• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Overclock a P4 on stock cooling? You HAVE to be kidding!

beatle

beatle

Diamond Member
Ok, so I'm firing up my IS7/retail 2.6C combo and after finding out that my TwinMos ram is complete garbage, I decided to see how much headroom the chip has by stressing it at stock speeds and stock vcore, then scaling up from there. I removed the black retail TIM and I'm using ceramique, applied as per instructions on AS's site for Intel cpus. Idle is 42C, which is not great, but better than my overclocked barton @ idle. Loading up 2 instances of P95 send the temps through the roof! Within minutes I'm brushing 60C, finally topping out around 61C. The NB (or PWM as Abit's hardware monitor calls it) is a blazing 54C! It reads my case/system temp @ 35C, but the temp probe on my Musketeer reads a cool 25C. Fans seem normal: 2400rpm for the CPU fan and 5700 for the NB fan.

MBM is also acting funny, sporadically sending a slight intermittant beep as if doing its alarm, despite the fact that I turned them off. I've since uninstalled Abit EQ after setting MBM 5.3.0.0's sensors to the ones matching Abit's temps.

EDIT: I shut prime off after temps had leveled off @ 61/54 and the temp on the CPU and NB dropped quickly. In under 60 seconds I was below the 50C mark on the CPU and the NB was close behind. I'm approaching idle temps as I type this. I'm going to verify I have the latest bios now. Yep, I do.
 
I'd be willing to believe it was user error in installing the cpu HSF since this is my first stock Intel install, but how do you explain the northbridge temp? 54C is hotter than my o/c'd Barton was under load!
 
Abit boards read anywhere from 5-10 degrees higher than they actually are. Don't worry about it. I've seen mine hit 69C before... The heatsink was barely warm to the touch.

As long as you're stable, temps don't mean anything.
 
Originally posted by: Nebor
Abit boards read anywhere from 5-10 degrees higher than they actually are. Don't worry about it. I've seen mine hit 69C before... The heatsink was barely warm to the touch.

As long as you're stable, temps don't mean anything.
Click to expand...

Exactly.

I bet your TwinMos RAM is Winbond CH-5, and that's where the problem is. Intel chipsets simply DO NOT like Winbond CH-5.
 
So the PWM is the NB chipset temp??? I though so....

54c seems a bit high...I notice at 3.276ghz w/ 1.63v (stable at 3.25ghz with same vcore) fails within 20 mins of TMPGenc (4hr encoding)...In that time the PWM hits 46-47c and the cpu hits 62c.....I then open side of case and crack window in office and let some of that 40f air come in....It drops the pWM to 38-39c and the cpu temps never hits 60c and stays around 58-59c....Now it does 4hr encode just fine as well as do other things during at least 2 hours of it....

I placed AS2 on the NB chipset early on and noticed a few degrees better with just that....

I was assuming my NB was too hot but you seem to be cranking 54c.....I thought maybe cpu but you guys are talking 69c.....system stays in the 27-28c range under load and 21-22c range with window opened....


What do you think...NB temps or CPU temps??? Tested it twice..Close window and let temps hit 63c cpu and 48c PWM and get an error...Crack window drop temps and 4 hour encodes.....



I have a swiftech MCX4000 with 92mm panaflo waiting to replace my apha 8942 w/ 80mm YStech fan but haven't installed yet.....I may also order one of those massive Zalmann heatsinks for the chipset..It is supposed to help a bit over the active cooling stock now...
 
I think it's the is7 reading the temps wrong.. Mine is doing the same, my temps at load:

with zalman cnps7000 + as5
CPU: 55c
pwm: 53c
Case: 36~37


Stock hsf, no thermal paste just the black thing.
CPU: 70c
PWM: 64c
Case: 37

Never crashed in eitehr circumstance though.. even when it was at 70c i was encoding divx movies for hours with no crash or reboot.



This is for the rig in my sig.




 
What bios version are you running???

I am runing IC7 and have ran bios v14,v17, & v18..no difference.....

Mine has to be temps and since it can't be the memory cause I have ran it with a 3:2 ratio and well below its rated speed and specs....Also can't be vid card cause it wouldn't error out the tmpgenc program....

I have never been happy with the abit temps though I have heard yes it reads 5-10c different then most others but may actually be the more accurate temps versus all the others...
 
IS7 BIOS 17
Version: 17 Released Date: 2003-11-04

I'm using that one above but my board came with 15 and i went thru 16 too, they all read the same.
Always really high, but never had stability probs due to temps.

