What is wrong?

[Muse]

Running PC was fine until 5 days ago. I'd post in the other thread I started in this forum a few days ago but for some reason every time I try to open the thread for the last hour or so I get an interminable hour glass so I'm starting a new thread. Anyway, there's a new twist. Details and history:

Details of the system when things started going wrong are at the bottom of this post.

5 days ago, I turn on the computer and get no video and no post beep.

I disconnect everything including the keyboard and mouse and remove all the cards except the video card and get the same behavior.

I swap out the BFG geforce 6600 GT OC AGP video card for another, a BFG geforce 4600, and have video and a message that no keyboard is attached. I conclude that my video card has died and I call BFG support (there's a lifetime warranty on the card) and they send me an email with RMA instructions.

I reassemble all components, hooking everything up exactly like it was before this all started except this time I have the alternate video card in there, the BFG 4600, and press the power-on button and nothing happens! :Q

People tell me, and my computer hardware book tells me that it's either the MB or the PSU, assuming the power-on button works (I tested that already, and it's good). So, having not treated the MB roughly, and the video card having evidently died, I conclude that the PSU is probably what caused the video card to die and that the PSU subsequently stopped working altogether.

Operating on this theory, I removed the Corsair PSU and installed my former PSU, an Antec True 430. I leave all the cards in, hook up the floppy and IDE devices, but don't hook up the cabling yet, except for my monitor to the video card and my keyboard and mouse, and push the plug into the back of the PSU and then the fan at the back of the PSU starts spinning for a second and then stops. I press the power-on button and nothing happens and there's a strong smell of burning electronic components! :Q

What's going on??? Thanks very much for insight and advice.

- - - - -
System details as of last Tuesday when things started going wrong:

Midtower case
Corsair HX520w PSU
MSI K8N Neo-FSR/ V V2.0 motherboard
AMD Athlon Venice core 3200+ 2.2 GHz CPU
Arctic Cooling Alpine64 HSF on CPU
BFG geforce 6600 GT OC AGP video card
Samsung Original PC2700 2x512 MB for 1 GB
HDs: Seagate Barracudas: 120 GB, 160 GB, 200 GB; Samsung 500 GB SATA HD (16 MB/7200 RPM, Model HD501LJ)
NEC DVD burner
Floppy drive
Firewire card
USR Model 2977 PCI hardware modem
MyHD HDTV 120 HDTV PCI with daughterboard

 

[olmer]

As long as you are not shorting your board and have connected everything correctly i would assume a dead/dying mobo. Try it out of case first. Use my stock answer ( Disconnect power. Reseat CPU and power cables, remove everything else. Check jumpers. Connect only power button. Remove 3V battery and short-circuit the contacts. Press power button. Replace battery. Add one stick of working memory into 1st slot (closest to the CPU). Install graphics/use integrated if any. Try to boot. ) to make sure. You can start and test PSUs on their own, but i do not think it is the case here. If you have a choice ? choose a better quality board next.

 

[Muse]

Originally posted by: olmer
As long as you are not shorting your board and have connected everything correctly i would assume a dead/dying mobo. Try it out of case first. Use my stock answer ( Disconnect power. Reseat CPU and power cables, remove everything else. Check jumpers. Connect only power button. Remove 3V battery and short-circuit the contacts. Press power button. Replace battery. Add one stick of working memory into 1st slot (closest to the CPU). Install graphics/use integrated if any. Try to boot. ) to make sure. You can start and test PSUs on their own, but i do not think it is the case here. If you have a choice ? choose a better quality board next.

Yes, at this point I'm inclined to think the mobo has died. I'm wondering two things:

1. Is it possible that the dieing mobo on its way out trashed my video card. I tested the video card in my 2nd system, and it's not working in that either.

2. Is it possible for a dieing/dead mobo to harm a PSU? If not, I won't worry about either of the PSU's I had in the system this week. But if so, I will want to test the PSU's before installing my spare mobo, while I wait for a replacement mobo (one that supports SATA HDs and thanks for the suggestion on upgrading the mobo... will see what transpires when I contact MSI with the bad news).

And thanks for the details on testing the mobo. I don't know that I'll do those tests. Honestly, I don't see how the mobo could possibly be OK given what's transpired (OP).

 

[olmer]

1. Yes. 2.Yes. The likelihood of burned parts increases if you had a short somewhere. It does not really matter what died first PSU or mobo as long as your second working PSU failed as well. If you have a voltmeter ? start PSU outside the case and check voltages/amperage ? it is not that scary.

 

[Muse]

Originally posted by: olmer
1. Yes. 2.Yes. The likelihood of burned parts increases if you had a short somewhere. It does not really matter what died first PSU or mobo as long as your second working PSU failed as well. If you have a voltmeter ? start PSU outside the case and check voltages/amperage ? it is not that scary.

I started testing the Corsair PSU, one 3.3v, one 5v and one 12v lead. The voltages were:

3.41
5.03
12.08

Do I want to test the other leads? These are on the connector to the mobo, and the PSU was removed from the case. To get the PSU to function, I shorted ground to the "green" lead with a paper clip.

How do I test the amperages? This multimeter is a fairly good one, and has DC (and AC) amperage testing, with two ranges:

2A fused
20A unfused

How do I go about testing the amperages? Do I just put the leads in the same places as for voltage with the multimeter set to one of the DC amperage ranges (which?)? This would be without load, right? Or it would be, in the absence of load, just the load created by the multimeter.

Edit: I kind of think there must be a short on the mobo because of that smell. It was pretty strong and I noticed it practically the instant that I tried to turn on the system with the second PSU in there.

