Temp01 Overheating - Rising to 100C+

[RockingSock]

Recently, I've been having problems with my computer. First, I thought it was the power supply "over heating" until I've bought a new one. After I've put in the power supply, I booted on the computer and it did the same thing on me.. shut off. After sending the computer in for diagnostics, they said the computer is working fine. After doing some research, I took things into my own hands. I downloaded SpeedFan to see the temperature of my CPU, Core etc. I've found out that Temp01 rises to 100C+ in a matter of 10 minutes. After disabling "Smart Fan" in BIOS (which seems to make the fan run at 3900rpm and 2200rpm) it seems to stay constant at 63C but shuts off in around 30-60 minutes or so. What could be the problem? What does Temp01 mean in SpeedFan?

[quest_for_silence]

What does Temp01 mean in SpeedFan?

Nothing exotic: it's just the very first temperature sensor which SpeedFan reads.
Use Open Hardware Monitor (or similar programs) to see if it calls that sensor in a more meaningful way.

[andyb]

A very quick and easy thing to do to check whether the sensor is calibrated correctly is to simply turn the computer on from cold (really cold, not been on for an hour), then load up speedfan straight away and make a note of the temp, also make a not of the CPU(s) temps as well.

Then watch the temperature rise making a note of the temperatures every 2-minutes.

Do this with the machine idle and get back to us.

If the start temp for any sensors are way off they can be ignored (i.e. start at -100C, or start at +255C), however you might have a sensor that is not calibrated, so it might start 40C higher than everything else, which would put it at 100C when other temps are at 60C, it might have nothing at all to do with the fault.


Andy

[RockingSock]

I think its the CPU overheating. I'm going to buy some thermal compound and see how it will work. Hopefully that is the problem.

[RockingSock]

After applying the thermal compound.. same thing happened.



Still don't know what the problem is.

[b_rubenstein]

Uh, the CPU core temps are no higher than 33ºC. That high temp is not for the CPU. The CPU is NOT overheating. Try putting your finger on the NB, and/or see if the HS on the NB is loose.

[RockingSock]

Uh, the CPU core temps are no higher than 33ºC. That high temp is not for the CPU. The CPU is NOT overheating. Try putting your finger on the NB, and/or see if the HS on the NB is loose.

What do you mean by NB and HS?

I just applied new thermal compound and screwed back in the heat sink tightly.

[MikeC]

NB = North Bridge chip
HS = heatsink

[RockingSock]

Thanks for the translation Mike.

Nothing seems to be "loose". It's all sturdy.

[andymcca]

The Temp01 etc on speedfan are often sensors on your motherboard. You could try feeling the NB and SB heatsinks after a crash to see if one is so hot you can't hold your finger to it. Of course, this is not a great indicator if the chip is not properly coupled to the heatsink (loose, voids).

You could always try re-applying the NB and SB heatsinks, but this might require taking the motherboard out. Wipe off all the grease and reapply with your new thermal paste.

[tim851]

Your readings are curious. Temp3 and *the other* Temp1 read at ~16c. That's really, really cool. Do you live on the North Pole by any chance? Santa, is that you?

If the mysterious Temp1 endured 100c before, it's very unlikely that it is now overheating at 63c. With the PC working fine while Temp1 is at a constant 63c and then just shutting down after 30-60 minutes, my guess is Temp1 is NOT your culprit.

One way to find out what Temp1 is, would be to get a relatively small fan, 80mm or so, hold it in your hand and aim it at various parts on the mainboard, wait for a while and observe if Temp1 ever goes down.
But there's a good chance that it's a "bogus" reading that you won't be able pin-point. I've got two of these.

[andymcca]

Your readings are curious. Temp3 and *the other* Temp1 read at ~16c. That's really, really cool. Do you live on the North Pole by any chance? Santa, is that you?

IIRC, if a sensor is supported by the chipset but not implemented (ie not routed on the motherboard), you see fixed default values. I'm not sure if typical motherboard temperature readings are taken with thermistors, but I imagine the fixed value is just the result of an "open" connection (ie no thermistor present).
That he can watch the other value increase, and that it seems to correspond with system lockups may point to a real problem somewhere though! Obviously correlation != causation, and it may be that there is some other issue which is related to the system warming up.

BTW: RockingSock, have you tried playing with your RAM? Try pulling out all but one module or RAM and seeing if the system locks up with just one module. Then try another module IN ANOTHER SLOT. Really you should try to use a different channel if you can figure out which slot that would be. The problem could be a module, a slot, a channel on the controller, or just the way the module is sitting in the slot. Obviously turn the machine off before removing/adding any RAM. Also: do the new-fangled 3-channel intel controllers work with a single channel populated? You may have to work in sets of 3 if you have a multiple of three slots on your board.

