How low is too low for the 12v line?

[Ketchup]

I have a Skyhawk GM570PC 570-watt power supply and am using Speedfan to monitor my voltages. I noticed that the 12v line stays down around 11.71 while I am gaming and goes back up to 11.9 in windows. Is 11.7 volts cause for concern?

 

[John]

You should ignore software readings. Spend $10-$20 on a DMM to test the rails.

The rails should be within ±5%

 

[Zepper]

Yeah, Sears has their $20. DMM on sale t his week for around $11. Every home should have one...

.bh.

 

[Navid]

Zepper,

Please let me know if this is not the one.

http://www.sears.com/sr/javasr/product....ers+%26+Accessories&BV_UseBVCookie=Yes

 

[dBTelos]

"The rails should be within ±5%"

So that would be 11.4V is the lowest you can have while staying in spec? Similarly 12.6V is the highest you can go while staying in spec?

 

[eelw]

I feel a 10% tolerance is acceptable.

 

[w00t]

Originally posted by: John
You should ignore software readings. Spend $10-$20 on a DMM to test the rails.

The rails should be within ±5%

I am going to have to go with John Here

 

[JBT]

Originally posted by: w00t

Originally posted by: John
You should ignore software readings. Spend $10-$20 on a DMM to test the rails.

The rails should be within ±5%

I am going to have to go with John Here

agreed

 

[John]

Originally posted by: eelw
I feel a 10% tolerance is acceptable.

±10% is only for the -5 & -12v rails. The main +3.3, +5 & +12v rails (the ones that people typically measure) need to be within ±5%

Scroll to section 4.1.4
http://www.formfactors.org/developer/specs/atx2_2.pdf

Most high quality psu's have their tolerance @ <2%

 

[Zepper]

Navid,
. Yes, that's the one.

I concur with eelw or +/- 1V is fine in real life usages. The +12 rail is not used to power anything but drives and fans directly. For all other purposes it is reregulated down to a much lower voltage and filtered locally so as long as there is maintained a substantial ratio between the input voltage and the output voltage of the local regulators and the 12V rail is capable of providing adequate current, the output voltages of those sub-regulators will be just fine.

And the servo motors on drives can easily compensate for +/- 1 volt. The standard on the +12 used to be (and not long ago either) 10%. IAC, the standard for the +12 now is at +/- 5%. So if it measures outside that into a reasonable load (2A on the +12 and 3 to 5A on the +5 - yes both should be loaded for accurate regulation) with an accurate DMM, then the PSU is RMA-able. I don't know why software is so unreliable for voltage readings - it must be the cheap health monitor chips used in PCs these days.

. OTOH if you're measuring without a load (or too little of a load - the higher the Amp rating of a rail, the heavier the load needed for accurate regulation) then you won't get accurate readings as the regulation won't be working properly on most switching PSUs.

.bh.

 

[SparkyJJO]

Originally posted by: John
You should ignore software readings. Spend $10-$20 on a DMM to test the rails.

The rails should be within ±5%

Agreed, software reads my 12V line at anywhere between 11.69 and 11.86 but a DMM shows that it is actually just about spot-on at 12.01V with none of the fluctuations that the software reads.

 

[lenjack]

Software often reads about .2v low.