Sign in with
Sign up | Sign in
Your question
Closed

EAH5770 cucore temps raising to 100c

Tags:
  • Radeon
  • Graphics
Last response: in Graphics & Displays
Share
November 6, 2010 9:00:53 PM

Hello everyone, first of all sorry if this isn't the correct section to post this but I believe it is.

Second... as you could guess by the title I'm having temperature problems with this card, my system specs are as follow:

Graphic Card: Asus EAH5770 cucore
Motherboard: GA-790X-DS4
CPU: Phenom 2 x2 555 BE (2,6Ghz)
RAM: 4Gb Kingston 800Mhz

Now I've tried almost everything, replacing the thermal compound by a new one (a cheap one) with no result at all, undervolting it to 1.15v, 1.2v with no difference in the temperatures and the only thing that worked is make it run at 2D speeds (400/900) so it runs at 60c celsius. The 100c (105c actually) comes from furmark 1 minute benchmark test at 1024x768, no AA, now I know furmark is made to do this but... seriously? 104c?, I live in a hot clima (Spain) but still at this time of the year the temperatures aren't that hot and I can't think what's going to happen in summer where we reach 40º easily (I have AA in my room, guess this is going to save me).

On games like Medal of Honor 2010 it maxes at 83/84 and I got an ATI driver error couple of times that I tried to OC it a bit, now it's set to default.

The default voltage setting on this card is 1.25v according to SmartDoctor, isn't it a bit high?, should I reduce it?, should it make any difference?, could all this be a problem with my powersupply (500w coolbox)?

Too many questions and no answer, I'm also thinking of buying an Arctic-Cooling Accelero L2 Pro and see what happens as I've read that people is getting -10º~ on temps.

Any help is welcome, thanks in advanced.

More about : eah5770 cucore temps raising 100c

November 6, 2010 11:47:49 PM

Updated to say that it now gets to 103º while playing MoH 2010 too, I'm kind of worried and I won't be able to play until I fix this, any ideas?
Score
0
November 7, 2010 2:05:43 AM

whats your case cooling like? and has it been this way since you got it?
Score
0
Related resources
November 7, 2010 10:22:02 AM

My case is open and the window aims right into it, besides that nothing really.

No, at first it was better, on Windows XP it rarely got to 75º, but now I'm not sure if that was because it used DirectX 9 and stressed the card less or it was anything else, now I'm using Windows 7 for DX11 mostly and at first it was raising to 82/83º max and only raising to 100º on furmark, but last night it got to 103º playing a regular game which never did before, I have to say that I reapplied the thermal compound and maybe that caused it, it's not my first time reapplying it so I know what I'm doing.

Nothing of the changes I make seems to have any significant drop on the temperatures, changing the fan sped from 40% to 80% sometimes doesn't even result on a 1º loss.

Now I ordered the Arctic- Cooling Accelero L2 Pro but it will take a week to arrive so I'm still looking for any solution or tip that makes the temperatures drop because right now I don't think I'll try to play again until I replace the heatsink or thermal compound for a better one.

As I said lowering the voltage from 1.25v to 1.15 (stable) made no difference and I still think 1.25v default voltage is quite high, am I wrong?, the only thing that I think could work is underclock it to a place where it keeps the card cool enough so I can use it until the replacements get here.

Edit: is it normal that it stays at 90% activity for most of the game?
Score
0
November 7, 2010 10:57:52 AM

1.125v is the stock for 5770s, so 1.25 is way too high.

It shouldn't get that hot even highly overclocked, the heatsink on the cucores is supposed to be pretty good.

If it's still on warranty i'd say return it
Score
0
November 7, 2010 11:31:29 AM

wiinippongamer said:
1.125v is the stock for 5770s, so 1.25 is way too high.

It shouldn't get that hot even highly overclocked, the heatsink on the cucores is supposed to be pretty good.

If it's still on warranty i'd say return it



Returning it is not an option as I bought it from eBay some time ago "as brand new" which it seemed, would it be a smart move to change the voltage to 1.25v with smartdoctor then?

The card is this one "http://www.agrupa.es/asus-hd-5770-1gb-cucore-radeon-ati..." and it comes with the voltage tweak thing, it comes with the voltage a little higher to allow some OC which I'm not doing anyway.
Score
0
November 7, 2010 11:54:41 AM

get it down to 1.125v
Score
0
November 7, 2010 12:09:04 PM

wiinippongamer said:
get it down to 1.125v


Yea sorry, that's what I meant.

I did set it to 1.125v now, let's see what happens.

