ECS K7VTA3 Capacitors?

  • Thread starter Thread starter Noozer
  • Start date Start date
N

Noozer

Just had one of my PC's fail. When the PS was switched on (back power
switch) the machine would blip on and off and stay off.

The board is an ECS v3.1 K7VTA3 board with RAID (specs here:
http://www.ecsusa.com/products/k7vta3_3.html )

Looking at the mainboard, the three capacitors at the rear of the mainboard
by the regulators were all swollen. These are 6.3v 3300uf capacitors by
"OST".

I got the board running again by pulling all three caps and installing a
single 10v 1500uf cap. DEFINATELY not the best solution, but it's the
closest match I could make and I need the machine running. PSU has been
swapped as I *THINK* that the original PSU may be straining the mainboards.

Does anyone know if this mainboard has a history of failing caps? Is there a
way to tell if a PSU is out of spec enough to be damaging the caps?

Thanks!
 
Swollen caps are a common failure for a wide
variety of mobos. There was a bad batch of caps
and a lot of manufacturers got bit.
 
Just had one of my PC's fail. When the PS was switched on (back power
switch) the machine would blip on and off and stay off.

The board is an ECS v3.1 K7VTA3 board with RAID (specs here:
http://www.ecsusa.com/products/k7vta3_3.html )

Looking at the mainboard, the three capacitors at the rear of the mainboard
by the regulators were all swollen. These are 6.3v 3300uf capacitors by
"OST".

I got the board running again by pulling all three caps and installing a
single 10v 1500uf cap. DEFINATELY not the best solution, but it's the
closest match I could make and I need the machine running. PSU has been
swapped as I *THINK* that the original PSU may be straining the mainboards.
Click to expand...

Since you need machine running, immedately source more caps, or replace
board, and underclock it as much as possible.

If you can follow the regulation circuit, note where in the circuit the
caps appear. Power supply problem would typically cause failure of those
on the high side of the circuit. That is, measured they'd be at (slightly
less than) 5V read by a meter. Overclocking would typically cause failure
of caps on low side, voltage of cpu vcore.

Does anyone know if this mainboard has a history of failing caps? Is there a
way to tell if a PSU is out of spec enough to be damaging the caps?
Click to expand...

Ost caps are not among the most commonly mentioned brands that were
subject to fail in past years' boards, but IF they are the problem you'd
be looking at replacing most if not all of them... from the picture at
link you provided that looks like 11 caps left of cpu, a few more around
memory banks and a couple above agp slot, though i can't see well enough
if all of those are same make and model.

You could take a multimeter, probes, to the back side of the ATX connector
where the leads enter and measure running voltage, but that will not be
sufficient to measure ripple. It's possible you have a power supply
design problem and/or component failure or inadequate capacity. You might
assess it's true capacity compared to that needed by the system, and
(after unplugging it for a few minutes) open it and inspect it,
particularly the caps nearer the wiring harness exiting the casing.

If the power supply is an overrated generic, the best course of action
would be to replace it. Since you don't have the spare capacitors and
there appear to be so many of same type, it might also be best to just
replace whole motherboard. At the very least, get more caps onto the
board, a single 150j0uF cap is not going to take the place of 3 3300uF for
very long, particularly if the power supply is the problem rather than the
caps themselves.
 
Pen said:
Swollen caps are a common failure for a wide
variety of mobos. There was a bad batch of caps
and a lot of manufacturers got bit.
Click to expand...

This I know... Just didn't think that this mainboard was old enough to be
part of that batch.

<snip>
 
Alceryes said:
Agreed, I doubt it was the PSU.


TO NOOZER: What makes you think it was the PSU?
Click to expand...

I've had a few mainboards that had the same problem... I've had the PSU for
over two years. Unfortunately with all the upgrades and hardware swaps we
done I can't be sure that this PSU has been in server for very long on all
of them. At this point I'm just taking it out of service to be on the safe
side.

BTW, th PSU is an ATNG Power Co. Model PS-300S.

Ratings:
3.3v- 15A
5v - 29A
12v - 11.5A
-5v - .5A
-12v - .5A
5Vsb - 1.5A

....and the mainboard was pusing a 2000XP CPU and some low power cards.
 
kony said:
mainboards.

Since you need machine running, immedately source more caps, or replace
board, and underclock it as much as possible.
Click to expand...

Sourcing more caps isn't the easiest thing. Looks like I'll be ordering
replacements from the net. $15 to ship and then convert to Canadian currency
will make this more expensive than necessary.
If you can follow the regulation circuit, note where in the circuit the
caps appear. Power supply problem would typically cause failure of those
on the high side of the circuit. That is, measured they'd be at (slightly
less than) 5V read by a meter. Overclocking would typically cause failure
of caps on low side, voltage of cpu vcore.
Click to expand...

Definately no overclocking in this machine. It just idles most of the time
in my basement as a file/email server.
Ost caps are not among the most commonly mentioned brands that were
subject to fail in past years' boards, but IF they are the problem you'd
be looking at replacing most if not all of them... from the picture at
link you provided that looks like 11 caps left of cpu, a few more around
memory banks and a couple above agp slot, though i can't see well enough
if all of those are same make and model.
Click to expand...

The rest of the caps look fine. That's one reason I suspect the supply.
Since I have to order replacements online I'm going to be sure to get my
moneys worth for shipping and get LOTS of replacement values.
You could take a multimeter, probes, to the back side of the ATX connector
where the leads enter and measure running voltage, but that will not be
sufficient to measure ripple. It's possible you have a power supply
design problem and/or component failure or inadequate capacity. You might
assess it's true capacity compared to that needed by the system, and
(after unplugging it for a few minutes) open it and inspect it,
particularly the caps nearer the wiring harness exiting the casing.
Click to expand...

PSU is two years old. At this point I'm just taking it out of service. I've
had three boards that had cap problems and its very possible that this PSU
spent some time running each of them.
If the power supply is an overrated generic, the best course of action
would be to replace it. Since you don't have the spare capacitors and
there appear to be so many of same type, it might also be best to just
replace whole motherboard. At the very least, get more caps onto the
board, a single 150j0uF cap is not going to take the place of 3 3300uF for
very long, particularly if the power supply is the problem rather than the
caps themselves.
Click to expand...

I'd replace the MB, but finding an equivilent would be a pain. I'm using the
RAID controller so I have to at least have the MB working long enough to get
the files from the RAID array.

Thanks for the feedback everyone!
 
At the very least, get more caps onto the
board, a single 150j0uF cap is not going to take the place of 3 3300uF for
very long, particularly if the power supply is the problem rather than the
caps themselves.
Click to expand...

Swapped the PSU out just to be on the safe side. Definately will be trying
to find replacement caps.
 
Back
Top