Sorry for the late reply. Have you had any luck?
Edit: I mention the RAM because it is the source of 80% of the intermittent lock-ups I've had to troubleshoot. A system can work fine for years and then magically have a RAM dimm go bad, or need to be re-seated in its slot. It's sort-of the first thing to check when having computer problems, since it is easy to test/remedy. If you really want to be sure, you could run MEMTEST86+ (involves burning a CD and booting off it)

[RockingSock]

I tried feeling out the hottest part.. only thing is the graphic card.

I will try to play around with the RAM. I want this piece of junk to work .

I also feel heat coming out of the power supply sometimes when the Smart fan option is on (Around 1200RPM fan speeds). I sometimes use a household fan. The fan seems to help a lot. Drops it to around 60C.

[tim851]

Yeah, the 4870 is one hot card. That is the heat you feel coming from the PSU, it's venting hot case air.
Since you have a Core i7, your mainboard must be of the X58 variety. The Northbridge should be pretty hot too, no?

Just out of curiosity:
Temp1 is within 1c of the GPU temperature. Do they correlate? SpeedFan shouldn't be able to read the Radeon's temp, but who knows.

[RockingSock]

All the fans seem to work fine. I tryed feeling around the hot spots. Hottest spot I felt was the NB and the graphic card (well around there).

[danimal]

speedfan doesn't tell you where those temps actually are, so it won't help... don't confuse the issue with it.

the problem here is that you are assuming that something is overheating, when there is no evidence to back that up... everyone is off on a tangent, that isn't helping to fix the problem.

if you think that the ram is bad, burn memtest onto a cd or dvd, boot off of it, and do some ram testing... memtest isn't the best, but it's a good start.

you need to provide more details on when the pc actually shuts down; and how many memory sticks are in it, for example... can you swap video cards? etc.

[Redzo]

Speedfan is best program there is for controlling fans, for temperature readings it is THE worst program currently available for download full stop.
If you are serius about controlling temps try HW32 or Everest or well ANYTHING else then speedfan and it is a mystery that nobody here on forum did not tell you this already.

[faugusztin]

And to the OP - because it is a Gateway board and there is next to none information about it, the best way to deal with this issue would be contacting Gateway and asking them directly. The OEM is probably Foxconn, but that won't help you too much.
In case of Asus, the motherboard temperature was actualy temperature near the memory voltage regulators, maybe check out that area if it is not overheating. There is also a question if it does the same with open case or not. If not, then it's probably a issue with airflow in your computer case.

@Redzo: please read up the whole thread and notice that Speedfan is not the only tool reporting that temperature and surprise, surprise, it is reporting the same temperature as others. Then there is your fact that "it is the worst problem for temperature readings". That is simply not true. There are sometimes issues with certain chips or boards, but for majority of things, the readings are correct (for example with my EVGA board, Speedfan is correct with all temperatures, compared to any other tool, including Coretemp, Everest etc etc).

[quest_for_silence]

@Redzo: please read up the whole thread and notice that Speedfan is not the only tool reporting that temperature and surprise, surprise, it is reporting the same temperature as others. Then there is your fact that "it is the worst problem for temperature readings". That is simply not true. There are sometimes issues with certain chips or boards, but for majority of things, the readings are correct (for example with my EVGA board, Speedfan is correct with all temperatures, compared to any other tool, including Coretemp, Everest etc etc).

+1.

[ces]

When you know you are running high temps, just grope around the board with your hands. You should be able to sense where the heat is coming from. Crude, simple and effective.

[RockingSock]

I think I've found out the problem. It has something to do with the heat sink or the microprocessor under it. Either the heat sink isn't giving proper air flow or the micro processor has something burned out in it, causing heat (or something like that).

Reasons?

The fan I use to keep the computer running just cools up the sensors. BIOS knows that the inside of the computer isn't over heating since the fan's cold air is hitting the sensors and around that. But actually, the chip (microprocessor) is over heating, causing the computer to heat up and shut down (without the fan). HW Monitor is reporting the the microprocessor is running at high temperatures causing shut downs and BIOS reports the inside of the system (case temp) is normal (with the fan on). If a microprocessor heats up and other parts around it, it creates more and more heat. Since my fan is "tricking" BIOS that the inside of the case is nice and cool, BIOS is thinking everything is alright and nothing is wrong.


I've also found out using my fan doesn't cool the microprocessor. It still runs a 85C-90C but doesn't bounce higher (most of the time).

Ehh... Tried my hardest to explain. Those might be the reasons. I still don't know what to do. I don't want to waste money for nothing :/ .

[darkb]

It's normal for the standard cooler with an i7 to get that hot under load, your CPU cooler is working fine. The high temp might be the voltage regulation circuitry for the CPU, that normally runs hot like that and can also cause an emergency shutdown. Check the heatsinks or chips near the CPU socket (between the socket and the ports on the back of the board, and between the socket and the top edge of the board)

[ces]

I think I've found out the problem. It has something to do with the heat sink or the microprocessor under it. Either the heat sink isn't giving proper air flow or the micro processor has something burned out in it, causing heat (or something like that).

