Sapphire Radeon HD 5570 overheating problem

rimE201

Reputable
Jan 25, 2016
6
0
4,510
Well I'm not sure what's happening with my graphics card.
Now, before you all tell me to change my thermo paste and clean out my fans I want you to know that I did that 5 months ago. So here's the story:
Last summer (5 months ago in august) I was playing League when all of a sudden I hear a sound in my headphones so I minimize the game to investigate what might be causing this sound. After a short search in the volume mixer I find that a program I use for monitoring my GPU/CPU was on alert due to my GPU reaching >80°C. I turn it off and ignore it for a while. Tomorrow I wake up, play another round and see that my GPU is at 91°C. Now that's where I kinda got scared.
I then changed my thermo paste and cleaned out my fan (it did not have a lot of dust on it) and detached the radiator from the fan and cleaned that too (the radiator had an unspeakable amount of dust in it). So after I cleaned it all up I put it back in and played another round of League. Then I checked my temps and they were at about 85°C. I updated my drivers, and thought that would fix it but it didn't. So then I just left it there to kinda burn because I literally had no idea what might be causing the problem. I considered calling a priest to exorcise any possible demons that might be hiding in my GPU, I'm that pissed... So I just leave the card to burn because I was frustrated that I couldn't figure out what the problem was.
15 days after I check my GPU temp while playing League and it's all like 65°C -75°C I was surprised and happy so I just "rolled" with it and totally forgot about the whole 10 days that my card was overheating.
Fast forward to present day (1/25/2016):
My CPU suddenly starts to get temps like 70-85°C and I open my rig up to investigate. I Clean the radiator (had a lot of dust in it), I clean the fans and I replace the stock thermo paste that was on it. I boot up my rig and what do I hear? THE DEEPEST AND MOST HORRIFYING BEEP THAT SOUNDED LIKE IT CAME FROM THE UNDERWORLD. Right then and there I knew I F-ed up.
My CPU temps were now about 40°C idle and I've never seen them go over 51°C even at 100% load. But oh god oh god if you could see my petrified face, I looked like someone just stabbed me in the back when I saw THAT MY GPU TEMP IS NOW AGAIN 80° IDLE!! I go and start up League to play a game and I see that my GPU temp is now 92°C.
Mind you, 92°C at like 50% load because I play League on the lowest settings possible. If you're wondering why I'm not using any testbench / GPU stress tester is because I'm sacred that if I overload it that it's gonna burn the whole rig down including my motherboard, my house, me and most importantly my kittens.

But I think this might give you guys a clue
The overheating problems started in august and I did something with my graphics card back in July. I'm not really sure what it was but it was some kind of program for rendering / parallel processing I installed and read it said somewhere that it might go faster if I edit the
Shader Buffer setting. So I put the shader buffer on 128 or 256 (I'm not sure) and yes it did go faster. But my GPU started overheating only a month after that.

I've also asked many other people online what to do and some told me that I should just give up on everything and just start cooking my breakfast on the card and some told me that those temperatures are normal for that card (I'm pretty sure they're not normal for any card), some others told me that I should update my drivers. I did that, nothing happened.

I'll link you to another guy that seems to have the same problem: http://www.tomshardware.co.uk/answers/id-1771936/ati-radeon-5570-overheating.html

So TL;DR
-August : I start having overheating problems, clean everything update drivers, nothing changes.
-From September to January: GPU temp seems fine
-January 2016: I take out my CPU and put it back in, boot rig up, GPU overheating again.

Please someone help me with this I'm out of ideas. I literally have no clue as to what could be causing this card to do this. Temp sensor seems fine because when I boot up the rig it shows temps in the 30°C and quickly climbs up to 80°C. And please don't tell me that this card is just old and that it's reaching it's end because I've been for about 5 years with this card and it's very dear to me...
If you need any screenshots just say it and I'll reply with one.

UPDATE: Yes I have 1 fan on the top side that seems to be blowing out hot air and my GPU is positioned that it's fan is taking air in from the bottom of my case. I assume that's good because hot air rises and cold air drops? I don't think airflow is causing any problems as my CPU is now perfectly fine with temps and my motherboard temps are okay too.
 
