It doesn't matter how many fans you have going. Ambient case temp has to deal with that.
No matter how many fans you have, there is no way for a case ambient temp to be under room air temp. And unless your room temp is between 45-50F, there is no way your case ambient temp is going to be low enough for a 38C core temp to be possible.
Mike
No matter how many fans you have, there is no way for a case ambient temp to be under room air temp. And unless your room temp is between 45-50F, there is no way your case ambient temp is going to be low enough for a 38C core temp to be possible.
Mike
My ambient case temp is 23 Celsius. Right now, my CPU temp is 42 C. I have the stock TaiSol HSF. I have 5 case fans, 3 blowing in, 2 out, plus the PS fan. My system has finally been running stable, since I disabled USB, and SB16 emulation. I'm guessing, something to do with the IRQ sharing. The temp still really bugs me though. When I get a chance, I'm going to go and buy some more thermal compound, and re-apply it to my CPU. Will post an update after I get that done.
Mem
Lifer
- Apr 23, 2000
- 21,476
- 13
- 81
OK guy`s first temperature in the FORTYS IS FINE(well within the safety range),mine is at 46-47c under full load, average temperature for my 900T-bird is 42c using the standard heatsink/fan(Taisol)with no special thermal paste,I`m also using 2 case fans,it seems too many Members are worring for nothing ,you will know if it`s running too hot normally by stability problems.
40s, and even low 50s for those overclocked into the Gigahertz and beyond range are more than fine for AMD Cpu's.
I'm tired of seeing people coming in and bragging about how their cpu is running in the 34-38C range, confusing all the other people running in the 40+C range. Especially when those reporting temps in the 30C range are reporting severely low misreadings.
Mike
I'm tired of seeing people coming in and bragging about how their cpu is running in the 34-38C range, confusing all the other people running in the 40+C range. Especially when those reporting temps in the 30C range are reporting severely low misreadings.
Mike
Mikewarrior, these people are getting temp ratings from motherboard monitors, of course the temps are off. When people are comparing temps, 99% of them are comparing the readings they get from the motherboards diodes, so who really cares what the REAL temp is? Noone is comparing that because noone knows the REAL temp. I think the man focus should be what bios they are running and mobo to clear up inconsistancies.
DO you not understand anything Mortanius? That is precisely why not to compare temps from mb thermistors. Not only are they highly inaccurate, and depend more on ambient air temp than does an internal diode reading, but lets look at why the readings cannot be comapred.
1) Certain bios revision's ADJUST THE TEMP UPWARD IN AN ATTEMPT TO MAKE Up for the thermistor inaccuracies. However, each mb manufacturer chooses a different compensation.
2) Can you guarantee that the thermistor is in the exact same spot on every single cpu/mb combination? NO. Because you have people bending the thermistors on kt7's, and you have numerous other mb's where the thermistor isn't moveable. Unless you have the exact same placement of thermistor on every single mb, you cannot compare anything.
3) The temps are off. That alone should stop people from comparing Socket A Cpu temps. When your readings are inaccurate, especially to the point of those who say their cpu's are running in the 32-38C range, how can you trust them? And even on the ones that are compensated for, you still can't trust the readings. Why? because the method to read cpu temps is inaccurate.
If your cpu runs fine, great. Why compare temps when you have people coming into each thread claiming their cpu runs at ridiculously low temps, only adding to the confusion of the orignal poster.
Mike
1) Certain bios revision's ADJUST THE TEMP UPWARD IN AN ATTEMPT TO MAKE Up for the thermistor inaccuracies. However, each mb manufacturer chooses a different compensation.
2) Can you guarantee that the thermistor is in the exact same spot on every single cpu/mb combination? NO. Because you have people bending the thermistors on kt7's, and you have numerous other mb's where the thermistor isn't moveable. Unless you have the exact same placement of thermistor on every single mb, you cannot compare anything.
3) The temps are off. That alone should stop people from comparing Socket A Cpu temps. When your readings are inaccurate, especially to the point of those who say their cpu's are running in the 32-38C range, how can you trust them? And even on the ones that are compensated for, you still can't trust the readings. Why? because the method to read cpu temps is inaccurate.
If your cpu runs fine, great. Why compare temps when you have people coming into each thread claiming their cpu runs at ridiculously low temps, only adding to the confusion of the orignal poster.
