Urgent\help needed from the PC tecchies

[tranquillo 10]

Sorry, tranquillo...I'm in a classified location, held captive against my will :) (Actually I'm near Atlanta GA, USA...I keep meaning to ask Olham to put me on the map and just forget). You guys can come visit if like...there's also a very lovely theater room I built down in the basement, adjacent to the shop.

 

But I have to say there's no spare room...you see, that's where the shop is *lol*

 

You know, that power supply you have/have ordered - the OCZ unit...that's a fairly high-quality unit I would think. I have such a hard time imagining that it being "off" by the 95msec for the PG signal is your problem. Hell, I have a hard time believing that it's really all that 'off', TBH. I guess the spec is what it is, but I just don't know.

 

I know you were unable to wait on parts being sent to you, but in hindsight, it almost seems it would've been worthwhile to have a graphics card headed your way as a contingency. By the time you get the PSU, etc. the graphics card could probably have been there.

 

I think you're right in all respects, Tamper.

[tranquillo 10]

Update.....................I've just seen my desktop

Not fully celebrating just yet though. My new psu arrived today but I had clients 'till early this afternoon so didn't get a chance to do anything with it 'till then.

I fitted the new psu, just running cables to the mobo............I got beeps and I could switch off using the front switch. I connected 1 hard drive and tried again..........got a series of beeps indicating no vga. Fitted the graphics card and just got the beep to indicate no keyboard. I checked the fan on the graphics card and it was spinning around quite merrily, as was I at this point. I connected the keyboard, mouse, and monitor and there it was, a desktop.

Rather than connect up my 2 optical drives and my other hard drive and aux fans, I decided to take the opportunity to upgrade the HD4830.

It seems that all is well and I'm digging out the invoice for the original psu as It might just be inside of its 3 year warranty which would be nice.

I also plugged the psu tester into the new one and all readings were virtually identical except the power good signal which was reading 320ms. So it seems that the pg signal does make a difference. Must admit, I didn't check the graphics output on the old one to see if that was faulty. I'll do that tomorrow.

Just left to decide on the card I want now.

[Tamper 9]

Well, I have been watching all morning to see if you'd post - that is indeed great news.

 

And I gotta say, I've learned something totally new...I never would've thought that little of a time difference would make the difference, but *wow*. You do realize this has only played out as it has, because I said it was probably the motherboard *lol*

 

(Now I'm thinking I might need to get one of those testers after all...) I wonder how often this PG is a problem (of all the "dead" PSU's one might have seen)?

 

At any rate, good to see you're on your way again!

[Von Paulus 8]

Glad that everything is working now, Tranquilo. And not having to change the board and formatting the system again, it's even better.

It's always good to have spare parts. So just keep the old graphic card for that purpose.

 

And I gotta say, I've learned something totally new...I never would've thought that little of a time difference would make the difference, but *wow*.

+1

 

(Now I'm thinking I might need to get one of those testers after all...) I wonder how often this PG is a problem (of all the "dead" PSU's one might have seen)?

Probably some, but not much. I've read that with lower brands they sometimes fake the pg signal, so it's always good. If it's true or not I really don't know.

I don't remember to have witnessed a PG good signal problem in my tester. It's true that I haven't used it so many times, as a dead PSU is a dead PSU.

[almccoyjr 7]

...I just wish I understood better why the 95msec would make that much of a difference. I get the specs, sure..but, that just seems so unlikely to me...I guess it is what it is.

Hey Tamper, maybe the following link will help regarding "good power signal" time interval. It has a straight forward diagram showing start up sequence in ms.

 

http://www.hardwaresecrets.com/printpage/Everything-You-Need-to-Know-About-Power-Supply-Protections/905

 

Even though it's probably not necessary, I've always checked the psu when swapping a new one in.

 

plug_nickel

 

ps: Absolutely, positively stay away from ANY psu that fakes the PGS on the 5v leg; you'll be sowing the seeds for a complete new system.

[tranquillo 10]

I'm grateful for the support I've had with this problem but, if I were to be honest, it's been no less than I have come to expect from this forum. If I get it all up and running tomorrow then there's drinks waiting for all in the mess. The very first thing I'm going to do is a full backup of both drives.

[Siggi 10]

"Nine times out of ten..." :wink2:

 

Glad to see you back up and running Tranq. :)

[Tamper 9]

"Nine times out of ten..." :wink2:

 

Glad to see you back up and running Tranq. :)

 

Yeah, yeah, OK...I just knew there would have to be some gloating. I was just waiting.

 

Good on ya, mate - glad you were right, sincerely. Like they say, even a blind squirrel can find a nut every now and again. But, as I have said (and others) nine times of ten, this would have had a different outcome. There are always exceptions to the rule, just not very often.

 

Like they say where I work, if you're not wrong every so often, you're not doing enough. I guess I get it right enough to have earned a living at it all this time, and/or convinced employer(s) the pay wasn't wasted

 

(Good thing I have no problem being wrong every now and then, and learning something from it too, because of course it does happen ...otherwise me and people like me would stop offering any help at all in forums)

[Von Paulus 8]

Yeah, yeah, OK...I just knew there would have to be some gloating. I was just waiting.

 

Good on ya, mate - glad you were right, sincerely. Like they say, even a blind squirrel can find a nut every now and again. But, as I have said (and others) nine times of ten, this would have had a different outcome. There are always exceptions to the rule, just not very often.

 

Like they say where I work, if you're not wrong every so often, you're not doing enough. I guess I get it right enough to have earned a living at it all this time, and/or convinced employer(s) the pay wasn't wasted

 

(Good thing I have no problem being wrong every now and then, and learning something from it too, because of course it does happen ...otherwise me and people like me would stop offering any help at all in forums)

 

Precisely my thoughts.