That can't be. Your cores are running fine.

100C means you can't touch it. That is the temp of boiling water. An overheating video card should not be capable of shutting down your computer.

So that leaves only your northbridge. Is it cool enough so that you can place your finger on it and keep it on it? If so:
(a) it is not 100C
(b) if your computer is shutting down because of high temps, it is not high temps, it is some kind of sensor problem.

If anyone sees that I am overlooking something, please let me know. There some very smart people contributing to this thread. This seems so obvious to me... yet not to anyone else... it makes me not so confident that I am right. But I can't see it any other way.

[faugusztin]

Unfortunately no, the temperatures detected by motherboards are usually not the northbridge, but random sensors somewhere on motherboard. As i wrote, for example in case of Asus you didn't have a nortbridge sensor in P35/P45 boards at all - you had the motherboard temp sensor right next to memory VRM. This is Gateway/Foxconn board, so it could be anything, could be northbridge, southbridge, CPU VRM, anything. That is why i said he should contact Gateway itself with a support question.

[tim851]

I think I've found out the problem. It has something to do with the heat sink or the microprocessor under it. Either the heat sink isn't giving proper air flow or the micro processor has something burned out in it, causing heat (or something like that).

Which heat sink? Which micro processor?

The fan I use to keep the computer running just cools up the sensors. BIOS knows that the inside of the computer isn't over heating since the fan's cold air is hitting the sensors and around that.

Didn't you say that you turned OFF Smart Fan? That means all fans are running at full speed. And you still get shutdowns. It's not a BIOS fault then.

But actually, the chip (microprocessor) is over heating, causing the computer to heat up and shut down (without the fan). HW Monitor is reporting the the microprocessor is running at high temperatures causing shut downs and BIOS reports the inside of the system (case temp) is normal (with the fan on).

I've also found out using my fan doesn't cool the microprocessor. It still runs a 85C-90C but doesn't bounce higher (most of the time).

Again, which microprocessor? Do you mean the CPU? From what you've posted here, the CPU temps are fine!

[ces]

Unfortunately no, the temperatures detected by motherboards are usually not the northbridge, but random sensors somewhere on motherboard. As i wrote, for example in case of Asus you didn't have a nortbridge sensor in P35/P45 boards at all - you had the motherboard temp sensor right next to memory VRM. This is Gateway/Foxconn board, so it could be anything, could be northbridge, southbridge, CPU VRM, anything. That is why i said he should contact Gateway itself with a support question.

I think he said that the only two areas on his board (other than the CPU, which is not overheating) where he could feel heat were the video card and the northbridge. The video card isn't causing the shut down. And if the northbridge is just hot to the touch, but not in the range of the 100C that he is reading... it isn't the problem either. Correct?

If that is so, it appears that false sensor readings are the only culprit left. Correct?

[RockingSock]

I'm not that good at explaining and I'm not that good at computers, considering I'm only 15.

I'm saying that under the heat sink, the cpu is overheating or something around there.

[andyb]

It is totally possible that your CPU (an example, could be any chip with a heatsink) could be running at 100c and the heatsink might not be very hot. The reason is very simple, if the heatsink is not making a good contact with the chip, then it wont be getting very hot.

This problem can be caused by a number of things such as, the heatsink is not mounted correctly, there is no TIM (thermal transfer material, silicon paste, CPU gunk, call it whatever you like), or there may be something that is physically stopping the heatsink from making proper contact with the chip.

I have seen the following over the years. An AMD K6-2 heatsink mounted back to front, thus only making contact with the extreme edge of the CPU, a P3 with a sticky label on it, and NO CPU gunk, a S775 heatsink only locked down with 1 (one) pushpin, and a CPU that had a (hard) foreign object on it that stopped proper contact with the heatsink, and many with either no CPU gunk, or they used shit stuff that did not work, as well as several that use push-pins where not all were "pushed" down and secured correctly. And finally I have seen one case where the CPU Heatsink actually came off entirely, the boot of a car is not a sensible place to put a computer when travelling - put them on a seat, and strap it in, if you have an enormous and heavy heatsink, lay it flat.


Andy

[RockingSock]

I've applied thermal compound to it already. Anyways, I'm gonna take a picture and show you the inside. It will help me explain since I'm not good at explaining.

[RockingSock]

Here is the inside of my computer;



My heat sink fan. When it runs at 4000rpm, the thing HWmonitor and SpeedFan says is over heating, stays at 63C and shut over time. When the fan runs at 1200rpm, it rise up to 85C+ before shutting down. And there is a fan on the side for blowing air out. That's nothing important. Tell me, what do you think?