Don't you have the ic7 duvie?
could be that our super high pwm high temp readings are specific to is7 boards.

My pwm stills exceeds my cpu temps sometimes..
Another thing i noticed, turning down the nb fan from 5k+ to 2k in speedfan doesn't effect the pwm temp.. so maybe it's not reporting the NB???
at idle:

Fan02: 5273 RPM PWM: 36c
Fan02: 1764 PM PWM: 35c
 
I will install speedfan and see if I can duplicate that by lower NB fan speed....


I do have the IC7.. Now way system can be the NB chipset cause I am at 19c right now and I am certain that is room temp with window cracked cause thermostat reads just about that...therefore it would be a case temp and no way can be a NB chipset cranked at 252fsb w/ 1.6v and 3 hours into a 4hours encoding...PWM is reading 38.5-39c right now....Seems liek a NB temp as my old i845pe chipset use to hit low 40's as well oc'd to 180fsb...
 
dropped fan speed from 5000+ to 2200+ and no difference as well in eithe the system or PWM temps....If I go 35 or belw the stupid speedfanand hardware monitor in sequence report 14000+ rpm...I call Bullshit on the whole crappy Abit thing....

I can't even hear the wind down when I dropped the NB fan done in speed and my alpha with ystech is not that loud that I could.'t hear it...
 
just did the fan test under FULL load..2x prime

same results.

fan02 @ 5273 RPM..
The second speed reading(temp3) which we are assuming to be pwm? and also assuming to be the NB temp is at 50c.

I drop fan02 to 40% = 2344 rpm
and the temp3 reading still = 50c

i'm not 100% sure what temp3 is measuring in speedfan but its the second one listed and usually second hottest under load or idle.
temp2 i'm pretty sure is cpu.. it's the at the top and always hottest.
temp1 is always lowest and on bottom of list which i'm assuming measures case temp.
 
beatle - In case you didn't believe any of the previous posters that said Abit boards show temps that are around 10°C hotter than other boards, here's a link for you: High temp readings on Abit mobos.

The thread linked above is the culmination of several very large and vocal threads about this issue in the Abit-usa forums. If you want even more background info, look up those threads too. There's also some big threads about this same issue over at OCforums from earlier this year.

Bottom line is: your temps aren't nearly as bad as they look.
 
Regarding the Abit IC7/IC7-G/IC7-Max3 and the temps. Yes the Abit method of "reading" the CPU diode temp shows an approximate 10C increase over other OEM methods. FWIW Abit claims that thier conversion follows the Winbond (monitoring chip) specs whereas others don't.

The way CPU temperature measured in the BIOS or in the hardware monitor is through an I/O chip on the motherboard. That I/O chip probes a voltage given by the CPU thermal diode, and then BIOS / hardware monitor reads this value and uses some mathematical way to calculate the approximate temperature
Since the temperature is ?calculated?, not ?measured?, the formula which the BIOS uses will make the outcome different. There is a parameter that is provided by the I/O chip vender, which we expect it to be the standard, so we use this parameter to calculate the temperature of the processor. Once this parameter in the BIOS is changed, the temperature shown in the BIOS PC health or in the hardware monitor tool will be altered. But, in the mean time, the actual temperature of the CPU is still the same. So the temperature you saw might be higher comparing to other motherboard with the same chipset, it is because we use the different parameter to calculate the temperature. We cannot say that we are more accurate, but we sure follow the standard provided by the I/O chip vender, and we believe it should be the proper one.
Click to expand...

Where are the "System" and "PWM" temps "taken"? Below are my findings:
the system temp reading is the area around the Winbond chip, near the PCI slots and thus should be the coolest temp as it's near the bottom of the mainboard. My MBM results seem to verify this. I also tested this by using a can of compressed air and have seen an immediate reduction in the sys temp reading as the cold air hit the chip/area.

You'll note that the Abit "knowledge base" in the definition What is PWM Temperature? says that the PWM temp can be higher than the CPU temp. Now, from what I've seen, the components are located above the CPU socket. You'll notice that the Max3 came with a plastic housing/fan above the socket, supposedly helping to keep the area cooler.
Once again, I have seen a decrease in the PWM temp by directing the cold air at this area.
I have never seen the "PWM" temp anywhere near the CPU temp under load but my PS has a 90mm lower fan just above this area which may be helping to keep it cooler than some.

