably more disassembly to get a good look at everything. i took some photos, but i'm really not motivated to go further.

No problems.

Even this is more or less enough for us to get a clue.
This is a Topower PSU (as evident by the "Ever-power" text on the main transformer), which was a decent brand overall... though they no longer exist anymore.

I can see the primary capacitors are made by Fuhjyyu - a brand that is notorious for going bad. The secondary side also appears to have some Fuhjyyu caps in there. Thus, I wouldn't be surprised if the entire secondary side is done with Fuhjyyu caps. Because of that, the PSU is very likely due for a recap at this age, as Fuhjyyu caps rarely last this long (save for the 2 big caps on the primary - those are usually the only exception.)

I know the PSU is pretty crammed and may seem hard to take apart, but you really only need to remove the 4 screws on the corners of the PCB and untie the output wires from the case. After this, the PCB should lift up and out nicely. I can't see if the input wires are soldered to the PCB or connected with a connector... but that's usually the only thing that stops the PCB getting completely pulled out. Once pulled out, bend the output wires away and see how the output caps look. Maybe take a picture again, if you can. The PSU looks well-built, so it would be a shame to trash it.