Fluctuating Vcore

[antipop]

I have a nf7-s and a TTGI 420W psu (verything is in sig) and my vcore is making jumps! It's set @2V in the bios but it's varying from 1.95 to 2.02 i've even seen quick jumps to 2.06! Should i be worried? I mean it's varying quite oftenly between 1.97 and 1.98. Can it cause instability at higher clock?

[TheDude]

I had the same problem with the 520 model. I upgraded to TrueControl 550w and problem is gone. It allowed me 100 mhz more overclock too. Cleaner power I think. :D

[Holst]

This might be the onboard sensor fluctuating not the vcore.

Do you have a decent multimeter to test ?

[antipop]

Originally posted by Holst
This might be the onboard sensor fluctuating not the vcore.

Do you have a decent multimeter to test ? I don't but maybe i'll get on to test

[Lunatic]

You could try to cool the mosfets, that can help :)

[felix88]

Holst, where do you measuer Vcore on the NF7-S V2.0?

[Holst]

I prefer to use the CPU power plane on the back (or front of a board)

I dont have an Abit bord myself, but if somebody posts a pic ill take a look at the layout. Try testing one of the 6 caps next to the socket or the fat trace they connect too.

[felix88]

i've got a few pics, unfortunatly they aren't the greatest quality.

www.canby.com/chale/images/NF7-S/NF7-S1.JPG
www.canby.com/chale/images/NF7-S/NF7-S2.JPG
www.canby.com/chale/images/NF7-S/NF7-S3.JPG

careful, they are 1600x1200. ;)

[Holst]

You can either grab it on the mosfets directly.

Or on the ends of the coils oive circled in red or on the traces the coils solder onto.

Be VERY carefull where you put the probe, if you short between the mosfets and ground or another component you will kill your board.

You might have to put your PC on its side or even take the oard out of the case to reach.

With my own boards I solder some wire to the Vcore, VDD, VDIMM and put them into a plasic wire holding block thingy, so I can easily probe voltages when my PC is in use and in the case.

[felix88]

so i can just measure the Vcore right off of the mosfets or from the inductors(the coils)? sounds easy enough.

how about points to measure the VDD and VDIMM(if you also own the NF7-S)?

thanks for your help Holst!

[Holst]

I dont have either an NF7-S 2.0 or 1.0 but I can point out measuting points if you gimme a pic.

VDD and VDDR are a bit less clear but I think this is pretty close.

Maybee somebody who has the board can tell us for sure :)

[felix88]

cool, i'll check those out.

one last question for you. on the mosfets, where do i measure the voltage? i've heard some people say the middle leg, but in the picture you posted above, it looked like you highlighted the metal plate attached to the back of the mosfet.

i assume you can use the case as your ground.

[Holst]

The middle leg is the same as the flat plate.

Using the middle leg is MUCH more awquard and easyer to short your board so its best to use the fat leg or a coil.

[felix88]

yeah, i tried it. thanks very much for your help!

either there are voltage drops between the mosfets(i measured from the leftmost mosfet, if that matters) and the CPU, or the hardware monitor is VERY inaccurate. i measured 1.88v for my Vcore, and the hardware monitor reports 1.82v. it's set to 1.85v in the BIOS.

also, the 2 mosfets you circled to the left of the AGP slot one reports 1.59v and the other reports 3.28v. i assume the 3.28v is PCI voltage, and the 1.59v is probably AGP.

where you have VDDR and VBT marked, there aren't any mosfets. so i'm not sure which pins to measure from on the IC.

here are my best guesses:
http://www.canby.com/chale/images/NF7-S/NF7-V.jpg

[Holst]

You see where I marked CDD and CDDR there are two 8 pin ICs, try testing the legs closest to the DIMM sockets.

Be carefull not to short anything..

[felix88]

i might need to get some different probes if i'm going to try testing the legs of such small ICs. i'll just go without it for now.

thanks again Holst!