 

[olmer]

No point to test any further ? it is fine as long as you have stable readings. (You cannot smell burned components from its exhaust I hope) If you can see the place of short/physical evidence of one only then you can be 100% sure, otherwise also assume a bad mobo. I personally do not use/recommend MSI boards but came across/repaired pcs based on them before. Some of the Neo models appear to be very good v4m and feature -reach (as long as they work that is), so if you have no choice but to get a replacement ? give it another shot.

 

[Muse]

Originally posted by: olmer
No point to test any further ? it is fine as long as you have stable readings. (You cannot smell burned components from its exhaust I hope) If you can see the place of short/physical evidence of one only then you can be 100% sure, otherwise also assume a bad mobo. I personally do not use/recommend MSI boards but came across/repaired pcs based on them before. Some of the Neo models appear to be very good v4m and feature -reach (as long as they work that is), so if you have no choice but to get a replacement ? give it another shot.

So, test the voltages on the PSU's and as long as I have stable readings, such as the ones I stated in my last post, don't bother testing for amperages?

In the bolded text above, you are thinking of finding a burned component on the mobo? Or are you thinking of the PSU?

What do you mean by "very good v4m"?

Assuming you suggest I be satisfied with testing for stable close-tolerance voltage measurements on my PSU's, and that I'm looking for evidence of damage on the mobo (and RMA it), I'm going to install my backup mobo in the system, an MSI KT3 Ultra2, and use one of those PSUs and keep my fingers crossed that my other components aren't damaged. That machine was chock full of components. That's the main reason I bought that mobo -- every PSI slot is used, and it has 5. The system also had 4 HDs, a burner and a floppy and two COM devices.

 

[Muse]

I just tested the voltages on my Corsair HX520w PSU, outside my systems. Voltages look OK except for one. The -12v pin, second from the left measures -11.09v. Is that acceptable or does that possibly indicate problems with the PSU?

 

[Muse]

The Antec PSU seems dead. Guess the mobo fried it.

 

[olmer]

11.09 is acceptable.

v4m/b4b - value for money

By the evidence of short i meant a burned area/mark with a corresponding mark on the case wall, otherwise i would not totally blame myself ; ).

Yes, i think it is safe to try this PSU with a new mobo/other components, trying bare board first obviously. If some of the other components are damaged they would not damage your mobo.

If you had your graphics card burned as well ? it sounds severe enough so indeed your memory would likely be damaged, CPU/PCI cards are less likely but you should consider that possibility as well. RE: burned Antec ? i hope it burned with only mobo connected, if not than it may have damaged more components, that is if any survived the first problem.

 

[Muse]

Originally posted by: olmer
11.09 is acceptable.

v4m/b4b - value for money

By the evidence of short i meant a burned area/mark with a corresponding mark on the case wall, otherwise i would not totally blame myself ; ).

Yes, i think it is safe to try this PSU with a new mobo/other components, trying bare board first obviously. If some of the other components are damaged they would not damage your mobo.

If you had your graphics card burned as well ? it sounds severe enough so indeed your memory would likely be damaged, CPU/PCI cards are less likely but you should consider that possibility as well. RE: burned Antec ? i hope it burned with only mobo connected, if not than it may have damaged more components, that is if any survived the first problem.

I should have had only the minimum of components connected when I hooked up the mobo with the Antec, but I foolishly assumed that the PSU was the problem and not the mobo. I'd hooked up my HDs, both sticks of RAM (512 MB each) and one of my better vid cards (I have 2 BFG 4600's with Zalman heatsink, i.e. fanless). Thinking I may have fried the vidcard, I swapped it out with the other, which is in my second PC. It brings up video, but now I have a problem in the other system. The HDs are not detected, except for one. The boot HD runs off a Promise PCI IDE controller, and suddenly (this wasn't happening the other day) the first channel of the controller card isn't seen. I have a second "identical" Promise controller card and swapped it in and the problem persists.

I've had similar problems in the past, and I think it has something to do with the motherboard in there, an Epox 8K7A. Well, I think that the fact that the video looks OK (except that Windows 2000 won't boot), is reasonable proof that the video card is OK, and I'll leave off troubleshooting the Epox system and get back to my main PC, and hope that I can get it running OK with my alternate mobo, an MSI KT3 Ultra2... hoping that my other stuff isn't damaged, i.e. the DDR you fear is damaged and the HDs and other components. It's been a bad week.

 

[Muse]

I removed the MB, DDR sticks, CPU and inspected the MB fairly carefully and don't see anything that indicates a short or blown capacitor, resister, chip or anything else. I still can't see how the MB isn't bad, seeing as the PC wouldn't start with either PSU, and the second PSU was killed when it tried to start the system. Can I safely assume the MB has gone bad and contact MSI for a replacement MB?

I'm about to test my backup MB, the MSI KT3 Ultra2, outside the system with just one stick of DDR, a video card, the Corsair PSU, a mouse and keyboard attached and try to post and go from there.

 

[Muse]

Some good news at last:

Connected the Corsair PSU to my spare MB, the MSI KT3 Ultra2, one stick of the DDR that was in the problem system (was therefore suspect), keyboard (totally different), the CPU that was on the KT3 Ultra2 already.

Posts, or appears to. Get post beep, asked to insert boot floppy.

Switched DDR, same result. I realize I have to run memtest on both my DDR sticks to be sure they are OK, but so far so good. Verdict has to be my original MB is dead, and worse, it killed a video card and my spare PSU, or likely it killed them.

Well, at this point I'll start assembling the computer bit by bit and see if I have a working system by the end of the day. I hope my cards and HDs survived.