• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Question about safe voltages

lbratch

lbratch

lbratch

Associate
Joined
13 Jul 2006
Posts
117
Location
Jersey
I'm currently running at 1.57V through my E6700 Conroe, and I'm getting 40 idle, 57 max under load.

Is the maximum safe voltage just to do with temperature?

Ie. can I run safely at 1.6V or 1.7V, as long as the temps are low enough? (Say below 65)

Or is there a voltage safety wall I will hit that will cause damaged regardless of the temp?
 
lbratch said:
Is the maximum safe voltage just to do with temperature?

Ie. can I run safely at 1.6V or 1.7V, as long as the temps are low enough? (Say below 65)

Or is there a voltage safety wall I will hit that will cause damaged regardless of the temp?
Click to expand...


No
No
Yes

Also maximum thermal limits before damage drop everytime you raise the voltage, so for example a cpus max limit is 100C at 1.4v, it may drop to 80C at 1.5v, as processes shrink the max voltage level obviously drops. Would'nt be too comfortable at anything over 1.475v-1.5v on air (just my opinion, could be way off). One of the best documented cases of voltage killing chips quite fast are with earlier Northwood chips, many were even phase cooled but still died, hence the term SNDS - sudden northwood death syndrome. Its really to early to draw conclusions on max core 2 voltages as they haven't been around that long.
 
Last edited:
1.55v is the maximum specified by Intel in the design sheets, with TCaseMax of 60.1C. Any more and you run the chance of killing it, but then again it could last ages.

As voltages rise about the design specs the temperatures rise very quickly, mine something like 6/7C for every 0.05v. Assuming you can cool it efficiently, high voltages regardless of temperature cause electron migration. Basically it errodes the tiny copper conductors until they break. Still, its a bit like "how longs a piece of string"

http://forums.overclockers.co.uk/showthread.php?t=17574496
 
Last edited:
fornowagain said:
1.55v is the maximum specified by Intel in the design sheets. Any more and you run the chance of killing it, but then again it could last ages.
Click to expand...

The Absolute Maximum Voltage is the highest voltage the chips can take for a short period of time. Extended use of the processor at or near these voltages will cause permanent damage to the chip. Reliable operation of the chip outside its nominal voltage range is not guaranteed
Click to expand...

http://users.erols.com/chare/elec.htm
 
Eh, I'll rephrase not clear enough. What is safe? The quote above is for wimps, how bright can a lamp run before to much current burns out the filament? It will take more without failing, but yes it reduces the lifespan and increases the chances of a spontaneous failure. There is a point it just dies. Intel design it to last a decade or more.

The maximum Intel stock voltage is 1.35v. Do you limit yourself to that for an overclock, no. 1.55V come from the extreme top of the curve for 65nm. And the maximum voltage based on Tc case TDP. Any more and from an overclocking point of view your on very dangerous ground. It will run at 1.55v for an extended period, but it's shortening its lifespan for sure. How long, who knows. There were some chips awhile back started dropping like flies after a few weeks. Then again I've seen massive overvolts last for years.

http://download.intel.com/design/processor/datashts/31327801.pdf

 
Last edited:
Yea i setup a mates XP 2200 at 2.3GHz on water, Tbred A at 2.1v for 2yrs and its back to stock in a mates rig now just dandy, death can be pretty random but with my own chips i love being a wimp, no point with massive overvolts for minor speed differences. Used to run an AXIA 1Ghz at 2.1v as well at 1.8Ghz on phase with no ill effect, but speed difference is so unnoticable these days i don't even bother going over stock volts by far.
 
Justintime said:
Yea i setup a mates XP 2200 at 2.3GHz on water, Tbred A at 2.1v for 2yrs and its back to stock in a mates rig now just dandy, death can be pretty random but with my own chips i love being a wimp, no point with massive overvolts for minor speed differences. Used to run an AXIA 1Ghz at 2.1v as well at 1.8Ghz on phase with no ill effect, but speed difference is so unnoticable these days i don't even bother going over stock volts by far.
Click to expand...
Yeah your right, as I'm sure you know, play with a chip for awhile and you can "feel" or plot F vs T, when the temps start to climb very quickly for a small increase in frequency, that's where its really dangerous when the curve head straight up. Looking at the TDP helps, extend the curves a bit. Design profiles of the chips can be (I love conroe) quite safe for massive overclocks without any real harm. The manufactures plans for the future, higher and higher clocks with small revisions based on the same silicon.

What gets me is the noobs just wacking up the volts for ten second screen grab, now that's just silly, it is damaging the chip. Moving to phase, or LN2 moves the whole curve down. I think small increments and lots of stability testing is the only safe (ish) way. But each to their own.
 
Last edited:
Soulja said:
i wouldn't want to go that high tbh, what have you got it clocked at? mines at 3.6Ghz 1.5v
Click to expand...
Could you give me any more details on your setup?

Even at 1.55V I'm not completely stable at 3.6GHz, but with 1.57V all is fine.
 
I wouldn't want to run more than 1.5v on air personally, not if you intend to keep the processor for a couple of years.

Overvoltage will kill a CPU over a period of time, regardless of the temperature. I've seen first hand P4 CPUs that have had 1.75v through them and have suffered electromigration as a result and they're just completely unstable (causing BSODs all the time).
 
Last edited:
lbratch said:
Could you give me any more details on your setup?

Even at 1.55V I'm not completely stable at 3.6GHz, but with 1.57V all is fine.
Click to expand...

Asus P5W DH Deluxe WiFi, GeIL 2GB (2x1GB) PC6400 800MHz Ultra Low Latency DDR2 Dual Channel Kit ,Thermaltake Big Typhoon 4 in 1 Heatpipe, runs fine at 1.5v
 
You must log in or register to reply here.
Back
Top Bottom