Unfortunately it's just an exception to the rule, otherwise it would be really easy to repair computers.

The customer who called me yesterday, unfortunately, was not so lucky. A brand new motherboard his coming on his way.

[Siggi 10]

Yeah, yeah, OK...I just knew there would have to be some gloating. I was just waiting.

 

Good on ya, mate - glad you were right, sincerely. Like they say, even a blind squirrel can find a nut every now and again. But, as I have said (and others) nine times of ten, this would have had a different outcome. There are always exceptions to the rule, just not very often.

 

Like they say where I work, if you're not wrong every so often, you're not doing enough. I guess I get it right enough to have earned a living at it all this time, and/or convinced employer(s) the pay wasn't wasted

 

(Good thing I have no problem being wrong every now and then, and learning something from it too, because of course it does happen ...otherwise me and people like me would stop offering any help at all in forums)

 

It's a weird thing though. I've found "nine times out of ten" during the years I did PCs, you found differently and the bloke at a local shop found otherwise too (always mobos with him). I think it must have something to do with chaos-theory and the laws of probability and stuff.

[Tamper 9]

Siggi,

 

Different experiences is all. What matters are two things:

 

- The mountain of parts (as pictured above) actually serves a purpose. Having a suspicion about what the problem is; that's one thing - anyone can do that. Having the parts to swap until you can prove/disprove...that makes all the difference.

 

- A 'guarantee' of sorts; anyone who approaches me about working on their machine, I tell 'em flat out: If I can't fix it, you pay nothing. I have never cost anyone money by being wrong while working on their PC. Which circles back to the 'mountain of parts' for confirming what I think, rather than heaving parts on the client's dime. These days, I don't charge for PC work anyway (just a hobby) - but still, I don't tell them they have to pay for parts unless I can prove those part(s) are actually causing their problem. On a few isolated occasions, I have had to buy a part myself because I suspected it and didn't have a suitable 'swap'. Far more often than not, that part was the one needed to fix the PC, so I just had the "client" pay me back for the part. But, in the very rare instances the part did *not* fix the problem, I just kept it.

 

And you see, that would get far too expensive if I were wrong all that often.

 

Unfortunately, neither of the foregoing two things really apply when you're trying to help someone halfway 'cross the world (when they cannot afford to wait for shipping, or maybe can't afford pay for it). You'll note I did say that, if there was to be anything bought for troubleshooting, a PSU is cheaper than a motherboard. And although we disagree, I maintain a motherboard is still far more likely to fail, overall. Simple probability...there are far more components on the motherboard to go bad, if one must guess. Power suppllies are fairly simple, by comparison. This case was a rare exception.

 

Also, further to '9 of 10 times' , I'd go so far as to say that, if you approached 1,000 people who work on PCs, the vast majority would have said motherboard. Moreover, the vast majoriy (999 of 1,000?) if they were being honest, would have to admit they didn't really know the PG signal was the problem here - even if they said PSU rather then motherboard. They might have said PSU as a guess because (for example) it's obvious the power isn't coming on, but that's not really a skilled or informed opinion...that's a simple case of "low-hanging fruit".

 

Now, let's be honest - did you really know the PG signal was wrong, in this case? (If so, I don't recall reading that as part of your theory...you just said the PSU).

 

I can tell you factually (perhaps it wouldn't matter) but I can accurately say that of all the the PSUs I've replaced, and quite a few I've repaired, this PG signal has never been the problem. I have seen primarily (85%+) the 'finals' go bad - the last voltage regulator in line on a 'rail'; the big ones mounted right to the large metal heatsinks inside. I have seen leaky and/or noisy caps, and I've seen fans that were left dirty and caused the PSU to overheat. I've seen cheap fans just get noisy or stop turning freely. These things are the most common failures by far, and they account for 90%+ ('9 of 10') of the PSU failures I've seen. A further 5% would likely be related to the 'smart circuit' that controls the over- and under-voltage/current protection (which includes the PG signal). But the PG signal is still only one part of what that circuit does, and it only matters for the split second the PC is being powered up.

 

I mention all these failures because, as I said above when I described the 'paper clip test': "...if the PSU comes on during the (paper clip) test, the chances are very high that your PSU is fine" I didn't say it's absolute, but the paper clip test will catch every one of these failures (assuming one has a meter) except the PG rise time. It just doesn't happen that often. Every one I've ever seen either fails the paper clip test, or is fine.

Edited January 15, 2012 by Tamper

[Siggi 10]

Gawd no, I had no idea what was wrong with his PSU. Just my experience over 12 years or so that nine times out of ten if a customer had a dead system, plugging my PSU in would bring it back to life and then I fitted a new one. Dead mobos were extremely rare for me. The guy in the shop, I suspected him of fleecing people.

 

I was also "No fix no fee". And that was going to people's houses so I really couldn't afford to get it wrong too often.

[Von Paulus 8]

My experience, tells that in case there is no power, no lights, no fans, 8 out 10 is only PSU. And that can be explained in my case, because, most people used cheap PSUs in Portugal (only in the last years the things changed). When I asked the clients if they want to pay a little more for a better PSU, 95% of them didn't want. You could explained them what were the advantages, but they didn't care. So basically only low quality PSUs.

But in case where the system has the lights on, with power, hardisks spinning, 8 out of 10 is motherboard. That's my personal and professional experience.

I worked in a computer shop (the same for 20 years) and in there, there were no PSU testers or multimeters. We would simply replace firstly the PSU with a new one, just to check.

Later when I opened my shop and after as a free lancer, it was when I begun to use the tester. And like I've said I don't remember to have seem a pg sign issue.

Edited January 17, 2012 by Von Paulus