• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

[ASUS P5B-Deluxe/Wifi] - Vcore spikes

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

avatar1983

New Member
Joined
Dec 13, 2007
Hi all,
I'm new around here and I have a question about this board.

I've been running a conservative stable overclock for many months of 320Mhz fsb with my E6600, resulting in a 2.88Ghz cpu frequency at 1.3Vcore, with the ram on 4:5 resulting in its rated 400Mhz. Recently I decided to see if I could push things a bit further. I've upped my overclock to 400Fsb, with ram on 1:1,and the cpu multiplier set to 8, for cpu frequency of 3.2Ghz and ram remaining at 400Mhz.

So far this has worked quite well, except I now occasionally get vcore spikes, both up and down. The up spikes will reach about 1.8V for a second or so, and the down spikes 0.4, (all voltages as reported by asus pcprobe) again only for a second or so. This happens only under load and even then infrequently, maybe one spike every 2 hours or so. They do not seem to be crashing any programs, and they don't cause Prime95 failure either.

I have experimented with slightly higher vcore settings then my usual 1.3, but that does not seem to make any difference. Temperature wise, I top out at 50° under load for the CPU, and about 32° for the motherboard sensor.

Any idea what could be causing these, and what I can to prevent this?
 
Ok, I will try with different tools. I was just surprised by this, since I never got these spikes before.

I have run a 10 hour orthos test last night, and that passed just fine. So whatever it is, it does not seem to affect stability.
 
if a piece of software is reporting a drop in voltage to 0.4v and your system doesn't lock up...then it is a bad reading. there is no way it could handle a voltage drop of that low. as the others have stated, double check with CoreTemp as it keeps a log but instinct tells me it is the CPU temp sensor that PCProbe is reading from.

either that or see if there is an update to PCProbe.
 
Ok, I have done quite a bit more investigation. The tools used to get these readings, by the way, do not have any effect. Asus PCProbe, SpeedFan, CPU-Z, same story everywhere.

I'm inclined to concur that something odd happens to the sensors at 400Mhz fsb, probably the Winbond sensor controller chip. Here is why:

As 3oh6 said, the fact that the system stability is entirely unaffected makes me think its a bad reading. I have also seen some very rare occurrences of the 3.3V reading briefly dropping to 0, with no effect on the system. And I have seen possibly the strangest of all: The CPU temperature reading goes odd sometimes. This will only happen after the system has been on for a a few hours at least.

What happens is this. The reading will flip from is proper value, say 40 for typical game load value, to -1 and then up to 80. First time this happened I was quite concerned, as you may imagine, and quickly turned of the computer. Upon further investigation (thank you speedfan log!) it turned out however, that the temperature reading delivered by the individual core sensors remained at their usual values, around 30 in this case. And after the system is restarted, the reading goes back to its proper value. When this problem occurs, an offset of 40 seems to be introduced to the cpu temperature reading. I have since discovered that this will right itself after a few hours. Strange or what?

So my new question is this. Has anyone else seen a sensor chip go so strange? Is there anything I can do to improve it's behaviour? Cool it maybe?

Thanks for all your help so far.
 
Back