Mike
1) Certain bios revision's ADJUST THE TEMP UPWARD IN AN ATTEMPT TO MAKE Up for the thermistor inaccuracies. However, each mb manufacturer chooses a different compensation.
- I know this,, thats why you can compare different mobos
2) Can you guarantee that the thermistor is in the exact same spot on every single cpu/mb combination? NO. Because you have people bending the thermistors on kt7's, and you have numerous other mb's where the thermistor isn't moveable. Unless you have the exact same placement of thermistor on every single mb, you cannot compare anything.
-Nope you cant, but the temp does give us a rough idea where we are at, a guy sitting at 30 c idle is prolly better off then a guy sitting 60c idle even tho the temps are way off.
3) The temps are off. That alone should stop people from comparing Socket A Cpu temps. When your readings are inaccurate, especially to the point of those who say their cpu's are running in the 32-38C range, how can you trust them? And even on the ones that are compensated for, you still can't trust the readings. Why? because the method to read cpu temps is inaccurate.
-You cant trust them, but again they will give you a rough idea in any case if you are running too hot.
- I know this,, thats why you can compare different mobos
2) Can you guarantee that the thermistor is in the exact same spot on every single cpu/mb combination? NO. Because you have people bending the thermistors on kt7's, and you have numerous other mb's where the thermistor isn't moveable. Unless you have the exact same placement of thermistor on every single mb, you cannot compare anything.
-Nope you cant, but the temp does give us a rough idea where we are at, a guy sitting at 30 c idle is prolly better off then a guy sitting 60c idle even tho the temps are way off.
3) The temps are off. That alone should stop people from comparing Socket A Cpu temps. When your readings are inaccurate, especially to the point of those who say their cpu's are running in the 32-38C range, how can you trust them? And even on the ones that are compensated for, you still can't trust the readings. Why? because the method to read cpu temps is inaccurate.
-You cant trust them, but again they will give you a rough idea in any case if you are running too hot.
That is precisely why you should not and can not compare temps. Because people reporting correct, compensated temps in the right range, like gimps are not tricked into believing that their cpu is running too warm.
All these people claiming low temp duron/t-birds make those with updated bios' think their cpu's are running too hot. When in fact, they are running just fine.
Since you obviously don't care about accuracy or even remote accuracy, i will leave you be. You obviously don't understand that when accuracy doesn't exist, you can't compare. It doesn't give you a rough idea because the temps are wrong. Differences of up to 10-15C are certainly attributable to case temp and bios version.
Mike
All these people claiming low temp duron/t-birds make those with updated bios' think their cpu's are running too hot. When in fact, they are running just fine.
Since you obviously don't care about accuracy or even remote accuracy, i will leave you be. You obviously don't understand that when accuracy doesn't exist, you can't compare. It doesn't give you a rough idea because the temps are wrong. Differences of up to 10-15C are certainly attributable to case temp and bios version.
Mike
If my mobo is reporting a temp of 60c+ , im gonna have a rough idea that something is wrong even if the temp gauge is not 100% accurate. Ive put together probably 100 kt7-raids and the temps are actually pretty darn close and i dont pay attention to bend the thermostat the same way every time.
THERE IN YOUR POST LIES THE EXACT PROBLEM WITH RELYING ON A MB THERMISTOR!!!
All the temps are fairly the same? Do you honestly expect every single cpu to run at the same temp? The reason they run at the same temp is more due to the fact you probably have a sustained room temp than it does with the chips running at the same temp.
And the fact that you continuously skirt the issue of people posting ridiculously low temps is really appalling. Are you suggesting that Gimp should worry because his temps are too high?
Again, the main problem lies in people believing the thermistor temp as core temp, and people posting erroneous temps that are unattainable. is that too hard for you to understand?
Might I also add that you were guilty of this, too. You claimed a 37C CPU CORE TEMP. YOu did not state thermistor. You claimed that as your cpu core temp at full load. And i am certain that seeing temps like that and FreakOCR's are what get people like GIMP concerned about their cpu temps. When the truth is, they should not worry about it.
Mike
All the temps are fairly the same? Do you honestly expect every single cpu to run at the same temp? The reason they run at the same temp is more due to the fact you probably have a sustained room temp than it does with the chips running at the same temp.
