Duron Burnt?

[doggyfreestart]

Well, I just found an MS-6378 V1.0 on the local recycling centre, with CPU & RAM. It was POSTing, but had 7 pcs bulged Chssi 1500uF 10Vs . I had some Frolyt (German manufacturer) 1500uF 10Vs lying around, so I installed them. Result: No POST. Later I discovered, that these have a rather low ripple rejection, so possibly not suitable for motherboards.
OK, then I bought Rubycon ZL 1500uF 10Vs. No POST either.
So where should I look for the error? May I have killed the CPU with the wrong caps? How sensible are the Durons against ESD? (I just had it lying on the desk while soldering.)
I know it was an idea to put just ther first caps I've found, but it shouldn't kill the CPU, should it?

Bye:
Georg

[davmax]

Re: Duron Burnt?

Do your systen fans start? What if anything happens when power is applied.

It is possible you created a short circuit when replacing the caps. A short can stop the power supply, it will self protect.

Did you read the Recapping Procedure and the type of caps in the FAQ section of this forum?

Re ESD and CPU it depends what cautions you took like a wrist strap. Working of carpeted floors is real risky.

[doggyfreestart]

Re: Duron Burnt?

Fans start OK. Does this signal that the PSU is OK? No sound, no beep, nothing on the screen.
Some vias were a bit messy, when soldering in, so the procedure in the FAQ couldn't be followed exactly. I soldered these holes with the "tweeze-in" method. But no raw force applied.
There are no shorts on the caps, I can measure 1.65V on the positive side.
I used genuine Rubycon ZL caps, the same rating as the Chssis were before.
I have an ESD-Safe floor in the room I use for soldering, but no wrist strap. I wasn't able to kill anything like this yet, and I'm usually repairing mobile phones.
It's my second recap, the first went like a breeze... And never had ESD probs before, respectively.
Could the wrong Frolyt caps have killed the poor Duron?

[davmax]

Re: Duron Burnt?

So you have fan running and 1.65V on VRM output caps. That sounds like PS is OK.

It is hard to imagine how the caps would kill CPU. Maybe the wrong impedance charcteristics caused the VRM circuit to oscillate and generate excessive voltage. That is about the only possibility.

Now you have better caps you need to borrow another CPU if you are sure the supply voltage is 1.65V.

Wishing you success.

[dood]

Re: Duron Burnt?

I've found that with problems like these, if I just go back and reflow the solder at each of the joints I was working with, the problem is usually solved. Also, check to make sure that none of your soldering work ran over any other traces, or that you burnt any other traces.

[MD Willington]

Re: Duron Burnt?

RMA test procedure for socket A processor.

[doggyfreestart]

Re: Duron Burnt?

None of the RMA things apply. The resistance shows some 100 ohms at the beginning, then goes higher. Nice description, by the way.
I've just tried the CPU in my desktop, and it didn't make any sound, either.
So the moral of the story is: always use genuine, trusted caps, otherwise it can cost you a CPU/Mobo or both. And you were working for nothing. Sad but true. And now I drink a glass of wine to tribute my first killed Duron.

[doggyfreestart]

Re: Duron Burnt?

After having the wine, I went to the computer cemetery again - and I've found an Epox K8A3+ MoBo. So I put the CPU inside - nothing to loose - and it started POST-ing, then asked for a floppy because of BIOS checksum error. OK, it's damaged as well - but at least it does something...
Now I'm confused: I've tried that shitty CPU in 2 known good MoBos, set to 100 MHz FSB, 1.65 Vcore & reset CMOS, and no sign of life. But to display anything on the screen, the CPU has to be alive, hasn't? Pls help me out, I'm feeling stupid.

[kc8adu]

Re: Duron Burnt?

when they die they usually die completely and stay dead..
make sure there is no thermal compound (esp silver) on the core bypass caps and l bridges.
overuse of compound can cause trouble and can even foul the socket.

[davmax]

Re: Duron Burnt?

No need to feel stupid. There can be so many contributing factors, these are very complex systems. All you can do is work systematically noting what does and does not work. Always make sure all system voltages are correct.