Solution
That would probably work as the power ist almost the same in all GPU fanheaders. I don't think there are any problems with a short test, but i would not recommend to leave the fans in the wrong cards, but i can't guarantee it.

rimE201

Reputable
Jan 25, 2016
6
0
4,510


I know it's a little long and confusing but I had to make it that long because I'm just out of ideas of what might be happening to my GPU

It's my GPU.
My CPU was overheating too, but I solved that by redoing the thermo paste. The stock one was crispy and dry. I changed it and then my CPU stopped overheating but at that same bootup after I redid my thermopaste on my CPU, my GPU started getting temperatures like 80°C at idle and 90°C+ at load. I tried to underclock it from 650 to 500 clockspeed so that it may live longer and I don't really need a hight clockspeed since all I do is play League on the lowest settings possible. The underclock brought the temps at high load from 93°C (stable) to 89°C (stable).
 

thehutti

Distinguished
May 12, 2014
351
0
19,160
I think the problem is your so called shader-buffer-upgrade thing. No idea what this prog did, but if it did something like a GPU-BIOS flash to OC the card, then you could try to flash the stock BIOS. Another probability would be that you put too much termal paste on the GPU while reappling it, or that you mountet the GPU cooler unevenly. Could also be a broken card of course.
 

rimE201

Reputable
Jan 25, 2016
6
0
4,510
Well I found out what the problem is... My GPU fan isn't working. Is it broken? How do I check?
http://postimg.org/gallery/x6hoirac/
Also if I need a new one it shouldn't be a problem should it? I mean fans are cheap but how do I know which one I get?
I have a 400 VA power supply (VA I'm assuming voltampere = watt) But how would I know if I have the right power cables for the right cooler? I have a few unused cables going out of a hole in my power supply and they're tied up and unused? How do I know that I have the right one that I need in there
Does any cooler work with Radeon 5570? I don't think that the big ones are gonna fit?
 

thehutti

Distinguished
May 12, 2014
351
0
19,160
I don't think you have the right connector on the PSU to test the fan directly. Also you will need a fan that is build for your heatsink. Not all fans will fit. Most will probably fit by size, but the mountingparts are not on the right place to add it. So it's best if you just look for a complete heatsink replacement for your model or a specific fan that you can add to yours. You can test the fan by applying DC current to it, but it would be easier to just plug off the fan and see if there is current on the pins of the GPU with a multimeter.
 

rimE201

Reputable
Jan 25, 2016
6
0
4,510


And that's how I'd use my multimeter, IF I HAD ONE :p
I have 2 cards in 2 different computers. Cards are similar, other specs are identical.
Is it possible for me to plug this cooler into the other card, and the other cards cooler into this overheating one? That way if the new cooler doesn't spin it's because there is no power in the pins and if my "broken" cooler spins on the other card I'd know that it's due to the card and not the cooler? Will it do something to my other card if I do that?
 

thehutti

Distinguished
May 12, 2014
351
0
19,160
That would probably work as the power ist almost the same in all GPU fanheaders. I don't think there are any problems with a short test, but i would not recommend to leave the fans in the wrong cards, but i can't guarantee it.
 
Solution

rimE201

Reputable
Jan 25, 2016
6
0
4,510
Thanks I just switched the cards with their fans and I found out that the fault lies within the fan, not my graphics card. My graphics card was providing power to the fan it just didn't work. So my sister got this wild idea to jumpstart the fan by rotating it with her finger and... it worked. I'm absolutely amazed. The fan is now spinning and is good to go. I thought it was a connection problem but I now see that all the wires are soldered to the metal pads correctly and the fan servos were probably just jammed by dust. Thank you for your answers, if you didn't tell me that the fault may lie in my card I would have never tried to switch my sister's cooler for mine and instead I'd waste money on a new fan.
Any way I can help you? Maby upvote your account or something? Just let me know :D