[zerodamage]

This is with a Antec 400 Watt "Smartpower". Are the numbers within spec of what a PSU should be?

[mkk]

Even if the +5V seems a wee bit low at a glance, it is still within specs for typical allowed variance. And that it's slightly low is better than if it had been slightly high. Only a die hard overclocker could possibly be hurt in the efforts to squeeze out the last clocks of something. That Epox has chosen not to use the available +12V connector on this motherboard might have something to do with this. With a less good powersupply I guess you might possibly have had a real problem.

[zerodamage]

How do you know that the Epox isn't using the +12?

[ToshiroOC]

I think he means the 4-pin square 12v connector was not included on the motherboard. The +12v rail is being used through the ATX connector

[zerodamage]

No actually it isn't. And now I know what he means. Thanks for clearing that up for me Tosh.

[zerodamage]

Originally posted by mkk
Even if the +5V seems a wee bit low at a glance, it is still within specs for typical allowed variance. And that it's slightly low is better than if it had been slightly high. Only a die hard overclocker could possibly be hurt in the efforts to squeeze out the last clocks of something. That Epox has chosen not to use the available +12V connector on this motherboard might have something to do with this. With a less good powersupply I guess you might possibly have had a real problem.

Well that +5 right now is basically where it's at with a descent OC on my system. I guess there is nothing to complain about.

[The_Neon_Cowboy]

Originally posted by zerodamage
This is with a Antec 400 Watt "Smartpower". Are the numbers within spec of what a PSU should be?

This should answer your question.




PSU is eather under cooled, over loaded, or both.
Remember 400W is a peak rateing most PSU's cannot handle close thier full wattage for very long

(max variable according to troubleshooting and repair experts within 5% is acceptable. )




here is a shot of my 400w for contrast

[Matth]

Another point to watch is that the monitoring chip on the PC is not calibrated, so the measurements cannot be relied on 100%.

PS. 16C/60F case temp? - are you wearing your winter woolies?

[zerodamage]

Originally posted by Matth
Another point to watch is that the monitoring chip on the PC is not calibrated, so the measurements cannot be relied on 100%.

PS. 16C/60F case temp? - are you wearing your winter woolies?

I was waiting for someone to say something about that. lol
My comp is in my house and it is not finished. So when I go down there after work, I then turn on the heat. It is usually in the 20's when I get there. At peak Temps, my CPU did not go over 105F, and that is on stock cooling and the temp in my house at about 76 degrees.

[zerodamage]

neon,

You said within 5% which is what I am aware of. I believe I am within that.

+5v 96.22%
+12v 98.25%
-12v 95.84% (very close)
-5v 98.8%

Now some of these may seem low but I think that it may be my mobo is showing some resistance between the monitoring and the PSU. I will do some actual testing on my PSU this weekend and see what they really are. This would make a nice guide..... I will work on it.

[toddsmack2k]

If you are not experiencing any random reboots or crashing under stress then let her rock Zero!

[zerodamage]

I seriously think it is my monitoring and not the PSU. I have a digital tester and I will do some testing and see what the actuall results are. Be very interesting.

[zerodamage]

I will be confirming the mobo's readings of my rails against an actual multi-reader tonight. We will see if my PSU is struggling.

[swimtech]

Originally posted by zerodamage
I will be confirming the mobo's readings of my rails against an actual multi-reader tonight. We will see if my PSU is struggling.

That's the ticket!

[zerodamage]

OK, I tested my rails with a multi-meter. And lets just say that the software on-board numbers are off by a decent amount. So there is some restriction on the board between the PSU and the monitoring chips.



Here is a side by side of what the software says and what the actual readings are.
*Keep in mind that this is with my cpu OC from a 2500 barton to a 3000 barton and my FSB OC from 333 to 400 fsb. 3 HD, 2 Opticals and a 9500 Pro that is also OC.




Here is the situation in summary.
My software says:
+5V ~~~ 4.784
+3.3V ~~~ 2.96
+12V ~~~ 11.795

True readings on 20-pin with Multi-meter
+5V ~~~ 4.93
+3.3V ~~~ 3.34
+12V ~~~ 12.22


Now keep in mind there is little to no variation while monitoring the rails with the multi-meter. On the +5 there was no variation at all. I checked each 3 times to make sure. The 3.3V there was only a +/- .02 variation. On the +12V only a =/- .03 variation. I am thinking that my PSU is in quite good quality. Right now it is comformatable and the air coming from my PSU exhaust is room temp.

Any thoughts?

[Sourcer_2002]

Hmm you dont wanna see my volts....my 400W antec brunt out and now im useing a genric 300W PSU with 2 opts 2 hdds Abit OTES 5600, 2500+ @ 3200+ and 6 80MM fans.

[swimtech]

I'd call it good - quite good actually!

[The_Neon_Cowboy]

My post to here yesterday disappeared ... WTF oh well not typing it all over again...

Basically gave a long description what I had in my system (extremely heavily loaded) at those reading I screen shot out of. Ands said that I distrust the official %5 rule … I usually go be three percent or less on my systems usually their dead on. Also liked to the PSU article and mentioned it was partially covered there.



But in your recent tests it looks good….

[Vampyromaniac]

Of course the stress from the CPU/chipset/RAM/Video card were removed during the measurement, were all your drives still connected and running when you took the measurement?

[zerodamage]

Neon, thanks for your thoughts. I highly value them.

I can attest though that no one has deleted any posts here. It may not have posted to begin with. I've been watching this thread closely.

My point now is that you can not trust what the software says about your Rails accuratly. I am going to try out a few more boards and see what their software says and then do an actual check via the multi-meter. I think there could be people tossing out PSU's thinking they are bad when in reality, their mobo monitoring is off.

[zerodamage]

Originally posted by Vampyromaniac
Of course the stress from the CPU/chipset/RAM/Video card were removed during the measurement, were all your drives still connected and running when you took the measurement?

I tested this while my CPU was running. That screenshot of my rails is off the 20-pin and verified through an unused power plug while my system was on. I tested right off the mobo.

My system was being loaded via running video capture card, watching tv, running 2 games in the background. I put on a little more than above load to verify.

[Vampyromaniac]

Oh! Very well done

[zerodamage]

Originally posted by Vampyromaniac
Oh! Very well done

Thank you!!

I'm smelling a dh guide......

[brc64]

I don't normally frequent this forum, but I came here to post a weird problem I was having and noticed your thread. I have the same motherboard as you, and interestingly enough, we also have the same power supply. I've been worried about my rails for a long time, but after your multimeter test, I feel a little bit better. My USDM screenshot looks almost exactly like yours (my -12 is sometimes slightly lower, but all of the other numbers were extremely close). I was worried that my power supply might need to be replaced, but it looks like I may be okay.

Thanks, Lee.

- Me