How to read the rails from multimeter

[neologan]

Hi guys. I was wondering if someone could tell me how should go about testing the 3.3v, 5v and 12v rails using a multimeter.

I have a DFI Ultra-D if that helps at all.

Thankyee :)

[The Mofo]

Hi, its quite easy... If you have an extra molex flopping around your case, go ahead and grab it. With the computer on, running 100% load, i use D2OL as my load, plug the red lead from the MM into the red lead of the molex.. Then black lead from the MM into one of the black leads on the molex. That will give you your reading for the +5V. For the +12V, move the red lead from the molex and place it into the Yellow hole of the molex. To get your 3.3+ value, it gets tricky. Locate the 2 orange wires on the 20/24 pin connector of the PSU to motherboard. Ground the black lead and place the red lead of the MM into each orange line.

You get a more accurate reading at 100% load. If your PSU is adjustable, you can simple raise the rails to make up for the low readings. Good luck!

[neologan]

Thanks Mofo, that 3.3v read out sounds rather tricky indeed...

[EMC2]

You can read the 3.3V rail on any of the bottom 4 jumpers for the Vmem source selection header (top right of board) when set to use the 3.3V rail (bottom leftmost 4 pins if setup to use the 5V rail). All orientations given with DIMMs on top, PCI cnxs on bottom as if installed in a chassis.

Peace :toast:

[T1Cybernetic]

Whats the best way to read all the rails, :confused: idle and under load?
Im running a nf7-s v2.0 on an antec truepower 430 if that helps at all?

I've done some googling etc but i've yet to find anything that helps, :(
im just curious thats all because i've just changed motherboards and
bios and software readings are reading are completely different from
when i had my last nf7-s, :confused: when i had the old motherboard my
readings were just what i thought were acceptable but since i got this
new board my voltages are better and my temperatures are a little better.

Im hoping that it's because it's a newer board but i thought id test anyway.

[neologan]

Thats much easier EMC2. Here my rails from these readings:

+3.3v = 3.36
+5v = 5.17
+12v = 12.14

This is with the following hardware:

A64 Winnie 3200 @ 2.6 (1.67v)
DFI Ultra-D
ATI X850XT PE (600/600)
OCZ 512MB 3200 EB

@ T1Cybernetic - Motherboards determine the monitoring of rails from software, so changing this will no doubt give you different results. Its probably the bios update but overall it doesn't really make much sense does it. The reason I've given up on software monitoring of rails is simply because it is not accurate at all.

[The Mofo]

Youre right, software is far from accurate. My Zeus read in software that the +12V rail was 11.77, +5V was at 4.81 and the +3.3 was at 3.01

The MM proved this theory wrong of course.. I did have to give a slight bump to the 12 and 5 rails with the delta monster i have turned all the way up.

[T1Cybernetic]

Oh yeah i know i know that software isnt as acurate and reliable
as it should be thats why i asked, How to read the rails, ;)
so that im able to get a much better reading, :fact:.

So it's just basicaly read from the molex connectors and hot spots
Ie: from reading various volt moding guides from around the net etc
while under full load and from idle /startup's etc and that will work ok?

[Squid_Spit]

here's a great guide http://forums.extremeoverclocking.com/showthread.php?t=137886

[neologan]

I noticed i can get all three reading from a SATA power connector (if you have any) With this method the +5v and +12v gave identical readings to the 4pin molex method. But the +3.3v reading was higher than the DFI Vmem source selection header method, coming in at 3.39v. So what does this mean? Its probably means the DFi method is not 100% accurate either.

[SoylentGreen]

I noticed i can get all three reading from a SATA power connector (if you have any) With this method the +5v and +12v gave identical readings to the 4pin molex method. But the +3.3v reading was higher than the DFI Vmem source selection header method, coming in at 3.39v. So what does this mean? Its probably means the DFi method is not 100% accurate either.

I think what you are running into is that Vmem is generated on the motherboard. The +3.3V at the SATA connector is generated in the main PSU, so they are different supplies. +12V and +5V are probably only generated in the main PSU so they are identical.

[Flib]

what you mean with generated in the main supply?

I measure my 3.3V Line likte this:

(System must be off!!)
Take the P4 Connector, put there a wire in to have contanct with the orange wire. Take some tape and tape the other end of the wire to your red lead of the DMM.
Easy ground the black lead of your DMM.

[EMC2]

Nope... the spot I pointed you to is directly connected to the +3.3V input at the 24-pin ATX connector of the MB and the best place to check it. Checking the voltages at the SATA cnx (or a HD molex cnx) gives you a misleading reading... those connections are not under load (have no current flowing through them) and thus don't account for the IR drops across the wires and connectors that your MB sees. The best place to read any PS rail is on the MB.

--- Flib ---

You can actually save yourself some trouble if you have good probes with slender leads... you can slip them right down into the back of the ATX connector on the MB ;) Not quite as good as a spot on the MB (doesn't take into account the drop across the cnx), but much better than a HD cnx.

Peace :toast: