Recapping Abit BG7m with polymer

[Barchetta]

My decade old Abit 478 pin P4 533 MHz FSB board finally ceased to function reliably and killed its ATX power supply. Visual inspection revealed eight 2200uF 6.3V HM(M) HD147 Nichicons bulging/leaking (and two empty spaces) near the processor, and one lonely 1000uF 10V HM(M) HD139 Nichicon bulging (and one empty space) near the ATX power connector. The other smaller caps appear ok.

After some research (mainly badcaps.net and this forum), I ordered 10 Nichicon 1800uF 6.3V polymer caps and 2 Nichicon 820uF 10V polymer caps from Digi-Key.

I could not find the ESR and ripple of the originals for comparison (I obviously don't know where to look) so I just settled on 82% of the original capacitance, the same voltage, and ordered enough polymer caps to fill the empty spaces. I decided to use polymer because it seems more interesting than returning the board to original specs.

The board would have very little value to me even if I fixed it (it's the slowest of the seven computers in the house), so this is more for my own entertainment (and I might learn something - like how inadequate my soldering skills are).

Will the caps I ordered work? or at least not burn the house down?

Filling the empty spaces seems like a good thing, I figure the spaces are there due to "cost improvement". Any reasons I shouldn't do this? What are the effects of adding parallel caps?

Is there anything I can measure with a multimeter or 30MHz oscilloscope (I know - it's like WWII era) to quantify my results (if any)?

I have a basic plan of what I'd like to do, but I am open to suggestions or warnings of impending doom!

[c_hegge]

Re: Recapping Abit BG7m with polymer

Those caps should be fine. I'd fill up the empty spaces as well. I've done that a few times before.

[Barchetta]

Re: Recapping Abit BG7m with polymer

I had been trying to do a clean install of XP before this repair; it passed the point where it was hanging with 33 mins. to go and completed the install. It was rebooting alot so I figured it didn't survive my soldering. Since half of the previous install was with the bad caps I started over.

XP is now installed with all the updates. I think the security updates could be considered a burn-in test just because of the number of them. I'm still having the same trouble with the dvd rom drive, where it works just fine for the XP installation only. Once XP starts, dvd access gives errors or fully locks the OS.

I've been running memtest for almost three hours with no errors, and nothing has exploded! Yet.

[silvergoat]

Re: Recapping Abit BG7m with polymer

How do you know what type of capacitor to add into the empty spaces?

[Th3_uN1Qu3]

Re: Recapping Abit BG7m with polymer

They're in parallel so you just add more of what's already there.

[PCBONEZ]

Re: Recapping Abit BG7m with polymer

They're in parallel ....

Not always.
You have to check.
.

[PCBONEZ]

Re: Recapping Abit BG7m with polymer

The "eight 2200uF 6.3V HM(M) HD147"
is actually:
- eight 2200uF 6.3V HM(M) H0147
That's 2001, week 47 and part of the defective production run.
.

[PCBONEZ]

Re: Recapping Abit BG7m with polymer

My decade old Abit 478 pin P4 533 MHz FSB board finally ceased to function reliably and killed its ATX power supply. Visual inspection revealed eight 2200uF 6.3V HM(M) HD147 Nichicons bulging/leaking (and two empty spaces) near the processor, and one lonely 1000uF 10V HM(M) HD139 Nichicon bulging (and one empty space) near the ATX power connector. The other smaller caps appear ok.

After some research (mainly badcaps.net and this forum), I ordered 10 Nichicon 1800uF 6.3V polymer caps and 2 Nichicon 820uF 10V polymer caps from Digi-Key.

I could not find the ESR and ripple of the originals for comparison (I obviously don't know where to look) so I just settled on 82% of the original capacitance, the same voltage, and ordered enough polymer caps to fill the empty spaces. I decided to use polymer because it seems more interesting than returning the board to original specs.

The board would have very little value to me even if I fixed it (it's the slowest of the seven computers in the house), so this is more for my own entertainment (and I might learn something - like how inadequate my soldering skills are).

Will the caps I ordered work? or at least not burn the house down?

Filling the empty spaces seems like a good thing, I figure the spaces are there due to "cost improvement". Any reasons I shouldn't do this? What are the effects of adding parallel caps?

Is there anything I can measure with a multimeter or 30MHz oscilloscope (I know - it's like WWII era) to quantify my results (if any)?

I have a basic plan of what I'd like to do, but I am open to suggestions or warnings of impending doom!

I wouldn't bother with polys outside the VRM.
Doesn't get you much unless you have heat issues.
-
With good Polys [ESR under .010], as long as the total uF in Vcore is 5500uF or more and you use 7 or more poly [for combined ESR] - you should be fine.
That is what it typically used on factory 'poly boards'.
Some boards do okay with less but you have plenty of caps 'spots' available so don't skimp.
[8x 820uF would be a great way to go.]
-
For the 16v VRM caps 3x 470uF or 4x 330uF works fine.
.
Historically with lytics they over-kill the uF in VRM's.
It's not about the uF, they used those caps to get the ESR lower.
.