FW: [SLE] Need help diagnosing hardware problem
On Tuesday, October 17, 2006 @ 9:31 PM, I wrote: -----Original Message-----
On Tuesday, October 17, 2006 @ 2:27 PM, Randall Schulz wrote:
Greg,
On Tuesday 17 October 2006 12:10, Greg Wallace wrote:
...
Wow. Sounds to me like there's no really accurate way to test the power supply. Maybe the best bet would be to exhaust all other possibilities and if the problem persists just replace the power supply to see if that fixes it.
Well, that's not really true. If you look closely at the power connector, you'll see that the sleeves of each pin come pretty close to the top of the connector, and multimeter probe pins can usually reach them easily, so measurement under load is probably not a problem. As for knowing what voltage to verify, most supplies list the pin-outs and color codes on their labels. Failing that, there are doubtless on-line resources with this information. E.g.:
- http://pinouts.ru/pin_Power.shtml - http://pinouts.ru/Power/atx_v2_pinout.shtml (ATX version 2) - http://pinouts.ru/Power/atxpower_pinout.shtml (ATX version 1.x) - http://pinouts.ru/Power/atx12v_pinout.shtml (Pentium IV and later)
These references give both voltages and color codes. They should be adequate for testing purposes.
Do you have a multimeter / VOM / DVM (or, god forbid a VTVM)? At the very least, you'll need that to perform a measurement. And a minimally acceptable DVM is cheaper than a high-quality power supply.
Greg Wallace
Randall Schulz
Randall: I don't have any electronic gear, so I'd be starting from scratch. Sounds like you're saying that to do the testing you unplug each wire one at a time and plug some device into that socket to test its voltage, repeating the process until you've tested each socket, is that right?
Greg Wallace
From what I just read in one of Felix's posts, you just stick the probe in alongside the wire and don't need to unplug anything to get the readings. At least that's the way I understood his post.
Greg Wallace
On Tuesday 17 October 2006 21:41, Greg Wallace wrote: .............
From what I just read in one of Felix's posts, you just stick the probe in alongside the wire and don't need to unplug anything to get the readings. At least that's the way I understood his post.
Greg Wallace
Greg, it is not complicated when you see once, but from your other posts is clear that you never did it before and it is kind of bad idea to start with power supply that can give enough amps for quite strong sparks if you cross metal tips on probes, and that can happen. So better call for help somebody that did that before, and learn. Text is not good medium to explain everything that one has to know for this kind of troubleshooting. It is simply to much text for one posting: - handling probes to avoid making short with tips, and many other incidents, - how to get reading if surface is dirty without forcing the probe to slip in wrong direction, shorting something; today normal probes are large comparing to traces and electronic components and one can make short easily with single tip, - how to be sure that probe is measuring "no voltage", comparing to probe touching wire insulator instead of contact tip which will give you the same reading. - proper voltage range selection, for cheap voltmeters without auto selection, - to pay extra attention not to leave instrument on Amps and measure voltage, which will fry something for sure, or on Ohms which will probably damage the meter. - to watch on static electricity that human body is loaded with, and that can fry motherboard components without any sign that something happened, - and so on ... There is to much for the first time that one has to know, to make sure that measuring will not produce more trouble that it can save from. If you really want to do favor to yourself, skip testing for the first time and ask someone for help, and learn. -- Regards, Rajko M.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Wednesday 2006-10-18 at 18:27 -0500, Rajko M wrote: ...
There is to much for the first time that one has to know, to make sure that measuring will not produce more trouble that it can save from.
If you really want to do favor to yourself, skip testing for the first time and ask someone for help, and learn.
I concur. - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) Comment: Made with pgp4pine 1.76 iD8DBQFFNr5jtTMYHG2NR9URAs9fAJoDTS8jdC3t05VX6gGDXppRQe27nwCdEkFp OKCds8pWdqVHXBLMhOaCyHA= =wt6n -----END PGP SIGNATURE-----
On Wednesday, October 18, 2006 @ 6:27 PM, Rajko M wrote:
On Tuesday 17 October 2006 21:41, Greg Wallace wrote: .............
From what I just read in one of Felix's posts, you just stick the probe in alongside the wire and don't need to unplug anything to get the readings. At least that's the way I understood his post.
Greg Wallace
Greg,
it is not complicated when you see once, but from your other posts is clear
that you never did it before and it is kind of bad idea to start with power
supply that can give enough amps for quite strong sparks if you cross metal
tips on probes, and that can happen. So better call for help somebody that did that before, and learn.