Thanks.
Score
0
November 7, 2010 12:32:08 PM

so does your case have any type of case fans in it? also if you can give me the link or the name of your case.


if you dont have cooling fans i would highly suggest getting a couple as well. but make sure their decently powerful fans.
Score
0
November 7, 2010 3:24:57 PM

My case doesn't really have any extra fan, it's this one "http://www.gamegear.co.za/asus-vento-ta861-2usbaudio-30...", was thinking on buying a new one "http://www.hard-h2o.com/images/reviews/cajas/nox-moonli..." or just add a couple of fans to my current one, but I find it very small and hard to handle when upgrading components and such, also being small I guess it's worse for airflow.

I undervolted it to 1.125v and on idle I see no difference, it's on ~50º (I have two monitors so the clocks are 400/900) and can only get it to 41º by turning AA on in my room.
Score
0
November 7, 2010 4:03:36 PM

but your case side is open right? if so then it shouldn't be a problem for the system heat
Score
0
November 7, 2010 4:35:54 PM

Yes, it is open, but since I live in a hot clima maybe it is better to close it and put some fans, I don't know.

I'm kind of worried about the temps my gpu is getting, the cpu is ok (30 idle - 45 full load degrees), what could cause the card to stress to 90% - 99% activity?, after being there for 10 minutes it starts to raise up to 105º.

What changes should have made to change the voltage to 1.125v?


edit: Editing to say that after playing for about 20 minutes and AA on in my room the card kept cool at 75º top but still got an ATI driver not responding error, don't know what's happening, could this be because undervolting it to 1.125v?

edit again: Set the voltage to 1.150v and crashed at 76º, trying 1.2v now
Score
0
November 7, 2010 5:55:44 PM

if you live in a hot climate i think it would be better to buy a couple case fans to cool the card off. my guess is having absolutely no cooling is whats causing the card to heat up so much.

my friend owns the cucore 5770 and it does get a bit hotter than most 5770's

replacing the whole case isnt really necessary, just add some extra fans.
Score
0
November 7, 2010 6:17:05 PM

Yes, I guess you are right, that would probably explain the heat problem but now that it still gives the ATI recovery thing even at 70º I'm starting to question what the real problem is, it could probably be caused because the undervolt, or just because it is faulty I guess?, wouldn't I see artifacts or other signs if it was faulty tho?

Also, what gpu benchmark could I use to test the stability?, one that doesn't stress it like furmark because I won't feel ok leaving it testing the card at 100ºc, I'll try 3dmark vantage or something like that.
Score
0
November 7, 2010 6:22:08 PM

I live in a hot climate my self and case fans are not an option you need them or its no pc as it gets to hot. Without fans you are only risking killing your hardware or worse a costly house fire. So buy them even the cheap ones as they are better than nothing. Don't get cheap just buy them.
Score
0
November 7, 2010 6:38:13 PM

I'll do that, it's not because I'm cheap to be honest, I just thought keeping it open would be better, guess I was wrong.

I'm still looking for ideas of why it crashes when playing sometimes when the temp is around 75º which is ok (of course to achieve that I have to turn AA on in my room) but just for the sake of testing I'm changing voltages, if you read this whole thread you will see I undervolted it to 1.125v as suggested but it gave ati recovery after 20 minutes, changed it to 1.150 with the same result and now I'm trying 1.2v, ran 3d mark vantage with no error but still I'm not sure where these crashes come from, if it's a voltage thing or just a faulty card, the default voltage was 1.250v.

Thanks.
Score
0
November 7, 2010 7:13:46 PM

For now I just installed a regular home fan directed towards the cpu, it got the temperature down from 51º (idle) to 41º, so I won't have to freeze with AA on until I get the new fans, I'll test the card on games with the default voltage and see what happens, if I still get an ATI recovery I don't know what I will do then.
Score
0
November 7, 2010 7:24:47 PM

You may have permanently damaged the card if you continue to get them then which means that you may have bricked you card but lets hope that you haven't fried it yet.
Score
0
November 7, 2010 10:34:01 PM

Played for an hour with the home fan directly to the open cpu and default 1.25v voltage, didn't get any error or crash to desktop with ati recovery, the card reached a maximum of 85º at certain points but nothing like yesterday (105º), let's hope the new Arctic Cooling L2 Pro with some Noctua NT-H1 fix it, I'll also add some fans to the case if possible, as I said it's very small and nothing really fits in, for now I'll use the home fan as a quick fix and I'll hope not to get more weird ati recovery errors.

