Thank you for being a valued part of the CNET community. As of December 1, 2020, the forums are in read-only format. In early 2021, CNET Forums will no longer be available. We are grateful for the participation and advice you have provided to one another over the years.

Thanks,

CNET Support

General discussion

Testing the Power Supply?

Jul 28, 2004 7:07AM PDT

What is the easiest way of testing the power supply WITHOUT plugging it into all the devices at the computer? I have heard that if defect, it may fry the mobo so here I am with a PSU in my hand waiting for somebody to tell me how to test it without putting it back in...

Thanks in advance.

Discussion is locked

- Collapse -
Re: Testing the Power Supply?
Jul 28, 2004 7:53AM PDT

If you don't know how or have never used a digital volt meter for such purposes, and have some reason to believe it is suspect, go get help or another power supply.

Except for the motherboard, you could check the other outputs using an old device like a floppy or fan with those connectors, but be careful.

They are generally cheaper to replace.

- Collapse -
Re: Testing the Power Supply?
Jul 28, 2004 8:22AM PDT
- Collapse -
Thanks to the both of you...
Jul 28, 2004 10:51AM PDT

.
I just wonder why the meter that you linked to at Compgeeks, wouldn't be enough to determine if a PSU works or not...
.

- Collapse -
Because it doesn't measure the voltage.
Jul 28, 2004 11:25AM PDT

And as we see in the Tomshardware article, it doesn't load the supply enough to see if it delivers the Watts. The oscilloscope is a great way to see if the supply is emitting CLEAN POWER. I've tracked down too many siezure prone machines to glitchy power supplies that would have looked ok with a voltmeter.

The sad truth is that in the US, we find 450 Watt new supplies for under 20 by the boatload. So the 4 thousand dollar scope only pays off when the customer wants "proof" its the power supply. To recoup the equipment costs, proof is suppied and we up the service charges by a "Ulysses S. Grant".

Hope this explains it.

Bob

- Collapse -
(NT) (NT) Thanks!
Jul 28, 2004 1:49PM PDT
- Collapse -
One last question...
Jul 30, 2004 4:12PM PDT

.
Do you think this is a good buy? It's slightly more expensive than the one Proffitt suggested and I know it is not enough to be a 100% sure about the PSU. But is this worth an additional $10 comparing to the other one?
.

- Collapse -
Here. Not worth 10 bucks.
Jul 30, 2004 11:40PM PDT

Since 450 Watt generics are under 20 by the boatload, we don't test many supplies. We toss them in the dumpster and move on. Testing would outstrip the replacement costs.

But if an owner insists on testing, we do so after they agree to the 100 buck minimum charge for "tests."

Some go ballistic, some see the light, others just say "replace it."

Bob

- Collapse -
OK, new PSU!
Aug 5, 2004 10:08AM PDT

Hey everybody and thanks for your input.

I finally got a new PSU which I hope is more reliable than the one I had. My new one is this one... What do you think? Let me know!

- Collapse -
Re: OK, new PSU!
Aug 5, 2004 10:54AM PDT
- Collapse -
OK...
Aug 5, 2004 11:03AM PDT

.
But does it have the 12V connector with the 20 pins?
.

- Collapse -
(NT) (NT) These are custom units, so they can do that...
Aug 5, 2004 11:10AM PDT
- Collapse -
(NT) (NT) How much are they and where on the web do I find them?
Aug 5, 2004 11:13AM PDT
- Collapse -
Figure it's by the pound (wieght) and at satcon.com
Aug 5, 2004 11:24AM PDT

Your other selection is supurb IMHO. My responses have been mostly leg-pulling...

Bob

- Collapse -
LOL!
Aug 5, 2004 11:55AM PDT

I know Bob!

But do you apporve of my new one? It's gotten high ratings almost everywere, so hopefully it is a good one...

Charlie

- Collapse -
Yup.
Aug 5, 2004 12:00PM PDT

That's a fine selection. It should power almost any system you'll find in 99.9% of the PCs we would see.

Bob

- Collapse -
Amp...
Aug 5, 2004 1:16PM PDT

Could you explain the importance of the Amps in an easy -to-understand-way? I know it is important... The old PSU that I had had 12V+ and 18A and the new one has 30A id I'm not too wrong. I know it is better, but not why it is better...

Thanks.

- Collapse -
Re: Amp...
Aug 5, 2004 11:51PM PDT

In short, as the computer designs rolled around, the number of amperes required on the 3.3V supply was growing almost out of control. Let me take the extreme example of the Athlon 64 that can suck 100 Watts. At 3.3 Volts that 33.3 Amperes. What wire size would we need?

http://www.rbeelectronics.com/wtable.htm points to number 2 gauge. In copper that's over 1/4 inch of solid copper. It's slightly bigger in stranded.

Yes, you can run multiple wires, but the next hurdle is the ATX connector. From memory, each contact is good for about 5 amperes. So you'd need 6 or 7 of these just for the 3.3V and more for the grounds.

This won't do.

This is why the move to 12V for the CPU supply. At 12V vs 3.3V that means for the same Watts, you need 0.275 or about 1/4 the wire and connector amperes. This is doable. And why you have seen the Amperes climb on the 12V portion of the PSU.

Bob