And the fact that you continuously skirt the issue of people posting ridiculously low temps is really appalling. Are you suggesting that Gimp should worry because his temps are too high?
Again, the main problem lies in people believing the thermistor temp as core temp, and people posting erroneous temps that are unattainable. is that too hard for you to understand?
Might I also add that you were guilty of this, too. You claimed a 37C CPU CORE TEMP. YOu did not state thermistor. You claimed that as your cpu core temp at full load. And i am certain that seeing temps like that and FreakOCR's are what get people like GIMP concerned about their cpu temps. When the truth is, they should not worry about it.
Mike
Heh. You probably made elite just trying to prove your point in this thread!!! LOL!!!!
Okay i understand now, but that doesn't explain the unstability!!!
Okay i understand now, but that doesn't explain the unstability!!!
I'm not worried about the high temp, because of other people's postings. I'm worried about it, because if a CPU is running hot, it can still run stable. Plus, I can't even get my TBird 800 to run stable at 850 (100x8.5). This makes me think that maybe it's getting too hot. My VCore is set to 1.75V, and I cannot hit 850. Plus, if you've read my postings, the mobo thermistor temp reading jumps over 20 degrees C when under load. At idle, it's a min of 39 C. Under load, it jumps up to over 50, and sometimes even the low 60s Celsius. Which really would mean something, wouldn't it?
this is also pretty weird. my tbird 800 is about 57c !!!! when full load and idles around 45-50c....this is pretty damn hot......and it's only running at a core voltage of 1.75v...i think this calls for a new heatsink/fan.
Ok, not even under a heavy load, my mobo thermistor is giving me a reading of 67 Celsius. Would the thermistor be giving a higher or lower reading compared to the true Die temp?
I'm running at 800@1Ghz, 32C at idle, 49C at full load. That's according to the VIA hardware monitor 2.04. Do not know how accurate that is though.
cheers
cheers
Gimps, it all depends on your ambient temp, proper heatsink application(ie removing teh blue strip from the PCTC), and which bios you are running. Proper heatsink application is absolutely critical for socket A heatsinks(due to high heat and small contact surface).
Since Most kt7's exhibit a 5 degree difference(or around there) between idle and full load, i would suggest that you remove the heatsink, and strip off the PCTC. If you have a quality aluminum or better based thermal paste, use that instead.
Also, maybe you aren't discouraged by teh low temp postings, but other people certainly have been in the past.
Mike
Since Most kt7's exhibit a 5 degree difference(or around there) between idle and full load, i would suggest that you remove the heatsink, and strip off the PCTC. If you have a quality aluminum or better based thermal paste, use that instead.
Also, maybe you aren't discouraged by teh low temp postings, but other people certainly have been in the past.
Mike
Hmm...I didn't get the Via Hardware Monitor with my Tyan S2390..I wonder if it would work on my board?
Running CPUIdle 5.5 results in an idle-temp of 43c.....Running a FlaskMPEG job makes it jump to 45-47c (using latest Motherboard Monitor).
T-Bird 800@800
Tyan S2390
128MB PC-133 ram
Creative Live! Value
Diamond V550
Maxtor ATA66 20gb 7200rpm
Maxtor ATA66 60gb 5400rpm
Creative Ovation 12x DVD
Smart&Friendly Turbowriter CD-RW
Running CPUIdle 5.5 results in an idle-temp of 43c.....Running a FlaskMPEG job makes it jump to 45-47c (using latest Motherboard Monitor).
T-Bird 800@800
Tyan S2390
128MB PC-133 ram
Creative Live! Value
Diamond V550
Maxtor ATA66 20gb 7200rpm
Maxtor ATA66 60gb 5400rpm
Creative Ovation 12x DVD
Smart&Friendly Turbowriter CD-RW
MortaniuS or Mikewarrior2: Question...maybe you just answered it for me, but.... At first, I almost broke my 1Gig CPU on my KT7-RAID, while trying to install it. It kept rocking back and forth but would not seat properly. Then I notice that the thermistor(I think that's what it is) was sitting about a quarter of an inch above the top surface of the socket. I considered resoldering, and bending, and opted for bending. In your experience, is this the normal on the KT7-RAID board? If so, can you figure out as to why they would leave the thermistor sticking up like that? The initial rocking left a discoloration/small dent on the bottom of the CPU at point of impact. Will this possibly shorten the life expectancy of the chip(area that will overheat because of slight damage)?
