Temps and vDroop questions regarding E7200 overclock

[tbhuang2]

I have an E7200 @ 3.6GHz (424 x 8.5) with 1.2375v on the vCore with ~1.20v vDroop, all cooled with a Xigmatek HDT-S1283. AS5 has been applied in 3 dabs along the 3 heat pipes and it has been about 5 days since the inital application. Right now, with an ambient temp of about 27C it's idling at around 43C and load can go up as high as 64C. The room I'm sitting in right now is a comfortably cool temp, maybe 24C.

#1) Are these temps normal? I read customer reviews on newegg reporting around 50C load with similar overclocks... Should I bother reseating the CPU?
Additional info: when not overclocked, idle is around 40C, load is 53C.

Screenshot at load for additional info:

#2) Also, the "system temp" as noted in HWMonitor is kind of quirky. As soon as the CPU is under load, it rises to over 50C! Similarly, as soon as Prime95 was stopped, it dropped to around 30C. What's up with this?

"Idle," taken very shortly after stopping Prime95.

#3) Finally, a question about vDroop. As stated above, I understand that this is normal. However, is an additional vDroop while under load normal? Under load, vCore drops from 1.20v to around 1.184v as seen in the screen shot. Is this normal or is my PSU inadequate?

Much thanks in advance!

 

[Shiggity]

That idle temp is way way too high, should be in the mid 20's - low 30's C. Try a reseat / lap if necessary. (Is the chip or heatsink concave in any way and not making good contact anywhere?)

Not sure about the vdroop question.

 

[ziggo0]

Can you try using Realtemp? Open it before you start Priming, let Prime finish with 8k and let me know the Load temp.

I'm sitting at about 63C with my e7200 @ 450x9, 1.4vCore under water.

 

[jason4207]

The drop from idle to load is vdroop. The drop from the BIOS setting to what you see in CPU-Z or OCCT is just a badly calibrated BIOS setting...it really has no consequence other than you having to set it higher to compensate.

System temp is probably your North Bridge.

RealTemp should give you more accurate temps for the CPU.

 

[tbhuang2]

Thanks for your replies, everyone.

@Jason, thanks for clearing up the vDrop vs. vDroop. I didn't look into them enough and assumed they were synonymous.

I'll install RealTemp and see what I can do. I have to admit, the whole "Calibration" process seems simple enough but still a little daunting.

 

[Craxbax]

I am about to start a gaming build based on an E7200/Arctic 7 on a P5K pro and Zotac 8800GT for friend. Do the E7200 run a little hotter than the E8x00s at the same mhz?

 

[tbhuang2]

ziggo, I installed RealTemp (didn't touch the calibrations) and immediately noticed that it reports much lower temperatures. Idle is now 35C, load is now 55C. Included is a screenshot of both RealTemp and HWMonitor running at the same time. Very interesting... Should I do any further investigation? I don't know if I'm comfortable knowing that one program reads temperatures almost 10C cooler than another...

 

[blazed]

95C tjunction max vs 105C tjunction max, use RealTemp. However the only real way of knowing the condition the CPU is in is monitoring the DISTANCE to tjmax, realtemp provides this in its top row. Keep it at least 20-30C away from tjmax and your good.

BIOs to idle is vdrop and idle to load is vdroop, collectively generalized as just vdroop. The lowest ammount of voltage CPU-Z reports during your stress testing is the CPUs effective load voltage, thats the least it gets when it needs it most. Compensate for it by using higher BIOs voltage settings.

 

[jason4207]

Most programs report Tjmax to be 105*C. After extensive testing the author of RealTemp concluded that Tjmax is closer to 95*C. That is why you're seeing the 10*C difference.

Some folks have even theorized that Tjmax is different on each chip. No way to know for sure b/c Intel is very tight-lipped on what the actual value of Tjmax is.

Blazed is correct in that 20-30*C Delta to Tjmax is probably the best way to look at it. Once you start to get near a max OC temps that high will keep you unstable anyway.

 

[wingman99]

.Some folks have even theorized that Tjmax is different on each chip. No way to know for sure b/c Intel is very tight-lipped on what the actual value of Tjmax is.

On a E 8400 it's at 94c then speed step kicks in, I tested it below

 

[jason4207]

On a E 8400 it's at 94c then speed step kicks in, I tested it below

All you've shown is that your CPU throttles at 1*C to Tjmax. You can change Tjmax to any value you like and you're CPU will always throttle at 1*C to that value.

There is no way to know absolute temps unless you take readings w/ a thermistor or similar. Tjmax is still unknown.

 

[wingman99]

All you've shown is that your CPU throttles at 1*C to Tjmax. You can change Tjmax to any value you like and you're CPU will always throttle at 1*C to that value.

There is no way to know absolute temps unless you take readings w/ a thermistor or similar. Tjmax is still unknown.

I could not get the E8400 hotter, i even disconnected the fan so that must be Tjmax.

 

[JamesXP]

tjmax is without a heatsink, and not measured with software.

 

[blazed]

All you've shown is that your CPU throttles at 1*C to Tjmax. You can change Tjmax to any value you like and you're CPU will always throttle at 1*C to that value.

There is no way to know absolute temps unless you take readings w/ a thermistor or similar. Tjmax is still unknown.

lol yes indeed. Fine experiment none the less wing, seems your thermal throttling is in working order.

 

[wingman99]

lol yes indeed. Fine experiment none the less wing, seems your thermal throttling is in working order.

I like to check it incase of a emrgency, i don't want it to over heat..