I'll assume the card is ok since I was able to play for a whole hour with no error where before I would get one after 15-20 mins, I'll assume it was because I underclocked since after reverting it to default only the temp was an issue.

I'm still open to any ideas or if any of you messed with underclocking can tell me if it's possible to receive such errors.

Thanks everyone.
Score
0
November 7, 2010 10:58:17 PM

With the fans and other things that you bought you should be ok but be sure to keep it clean or you will be back ware you started.
Score
0
November 8, 2010 10:57:06 AM

im guessing that most likely the underclocking was the reason for the errors. usually when i get those its a voltage thing.

just keep the 5770 under 85C and you should be fine, cooler is always better though. as for the case
http://www.comx-computers.co.za/computer-store-specific...

as you can see from the specs you should be able to squeeze a couple 80mm fans in there. also like he said keep your case clean, make sue you regularly clean the dust off all your fans,
Score
0
November 11, 2010 6:07:29 PM

Since You're using smart doctor, do oyu use the smartfan option ? and when you get the 90c~ temps is your fan running at 100% speed? if no then adjust the settings so that your fan would blow at full capacity and that should do alot of cooling :) 
Score
0
Anonymous
November 11, 2010 7:08:32 PM

You need to see this is the worst case senario can happen to HD5770 (stock speed):

http://admins.bnb.gr/~giatrakis/Prime95-toruture-4Gbyte...

Also this the idle after the test it needed some time to cool:
http://admins.bnb.gr/~giatrakis/idle-HD5770-stock.jpg

I did open 2 halls in my box and i have 2 NF-P12 noctua hitting the card I also have NH-D14 noctua cpu cooler, 1 NF-P12 infront and 1 NF-P12 rear, the PSU has 140x140 fan that really helps.

You will get better results from me with the same equipment for sure.
Score
0
November 11, 2010 11:50:25 PM

SOLVED!

I've been having similar problems - 2x 5770's in crossfire, with an ASUS D2X audio card in the 3rd PCIe slot (Rampage II Ext. MB). Running the 'Furmark' stress test was getting temps of 106-107C! 100+ in COD.
Tried updating BIOS - No success.
Removed one card - and the overheat problem went away. I could even overclock without any problems.

At this point, I figured the problem was insufficient air getting INTO the fan on the GPU (as I have an ANTEC 1200 case, with the side fan installed, it wasn't due to not enough air getting into (or out of) the case), and noted that the clearance between the GPU fan intake and the back of the adjacent card was only a couple of mm's.

I noted that the cards (Gigabyte R577UD-1) have a ring (sort of) shaped 'shroud' on the top of the plastic fan cover, which obstructed several millimeters of the clear space between each GPU and the rear of the adjacent card - so I tried cutting a 3cm slot out - tested - still no change. THEN I removed the entire shroud (just the ring shaped bit on top - it unclips easily), and tested again - SUCCESS! AN instant 20+ deg C drop in max temperature, and 15-20 C drop in idle temp. I can now overclock using the ATI auto-clocker, and general graphics performance and reliability has improved.

Score
0
Anonymous
November 12, 2010 11:18:11 AM

steveb64 said:
SOLVED!

I've been having similar problems - 2x 5770's in crossfire, with an ASUS D2X audio card in the 3rd PCIe slot (Rampage II Ext. MB). Running the 'Furmark' stress test was getting temps of 106-107C! 100+ in COD.
Tried updating BIOS - No success.
Removed one card - and the overheat problem went away. I could even overclock without any problems.

At this point, I figured the problem was insufficient air getting INTO the fan on the GPU (as I have an ANTEC 1200 case, with the side fan installed, it wasn't due to not enough air getting into (or out of) the case), and noted that the clearance between the GPU fan intake and the back of the adjacent card was only a couple of mm's.

I noted that the cards (Gigabyte R577UD-1) have a ring (sort of) shaped 'shroud' on the top of the plastic fan cover, which obstructed several millimeters of the clear space between each GPU and the rear of the adjacent card - so I tried cutting a 3cm slot out - tested - still no change. THEN I removed the entire shroud (just the ring shaped bit on top - it unclips easily), and tested again - SUCCESS! AN instant 20+ deg C drop in max temperature, and 15-20 C drop in idle temp. I can now overclock using the ATI auto-clocker, and general graphics performance and reliability has improved.


Finally it didn't worked. After testing i put the ring back on and i did close the 2 edges in the rear as they had it I guess it had allot to do with the way the air flow goes to your box. The reason I have -2 C in the box is because now I have the 2 edges in rear sorter but I guess all this had allot to do again with the way the air is coming to every box.
The extra noise is from the ring and is not placed well is big problem.
Score
0
November 12, 2010 12:42:46 PM

I finally got the Accelero L2 Pro, after installing it the temperatures dropped quite a bit, from 104º on furmark to 80º, now it sits at around 39º idle and 70º when playing.

But I can't control the RPM on the Accelero FAN, I had to use the adapter molex and it's on 12v but I can easily tell that it isn't running at max speed because I can't even hear it, also it makes no difference using the 7v adapter or the 12v, any suggestions? could it be an option in the bios?

ps: trying to control it from ccc or fanspeed has no effect because it's not connected to the graphic card (the eah5770 has a very tiny 3 pin connector while the accelero has 2 bigger ones that won't fit).
Score
0
Anonymous
November 12, 2010 1:51:07 PM

I've seen Accelero L2 Pro the only thing it has better from the stock cooler is the 92x92 fan in fact the stock is better because the large copper that touches the GPU and I’m sure if a manage to find a 92x92 fan to the stock will perform better.

In the reviews of Accelero L2 Pro they were getting -15 C you claim you got -25 C can you run the same test I did with the Prime95 and the Furmark extreme for 12 minutes and give us a screen shot? You have seen i was 63 C and i went to 60-61 C and if the data you giving are correct it may also help us.

I toled you you have to increase the airflow to the box and i did it with 2 NF-P12.
Score
0
November 12, 2010 2:06:52 PM

Quote:
I've seen Accelero L2 Pro the only thing it has better from the stock cooler is the 92x92 fan in fact the stock is better because the large copper that touches the GPU and I’m sure if a manage to find a 92x92 fan to the stock will perform better.

In the reviews of Accelero L2 Pro they were getting -15 C you claim you got -25 C can you run the same test I did with the Prime95 and the Furmark extreme for 12 minutes and give us a screen shot? You have seen i was 63 C and i went to 60-61 C and if the data you giving are correct it may also help us.

I toled you you have to increase the airflow to the box and i did it with 2 NF-P12.



I have no problems doing this, but I would prefer getting the fan to work properly before doing so.

I sent an email to arctic cooling support and now I'm waiting for a reply, as soon as I get it to work I'll try what you said, can you tell me what tests did you do exactly?, I never really used furmark that much and never tried Prime95, so any detailed instructions would help do the test properly.
Score
0
November 12, 2010 3:25:33 PM

Could I just connect the 3 pin header to a sys_fan connector on the MOBO?
Score
0
Anonymous
November 12, 2010 4:44:11 PM

I have all the fans directly plugged in the psu and only in the power fan the front and the rear fan in system. if you have power fan connectors (see your manual will play better) this case of fans are working mostly with the PSU power leave it like this, is better, needs 4 pins fan to plug it to the card and have speed control from there.
Score
0
Anonymous
November 13, 2010 8:10:10 AM

Shuryou finaly i did put all back i have hopes in you if you managed to make it work and you do the test with prime95 and furmark they way i did it and you hit less than 63 tell me about ok ?
Score
0
Anonymous
November 14, 2010 10:31:18 AM

Shuryou Confirmed and be careful finally in the sys fans the volts dropped when the CPU, card work at max and I have only the rear fan (NF-P12) in the Power Fan all the others are plugged in the PSU by doing that even in the hardest tests the speed of rear fan never drops under the 1398 and idles at 1403 minimal. You can see a HARD prime95 test 4096 FFT 4Gbytes, furmark 1.8 extreme test 35 minutes here:

http://admins.bnb.gr/~giatrakis/Prime95-torture-4Gbytes...

It never breaks 65 C, now is 14:16 here at Greece and the day is hot, and it plays 63-64-65 66 max. In here the Idle after the test -3 C by putting only the rear fan to Power Fan and all the others to PSU:

http://admins.bnb.gr/~giatrakis/Prime95-torture-4Gbytes...

Bottom line the OC to the card isn’t worth it the +10 C that gives is HUGE Problem and the benefit is very small 3-10 FPS the additional watts are 8 at idle and over 20 at full load i have the card at stock speed. Only the Power Fan connector drops the voltages less that all the other fans in the motherboard and if you have one connect the rear only there. A strong psu can deliver more stable power to the fans and keep your box cool.

Tell me about if you got stable ok?
Score
0
November 15, 2010 10:25:30 PM

My temp is higher than yours Giatrakis, just 1 minute of furmark puts my temp to 76º.