With my IC7-G and a P4 oc to 3.1G/1.60 vcore set, 1.575 under load/ I see the following temps using MBM 5.3x and crosschecking in the BIOS:

Ambient room air at lower front fan inlet 18C
System temp 20C
CPU with Alpha 8942/39CFM fan 34C
PWM temp 24C

Under load the system temp may rise 1-2C
CPU (Load) 45-48C
PWM 30-32C


How accurate is all this? I don't know but this is what I've seen with my IC7-G.

Mike G
 
That's great, abit could be reading it properly and others may not.
but, since the other manufactures are considered the norm and abit the exception. they should at least put in a notice with the mb to warn us that temps may be reported 10c or so higher than the other motherboards (even if the other manufactures are not measuring properly to abit's standards) just as courtesy to their customers.

This will clear up the confusion and cause less worry for users. When they hear that p4 load temps should be 50c max etc.. then see that their cpu temps are much higher under load the they worry and end up re arranging heatsinks, buying extra fans, bending over backwards just to lower temps to what would of been read as 50 c on any other mb to begin with.
 
Your temps dont seem out of line.
Here is a picture of where "PWM" is measured on my shuttle. It is the little baby blue thermistor just west of the large cap and just east of the mosfet.
Do the ABIT boards have a thermistor reading temps near a mosfet? You can tell by watching the temps and putting your finger on the thermistor. The temps change very fast.
 
Well I feel better this morning after reading these posts. 🙂 I had just read too many posts of "I'm at 36 idle/42 under load @ 3.2!" 😱 I'm coming from an NF7-S where the temps were read high (but not this high!) If Abit is more accurate than other boards, so be it. I just wish there were a standard that manufacturers could agree on so stuff like this doesn't happen. I'm going to apply Ceramique to the NB and see if I can drop those temps lower. I'll also check my CPU to see if the spread of TIM looks like it does on AS's site.

UPDATE: Ok, I just reapplied Ceramique to the cpu hs. It looked pretty much like it did on AS's site, but slightly larger (my circle went to the edges) which means I put a bit too much on. I'll see if this has any impact. The NB uses the "chewing gum" TIM which is a major pain to remove. I did my best with 91% isopropyl, but I couldn't get it looking new.

Also, my TwinMos uses BH5 winbond modules. I bought them earlier this year when they were hyped to be as good as Corsair's better stuff.
rolleye.gif
 
Originally posted by: beatle
The NB uses the "chewing gum" TIM which is a major pain to remove. I did my best with 91% isopropyl, but I couldn't get it looking new.
Click to expand...

That stuff called "GooGone" will take off that bubble gum TIM in no short order, it works super good. Just follow up with a wash of isopropyl alchohol to remove all the GooGone residue.

 
Dang, and I think I actually have Goo Gone too... oh well, maybe when I upgrade the NB cooling. 🙂

I'm currently stable over 8 hours P95 @ 231x13 @ 1.575v (1.47-1.49v actual). This board undervolts like crazy. Temperatures are absolutely molten: 70-72C CPU / 67-70C NB! I tried 235x13 at this voltage and P95 errored out in under 5 minutes. To play it safe, I'll likely go with better cooling before pushing this chip harder. I'm hoping for 3.2! 😀

Duvie, do you or anyone else have any reviews handy of that Zalman passive heatsink showing its performance in comparison to the stock IS7 (does IC7 use the same?) NB cooler?
 
try 8 hours P95 @ 200x13 @ 1.575v and see how high the temp gets.. that will give you an idea of how high your is7 reports the cpu temps without any overclocking for comparison. I highly doubt your NB is really 70c after 8 hours of prime95 just due to some minor overclocking. Either pwr is not = NB or is7 isn't reporting it correctly.



 
It's entirely possible that PWM isn't the NB temp. I've heard PWM in terms of pulse width modulation (used in voltage regulation) so it could be another way the board is reading the CPU temp. I'll give it a shot tonight @ 200x13 1.575v. Are you sure you want 1.575 and not 1.525 (default)?
 
Well, if you keep the voltage the same then the results are more meaningful.

You will clearly see whether a change in clock speed results in a change in temps while everything else is held constant.

If it doesn't change much then that means that the super high temps are not a direct result of overclocking but rather just high temp reporting issue of the motherboard.

Is7 undervolts so a 1.575 setting in bios is usually 1.525(default) anyways.


Ie.. if 2.6c @ 2.6 load reports 65c, then you shouldn't worry about your 2.6c @ 3.0 load temps being 70c.



 
CPU temps were 65-66 @ 200x13, 1.575v. PWM was 61. Looks like I'm ok. Can someone clarify what this PWM reading actually means?
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top