Bending the thermistor should pose no problems for you. The dent it made on the cpu should also likewise not effect the cpu in any adverse ways. Keep in mind that the backside of a fc produced cpu(duron/t-bird/p3) is idle substrate, with a few resistor's thrown on there.
Also, it is common for kt7's to have the thermist somewhat sticking out of the socket.
Mike
Also, it is common for kt7's to have the thermist somewhat sticking out of the socket.
Mike
Mikewarrior2: Thanx for the reply to my question!!!
Now back to Gimps problem...anyone have any ideas why he can't get his CPU to clock 50Mhz higher than stock, and why is there no stability at normal clock speed? I'm concerned that the heatsink does not feel hot. If it is pulling heat away from the chip, it should feel hot. I read somewhere that an AMD Tech stated that a 1Ghz T-bird(yes, I know he has an 800)would self destruct within 8 seconds if it were not sinked properly. To me that means they run HOT, and therefore the heatsink if it is effective, should be hot as well. I'm not familiar with his particular heatsink, but with mine(Alpha PAL35T), I had to remove some plastic feet from the mating surface to insure proper contact. Also with my heatsink there were two different locking clip configurations...one for Intel, and a much tighter clip for AMD. I'm wondering if there is good contact between sink and chip.
Mikewarrior2...I agree with you about these arbitrary temp readings, but I still wonder if he might possibly have a heat related problem. Disreguarding how accurate his temperature readings are, if the temp. is jumping 20 degrees C, that would give me pause for concern!
Gimps:...Can you put your chip in someone else's system to ascertain if the problem IS your chip? Also, just for some peace of mind, you might want to try some Arctic Silver thermal compound. I've read that it can reduce temperatures substantially in comparison with most of the other products on the market...
Now back to Gimps problem...anyone have any ideas why he can't get his CPU to clock 50Mhz higher than stock, and why is there no stability at normal clock speed? I'm concerned that the heatsink does not feel hot. If it is pulling heat away from the chip, it should feel hot. I read somewhere that an AMD Tech stated that a 1Ghz T-bird(yes, I know he has an 800)would self destruct within 8 seconds if it were not sinked properly. To me that means they run HOT, and therefore the heatsink if it is effective, should be hot as well. I'm not familiar with his particular heatsink, but with mine(Alpha PAL35T), I had to remove some plastic feet from the mating surface to insure proper contact. Also with my heatsink there were two different locking clip configurations...one for Intel, and a much tighter clip for AMD. I'm wondering if there is good contact between sink and chip.
Mikewarrior2...I agree with you about these arbitrary temp readings, but I still wonder if he might possibly have a heat related problem. Disreguarding how accurate his temperature readings are, if the temp. is jumping 20 degrees C, that would give me pause for concern!
Gimps:...Can you put your chip in someone else's system to ascertain if the problem IS your chip? Also, just for some peace of mind, you might want to try some Arctic Silver thermal compound. I've read that it can reduce temperatures substantially in comparison with most of the other products on the market...
See my Sig below for what I have.
My temps are 35c Idle and around 41c under load of RC5. Mother board Monitor, Via Hardware Monitor, Hardware monitor Pro, ect. all say the same and they are about 12c LOWER than what my homemade external probe with LCD are reporting. It is placed next to the core between the pcb and heat plate (Slot-A T-bird with case off)
Bottom line: I dont trust the internal sensors and software.
My temps are 35c Idle and around 41c under load of RC5. Mother board Monitor, Via Hardware Monitor, Hardware monitor Pro, ect. all say the same and they are about 12c LOWER than what my homemade external probe with LCD are reporting. It is placed next to the core between the pcb and heat plate (Slot-A T-bird with case off)
Bottom line: I dont trust the internal sensors and software.
andygold,
Yes, it's normal for KT7s to have the thermistor sticking out, in fact, all KT7s ship with the thing sticking out. Bending it in a little is the right thing to do... but you don't wanna bend it too far down. closer is better.
Yes, it's normal for KT7s to have the thermistor sticking out, in fact, all KT7s ship with the thing sticking out. Bending it in a little is the right thing to do... but you don't wanna bend it too far down. closer is better.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes Discussion Threads
- Started by Tigerick
- Replies: 21K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter