Normal for Corsair 520HX 12+ line to drop to 11.58 under load?

[poohbear]

Hi all, i just "upgraded" from a 420wt enermax noise-taker 2 psu to a Corsiar 520HX cause i'm planning on going SLI.

The thing is my old enermax NT2 showed regular readings for the 12+ line @ 11.9 idle and 11.78 under load. So, i thought the Corsair would best this or at the very least match it. But now after installing it my 12+ line drops to 11.58 under load and idles at 11.84. What gives?? i've been reading review after review stating this corsair psu is a stellar, high quality psu, but why is my voltage regulation so crappy?

I've been getting these readings through Smartfan, both for my old Enermax NT2 and my Corsair 520hx. Is my Corsair faulty or is this normal?

Thanks in advance.

 

[HOOfan 1]

Form Letter Auto reply-

Software readings are not to be trusted, if you want accurate readings you will have to use a multimeter.

 

[poohbear]

sure, but i used the same software to measure my enermax, and that software measured a drop in voltage, so regardless of the software/hardware im using, it still measured a drop between the 2 psus, yea?

 

[dangman4ever]

Again, software reading should be ignored. No matter what your software readings show, they're still HIGHLY inaccurate and should not be used as a gauge of any sorts.

 

[poohbear]

ok sure its innaccurate, but its showing a huge drop in voltage between the old psu and new one, so its still measuring a drop, regardless of whether its not the actual voltage. Regardless of the measuring instrument, its detecting a drastic change in voltage regulation.

At any rate, is 11.58 under load on the 12+ line something to worry about?

 

[HOOfan 1]

[ObiWan]Software readings can decieve you, do not trust them[/ObiWan]

therefore you cannot compare the software readings for two different power supplies. Even for the somewhat more trustworthy Kil-A-Watt meters they will give fairly accurate readings for some PSUs but completely inaccurate readings for other PSUs. If you want to accurately read the voltage on your PSU...buy a multimeter.

To answer your last question, ATX specification allowed + or - 5% deviations so 12x.05=.6
12-.6=11.4 12+.06=12.6 Thus as long as your voltage is between 11.4 and 12.6 it is within ATX specification.

With that said, I would still expect the voltage regulation on the HX520 to be better, but as already stated, software readings are not reliable.

 

[jonnyGURU]

It's not so much that software is more accurate reading some PSU's and less accurate when reading others. It's because the equation used to calculate voltage is exponentially off the farther you get from median voltage.

That said, unplug your 24-pin connector, wiggle the wires around real good, and then firmly replug it into the motherboard and see what the software says.

 

[poohbear]

thanks HOOfan and johnnyguru. I tried what u suggested jonny but it didnt make a difference. Didnt wanna wiggle the 24pin wires too hard as i was scared one might come out (had that happen b4 and it shortcircuited), but wiggled them well and still same readings.

guess i'll just have faith in corsair and buy a multimeter down the road if there are any probs.

 

[magreen]

Where would you stick the leads to your multimeter to test under load? into an unused molex from your psu? But since it has multiple rails, do you have to find a 12V from each rail to test?

 

[jonnyGURU]

Originally posted by: magreen
Where would you stick the leads to your multimeter to test under load? into an unused molex from your psu? But since it has multiple rails, do you have to find a 12V from each rail to test?

http://www.bfgtech.com/bfgpowe...shooting.html#software

And even if you have multiple +12V rails, there's still only one source for that power. Any voltage drop would only be contributed to resistance between the PSU's PCB and the end of the connector.

 

[magreen]

cool, thanks

 

[SparkyJJO]

I had something similar when I changed PSUs, software readings were quite a bit different. Multimeter proved otherwise. Software voltage readings = crap, I've found you can't even use them in the same system to compare different PSUs.

 

[poohbear]

Originally posted by: SparkyJJO
I had something similar when I changed PSUs, software readings were quite a bit different. Multimeter proved otherwise. Software voltage readings = crap, I've found you can't even use them in the same system to compare different PSUs.

thanks sparky, that's reassuring. Cheers for that.

 

[Modelworks]

Think of it like this.
You have a long piece of metal that gets thinner as it gets longer to the point it is a hair thin. If I measure at the end that is closest to the source I get a high reading. If I measure it at the other end I get a low reading. That is the problem with software monitoring. You don't know where the device doing the measuring is located. If its just on a circuit board trace with a tiny tiny connection at the very end, its always going to read wrong as opposed to the things attached at the start where the power supply connects.

A lot of boards place the chip doing the measuring on the side of the board where the front panel connectors are for reset and power, usually towards the back. That is a long way from the power connectors to the board and the cpu power supply.