Text is not good medium to explain everything that one has to know for this
kind of troubleshooting. It is simply to much text for one posting: - handling probes to avoid making short with tips, and many other incidents, - how to get reading if surface is dirty without forcing the probe to slip in wrong direction, shorting something; today normal probes are large comparing to traces and electronic components and one can make short easily with single tip, - how to be sure that probe is measuring "no voltage", comparing to probe touching wire insulator instead of contact tip which will give you the same
reading. - proper voltage range selection, for cheap voltmeters without auto selection, - to pay extra attention not to leave instrument on Amps and measure voltage, which will fry something for sure, or on Ohms which will probably damage the meter. - to watch on static electricity that human body is loaded with, and that can fry motherboard components without any sign that something happened, - and so on ...
There is to much for the first time that one has to know, to make sure that
measuring will not produce more trouble that it can save from.
If you really want to do favor to yourself, skip testing for the first time
and ask someone for help, and learn.
-- Regards, Rajko M.
Definitely sounds like lots of ways to screw it up. If my re-seating the video card doesn't fix the problem, I'll probably let a pro work on it, unless the diagnosis lights on the back point me to some problem that seems fairly simple to fix (unlikely). Thanks, Greg Wallace
On Wednesday 18 October 2006 20:19, Greg Wallace wrote: .........
Definitely sounds like lots of ways to screw it up. If my re-seating the video card doesn't fix the problem, I'll probably let a pro work on it, unless the diagnosis lights on the back point me to some problem that seems fairly simple to fix (unlikely).
Thanks, Greg Wallace
Yes. BTW, although it might have different meaning, most of the boards today beep twice if they can't access video adapter. So it can be video adapter, but it can be something else that prevents test in memory range where adapter should be located. I still have one older computer that has problem with RAM connector. Every time I pull RAM out, for any reason, I have problem when is time to place it back. Sometimes I have to reseat memory modules few times before it stops beeping, with blank screen as you described, but it works after that without problems. You may also, disconnect all components except one RAM module and video adapter. If you can see BIOS messages, than video is most probably OK, at least in text mode. It is easy to troubleshoot if you have spare parts and know what part can be used as replacement, if not than you have to make estimate what is cheaper: - to take to shop instantly - to buy another computer - to buy another video card that might solve the problem, but it is not sure, and it can happen that you buy memory, than something else, and at the and to take to shop, or buy new computer, anyway. -- Regards, Rajko M.
On Wednesday, October 18, 2006 @ 11:31 PM, Rajko M wrote:
On Wednesday 18 October 2006 20:19, Greg Wallace wrote: .........
Definitely sounds like lots of ways to screw it up. If my re-seating the video card doesn't fix the problem, I'll probably let a pro work on it, unless the diagnosis lights on the back point me to some problem that seems fairly simple to fix (unlikely).
Thanks, Greg Wallace
Yes.
BTW, although it might have different meaning, most of the boards today beep twice if they can't access video adapter.
So it can be video adapter, but it can be something else that prevents test in memory range where adapter should be located. I still have one older computer that has problem with RAM connector. Every time I pull RAM out, for any reason, I have problem when is time to place it back. Sometimes I have to reseat memory modules few times before it stops beeping, with blank screen as you described, but it works after that without problems.
You may also, disconnect all components except one RAM module and video adapter. If you can see BIOS messages, than video is most probably OK, at least in text mode.
It is easy to troubleshoot if you have spare parts and know what part can be used as replacement, if not than you have to make estimate what is cheaper: - to take to shop instantly - to buy another computer - to buy another video card that might solve the problem, but it is not sure, and it can happen that you buy memory, than something else, and at the and to take to shop, or buy new computer, anyway.
?--
Regards, Rajko M.
If it happens again and I don't get a diagnostic code I will call in a technician and have him check power supply, connections, etc. I would think of it as a training session. In the end, he would only replace whatever parts are necessary, though I would be out some for labor, which, again, I could chalk up to training. Greg Wallace
<snip> If it happens again and I don't get a diagnostic code I will call in a technician and have him check power supply, connections, etc. I would think of it as a training session. In the end, he would only replace whatever parts are necessary, though I would be out some for labor, which, again, I could chalk up to training. For all that has gone into this thread, it would have been much simpler just to get a replacement power supply -- and if that didn't solve the
On 19/10/06 20:09, Greg Wallace wrote: problem, then at least you have a spare power supply. ;-)
participants (4)
-
Carlos E. R. -
Darryl Gregorash -
Greg Wallace -
Rajko M