False gpu temperature readings and what to do?

tomahawkhead

Honorable
May 3, 2013
7
0
10,510
HIS Radeon hd4830 shows 80c on startup, idles on 96. Clearly wrong readings. Makes the fan run 80% half the time.
I already checked and cleaned the cooler, repasted and reseated the heatsink and fan, everything seems completely in order. Checked the temps with every possible program, same results.
What is happening?
 

tomahawkhead

Honorable
May 3, 2013
7
0
10,510

Well, for one thing, my hands. On startup, those 80c would be already burning hot, but the heatsink and the card are mildly warm to the touch, i would say body temperature or so.
At one point, i forgot to mention, i had normal readings: 36 on start, idling around 47. Just one time.That was when i first cleaned it.
On next restart after couple of hours - back to 80.
And not to mention that it doesn't make sense that any card should be at 80 degrees right after boot up.
I am sure that the drivers and CCC are up to date.
Something is wrong though.
I am using Windows 8, by the way.
Cheers


 
Normal idle temp of the Radeon HD 4830 is around 37°C.

Normal load temp of the Radeon HD 4830 is 61°C.

I would try to make the heasink pull tighter against the GPU by adding a thin washer to the back of the graphics card at each heatsink mounting point and place the rear mounting bracket on top of the washers.
 

tomahawkhead

Honorable
May 3, 2013
7
0
10,510

Yes I'm aware that the heatsink is cooler, the back of the card where the chip is also not near hot enough to reach 80 i think.
I used a fair amount of paste, and fastened the heatsink as tight as it would go.


 

tomahawkhead

Honorable
May 3, 2013
7
0
10,510


That's a good idea, i thought of something similar myself, but i figured that it would be obvious if it wasn't in contact with the processor, maybe it would not be heating up itself or cooling at all.
I will give that a try, thanks.



 

tomahawkhead

Honorable
May 3, 2013
7
0
10,510
Sorry to say that the trick with the washers didn't work.
The heatsink is tight to the board, repasted, still the same thing going on.
I am back to believing that the temp readings are wrong for some reason, or something else must be.
Anyone heard of this? I am open to suggestions of any kind, this is really bugging me, I believe this is a very nice graphic card.
Cheers
 

tomahawkhead

Honorable
May 3, 2013
7
0
10,510


Yes, I did. GPU-Z, MSI afterburner, Speedfan, AMD Vision Engine Control Center, all report the same high temperatures.
At least i made a custom fan curvature in MSI afterburner, so that the fan doesn't drive me insane spinning at 70% speed all the time, making noise. It is now spinning at 40% speed, keeping the reported temperature idling around 87c.
I can use the card playing games at highest resolution, highest settings with no artifacts or problems, which i think I should experience when the temperature reaches 100 and more, which it reportedly does.


 

tomahawkhead

Honorable
May 3, 2013
7
0
10,510
Hey guys, sorry if I'm boring anyone with this thread, but I find what's happening here very interesting.
My card , and/or temp. readings have gone back to "normal".
This morning i turned the computer on, and saw the temperature of the card was 41c.
It is now idling quietly around 47.
Couple of restarts later, it's still good. (knock on wood)
Not sure it will last, but anyway.
I did not touch the card anymore for the last couple of days, didn't mess with the cooler, or did anything to the system or drivers.
I did, however, get rid of a virus, performed a full scan with Windows defender, and it found a Trojan dropper of some kind somewhere, don't even remember where it was.
Anyway, I don't really believe that was the reason for things to come around, but I'm hoping this keeps up.
Cheers