Identifying a CPU killer (ahead of time) help

[SonicSteve]

Hi guys,
Can anyone tell me how to know if a board will be a CPU killer before I pop in a CPU. To this point I've just chalked it up to cost of business and I use the cheapest athlon socket A chip I can find. I've probably fried about 8 in the last year and I would love to stop frying them.

Can anyone help?

[Rainbow]

Re: Identifying a CPU killer (ahead of time) help

When testing a board in unknown state, I do this:
1. Check (by using ohm meter) that there is no short between 12V, 5V, 3.3V and GND
2. If all caps are present and none of them is bulged, power it on without anything installed and check DIMM and CPU voltages

[SonicSteve]

Re: Identifying a CPU killer (ahead of time) help

Great,
Now I need to know exactly how this is done. My background is computer hardware and IT not electronics. I suppose I'm coming at the field of mainboard repair from a disadvantage.

[kc8adu]

Re: Identifying a CPU killer (ahead of time) help

check vcore by reading across one of the group of large caps close to the cpu.
vdimm across the lytics nearest the dimms.

[gonzo0815]

Re: Identifying a CPU killer (ahead of time) help

I usually power the board up without any proc. The i grab a voltmeter and measur either on the backside on the big caps if any Vcore like voltage is there. Or i put on probe to a ground point (e.g. a circle around a screw ) and put the other probe to the legs of those rectangular transistors or diodes. Usually on of those pins is on 5v or 12v, one is on 10v and one is on Vcore. Se atachement for an example. The layout of a Vcore circuit can differ, but they are always the same basic shematic. There are usually on every board an inputinductor, the input caps and hte o/p section. It is not hard to fifure it out.

PS: mesuring the o/p for short with 5v or 12v line to ground can be misleading. On any of my recapped boards, i have always measured a short with my ohmmeter (may be due to the small supply current of the DVM and the high capcitance of the caps). But all of them are working fine.

[SonicSteve]

Re: Identifying a CPU killer (ahead of time) help

Wow thats awesome,
Thank gonzo0815 your pic is exactly what I needed.

[MD Willington]

Re: Identifying a CPU killer (ahead of time) help

Knew it was an ECS without even zooming in on the picture.. lol

[Shroomie]

Re: Identifying a CPU killer (ahead of time) help

Haha MD, me too. Heatsink's a dead giveaway.

Nice pic, anyway, Gonzo.

Edit: 250? Dang.

[linuxguru]

Re: Identifying a CPU killer (ahead of time) help

Yeah, the VRM layout looks similar to my K7SEM (M810*) which once ate up a nice Thoroughbred B due to bad caps. If you've lost 8 CPUs this year, check the PSUs first for bad caps.

[MD Willington]

Re: Identifying a CPU killer (ahead of time) help

VRM and orange "Blob" are always the indicators.. lol

[gonzo0815]

Re: Identifying a CPU killer (ahead of time) help

If you have seen once, you know them all

[tronxkid]

Re: Identifying a CPU killer (ahead of time) help

This is exactly what i want.
Thank you guys.

[Per Hansson]

Re: Identifying a CPU killer (ahead of time) help

There are also special cards you put in the CPU socket with LED's on them that will show if it is working or not, like this one for example...

http://cgi.ebay.com/940-CPU-Socket-p...item3364f38245

You need a different one for every type of socket of course...

[kaboom]

Re: Identifying a CPU killer (ahead of time) help

I usually power the board up without any proc. The i grab a voltmeter and measur either on the backside on the big caps if any Vcore like voltage is there. Or i put on probe to a ground point (e.g. a circle around a screw ) and put the other probe to the legs of those rectangular transistors or diodes. Usually on of those pins is on 5v or 12v, one is on 10v and one is on Vcore. Se atachement for an example. The layout of a Vcore circuit can differ, but they are always the same basic shematic. There are usually on every board an inputinductor, the input caps and hte o/p section. It is not hard to fifure it out.

PS: mesuring the o/p for short with 5v or 12v line to ground can be misleading. On any of my recapped boards, i have always measured a short with my ohmmeter (may be due to the small supply current of the DVM and the high capcitance of the caps). But all of them are working fine.

That board doesn't even have synchronous rectification!

ECS...

[lmcancu]

Re: Identifying a CPU killer (ahead of time) help

The top 3 caps plus 2 at the bottom look like they are bulging ...

[ukvince]

Re: Identifying a CPU killer (ahead of time) help

That's exactly what i was thinking ....