I still claim that I got 25º less degrees but it doesn't mean it's all thanks to the accelero l2 pro, maybe my stock heatsink/fan was faulty, but the reason I opened this topic was because my card was getting to 105º degrees playing games, and now it varely reaches 75º with 910/1375 OC which is 30 degrees less, I guess we can end this here, thanks everyone for the help.
Score
0

Best solution

Anonymous
November 16, 2010 9:44:18 AM

Yes definitely is nuts case I mean think about it 40nm GPU you know what it means? The Processor is smaller than the HALF of the head to a needle!! A 6pin 12 Volt connector from the PSU can deliver 75 Watts (normal, OCed....)!! The heater for the water we have in our homes never in our wilders dreams have such an efficiency of creating heat !!! Finally I have down clock even more the half head of this needle and it was HUGE mistake to cut the edges with some tape I manage to drive the air to the exit of a card and drop more the temperature in the box. See here 1hour testing running the HD5770 at 750Mhz GPU 950Mhz GDDR5:

http://admins.bnb.gr/~giatrakis/prime95-4Gbytes-downclo...

You can compare these results when I open a hall in my box and added a second NF-P12 to hit the card. I was OC to the card 950Mhz GPU, 1350Mhz GDDR5 in here:

http://admins.bnb.gr/~giatrakis/Prime-Turture-4Gbytes-F...

As you can see the drop is HUGE and to the watts and to the temperature in the box -5C the benefit was +9 FPS at average and +9 FPS at Max and the Temperature in the box was 41C the tape and the ONLY the rear fan to the power connector helped me never to go above 38 C the box!

The other NUTS case is the SYSTEM fans speed and the Power connector speed!! I mean think about it I’m lucky to have the extra 12volts ATX molex to pci-e connector on board YET the card and the processor when they work at maximum as you see above the drop in the volts that the motherboard can deliver to the fans is HUGE!! Compare the The rear fan only to the power connector it operates in 1395 rounds per minute as you can see the idle here:

http://admins.bnb.gr/~giatrakis/idle-changes.jpg

I have 1403 rounds per minute MINIMUM to the rear fan finally you getting why I say these chips they look to me like BLACK HALLS? They can eat all the power from your motherboard and if you don’t have enough capacitors you may have BIG problems!

Finally this motherboard compared to the old one is the only choice i have to help me go Above 4 Ghz to the CPU and CF much more better:

http://www.tomshardware.co.uk/forum/forum2.php?config=t...

This is WHY I wrote the F**** word to the companies that did the things like that and I hope Tom’s hardware to forgive me about it.

Ps: in order to make this tread more reliable here Prime95 Torture 64bits 4096 minmal FFT, 4Gbytes ram (4000), Furmark 1.8 Hd5770 HD5770 Oveclocked at 950Mhz GPU, 1350Mhz 45 Minutes test:

http://admins.bnb.gr/~giatrakis/Prime95-4Gbytes-Furmark...

Even to 1 hour nothing is changed trust me i don't have so mutch time right now the time is 16:13 in Thessaloniki, Greece. Finally the Tape I used to drive the air out of the box worth it -3C to the box confirmed don’t remove the ring from the Gigabyte R577UD-1 as previous suggested it helps to keep the air out of the box. Is the OC to the card worth it? I believe NOT in the winder with these results the summer what will happen? figure it out...

Choose the best answer and close this thread we all did our best it’s really a nice thread. I need to test this cooler you say i may get it.

Ps 2nd: NH-D14 had 140x140 Fan on it and i put it to the Power Fan connector hehehehehe i mean is even large enough and hits under below the heat pipes also closer to the motherboard and what a difference ? I have 69 maximum to the card oced, in the CPU the none of my cores goes to above 58 and the box is stable at 36!!! I mean in the end a motherboard with 4 power fans and some Noctua 140x140 connected in there can drop everything to bad i don't have much space to put an 140x140 to a hall in box midi tower...

It get 1250 rpm to idle and 1245 in full load the drop to the motherboard is big... But now the temperature is very low night I have to run more tests tomorrow at the same time to confirm what I already know is better this way Shuryou do it if you can this is the solution but be careful to find Filters because in these speeds the fans will bring HUGE dust inside the box =;-)

NF-P14 FLX noctua:

http://www.noctua.at/main.php?show=productview&products...

NH-D14 cpu cooler:

http://www.noctua.at/main.php?show=productview&products...
Share
November 23, 2010 7:11:18 AM

Best answer selected by Shuryou.
Score
0
November 23, 2010 8:35:00 AM

This topic has been closed by Mousemonkey
Score
0
!