• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Inconsistent CPU Voltage From Different Sources Including Probe

S

SiRCYRO

Weaksauce
Joined
Jan 21, 2003
Messages
111
mI'm currently running the rig in my sig, i7 920 @ 4200MHz 200MHzx21, CPU voltage set to 1.3563. I've been doing some stress testing, trying to find a sweet spot for my CPU and I've been using a bunch of different programs to monitor temperatures and voltages.

At first I was just using Core Temp and CPU-z for temps and voltage then I started looking at Everest (for a completely different reason, to make sure every core/thread was at 4200MHz). I noticed that while idle my CPU voltages in all 3 programs were in the same ball park. When stress testing CPU-z showed a bit of VDROOP while the voltages in Everest increased.

Since I have a Rampage II Extreme motherboard I decided to go through the built in LCD to monitor voltages, which also seemed to go up versus down under load. I also hooked up my multi meter and measured the CPU Vcore with the provided probes and the result was even higher.

Here's some numbers.
BIOS Vcore 1.3563V

Idle:
CPU-z 1.352V
HWMonitor 1.35V
Everest 1.35V
LCD Monitor 1.343-1.349V
MM Probe 1.353V

Load (approximately 30 minutes into Prime95):
CPU-z 1.344V
HWMonitor 1.34V
Everest 1.38-1.40V
LCD Monitor 1.376-1.389V
MM Probe 1.396-1.398V

So if I was to hazard a guess, I would say the included probe would provide the most accurate results. Perhaps there is some overhead in the results derived from Everest/LCD/Probe as I've only ever really encountered VDROOP. I wonder what the degree of accuracy is with many of these programs. Others have said to test their values with actual values, which I guess is what I've done.

I was going with just my CPU-z/HWMonitor results and thought I had a decent over clock at close to 1.35V. With these new results, perhaps CPU-z is not as accurate as I had though and maybe I should tone down the overclock. My temperatures are respectable, 77C and under for Prime95 water cooled.

Should I tweak my overclock/lower it? I'm thinking I should still be safe even with 1.398V that the probe is topping out at if that is the true voltage, I don't want to damage my processor.
 
I desperately want to say to trust the DDVM and since the LCD monitor is closer to the DDVM than the software that is what I would go with.

The question I would have is do you have the Load leveling (vdroop elimination) or whatever Asus calls it enabled in the bios ?
(this is the only reason I can think of as to why Vcore would increase at load - load leveling increasing Vcore to compensate for IR losses)

2nd question (perhaps Asus tech support could answer) is exactly where does the DVM contacts go ? Do they go to the Vcc pins at the CPU socket (that would be actual Vcc and what I would want) or do they go to some other part of the CPU voltage regulation circuit.

I guess for now, if it where me, the safest thing is to assume the highest values are correct.

http://download.intel.com/design/processor/datashts/320834.pdf

Absolute Vcc max per data sheet is 1.55 V. (Table 2-6)
Per note 2 I would reduce that by the 50mV overshoot allowed in the voltage regulation circuitry so that leaves 1.50. And because I am conservative I would stay under 1.45. But thats me.
 
1. I do have Load Line Calibration enabled in BIOS which is Asus' vdroop eliminator. At all speeds/voltages I've tried the system is not stable without it. With vcore still set to 1.3563V in BIOS I get idle readings of:

CPU-z 1.336V
HWMonitor 1.34V
Everest 1.33V
LCD 1.336V
MM Probe 1.339-1.340V

As soon as I start Prime95 the system locks up, obviously the voltage is too low when it's set to over 1.35V. That would make sense that Load Line Calibration is causing the voltage to increase, too bad with it off it starts below 1.35V already. I tested other terminals such as DRAM voltage and that stayed steady under load which I am happy about.

2. I am curious about this as well. I'll have to put in a ticket with Asus see if they can give me a definitive answer as to where the contacts go. Maybe there is an offset due to where it is connected.

So by that data sheet, even if the highest reading is correct I'm still in the safe/minimizing risk of damage. Good to know.
 
You must log in or register to reply here.
Back
Top