Do IC7/G/Max3 boards *really* undervolt ?

We all know the hardware chip says it undervolts the voltages (thus the droop mods), but does it *really* undervolt?

Has anyone actually taken a DMM/Multimeter and tested the core voltage, to see if it is really undervolting the cpu voltage?

Do the droop mods actually stop the undervolting, or stop the monitoring chip from reporting low voltages?

I can't say for the core voltage, but the vdimm actually overvolts while software monitoring shows undervolting. (ie 2.8v in BIOS = 2.83-2.85v actual). I've read posts from others that the BIOS settings for vcore are nearly spot on at idle when measured with a DMM, but I haven't measured it myself as I believe the measurement point is on the back of the board.

I was reading on another forum, posted by hipro 5 (guy that did a lot of ic7-g) mods, that the voltage is dead even, and that the monitoring software is off.

I am not sure I agree, because my current cpu always freezes right after a drop in vcore when it's being pushed hard.

We're not saying that the voltage doesn't droop under load - it does. But monitoring software shows lower voltages than what can be measured on the board. The "dead even" comment means that the voltage setting in the BIOS and the measured voltage on the board are nearly the same at idle, not lower.

Do new northwoods use VID?

@KrisMCool
We need someone to test the voltages at -load- on something like an IC7 series board, to see if the voltages do drop at load. (the voltages the board is sending the CPU, NOT the voltage -inside- the CPU.)

Note that according to the Intel VR-Down specificiation, the Northwood CPU's drop the incoming voltage by a small amount, the larger the load, the more the vcore will drop.

But the voltage the motherboard is sending the CPU shouldn't drop under load. (unless the board is giving you the cpu voltage that is -internally- read by the chip itself).

Anyway, can someone with a dmm/multimeter test this?



@jenko:
Yes they do, but the motherboard can still send a too low voltage, etc, or a higher one.

For instance, if the max VID on a chip is 1.55, then how can a voltage of 1.625 work, without the CPU supporting that VID? (that's grounds for technical people to answer).

That's a good point, Falkentyne. I didn't realize there would be a difference in the external voltage supplied to the CPU and the voltage used internally by the CPU. When I muster the courage to install my Madcat vtt mod circuit, I'll also attach a read wire to the vcore measurement point (I don't remember where I saw it, but there are pics of where to attach a wire on the back of IC7 boards for vcore measurement). I solder copper nearly every day (plumber), but my hands shake badly so small wire soldering would be difficult. I'll try to find a link to the pics I spoke of.

According to some posts (another site) from Hipro5 (IC7 Mod God), vcore can be measured at the base of the vcore chokes surrounding the CPU socket. Check out his IC7 Mod Guide.
http://www.thelab.gr/reviews/ic7mods_en1.php

KrisMCool said:

That's a good point, Falkentyne. I didn't realize there would be a difference in the external voltage supplied to the CPU and the voltage used internally by the CPU. When I muster the courage to install my Madcat vtt mod circuit, I'll also attach a read wire to the vcore measurement point (I don't remember where I saw it, but there are pics of where to attach a wire on the back of IC7 boards for vcore measurement). I solder copper nearly every day (plumber), but my hands shake badly so small wire soldering would be difficult. I'll try to find a link to the pics I spoke of.

Click to expand...

Great.
I typed a 20 minute reply and it said "error: server load too high" and everything was lost. I won't bother to retype it, it was too much work.

Anyway, I'll just ask if anyone can please test the voltage being sent through the motherboard at full load. Hipro? Anyone else with a multimeter who can test?

And also : the motherboard voltage you set in BIOS: that is the VID, right? And the voltage readout you get in monitoring (BIOS/software), that is also the VID that the board is sending, right?

and about the post I wrote that was lost:
It concerned the relationship between VID, and the vcc (the voltage the chip is using or is reduced). That is on intel PDF 29864312, you can see for yourself.

Hey guys, I've been posting all overe these forums that the Vcore is spot on on my voard when measured with a DMM I actually soldered a wire to one of the caps in the vcore circuit. It does drop as you would expect, about 0.02 under load, but than it's rockstable. Anyways.. read through my posts, and you'll find the whole story.