the 4870 LEDS

[painkilla]

So i just installed accelero s1 on my 4870 and on booting i saw 3 leds flash up. In windows all 4 of them are off until i start furmark for about 2 minutes and the core temp reaches about 50C. Then the second led turns on and furmark crashes. What can be the problem and maybe more important what do all the leds mean?

I have to say i didnt hava sinks on the rightside components( those black small square things 3 down 2 up next to the big vitec and pulse components) as i have only 1 long sink which doesnt fit on the cart.

Already plugged out some devices and reset my cpu overclock but didnt work so power is not an issue. Also the core and memory temps are fine so...

[Willhub]

So nothing on the ram? Could be overheating maybe?

[painkilla]

well the strangest thing is when i put back stock cooler i have no problem at all. ram does have heatsinks only the little blocks next to Vitec and pulse chips not. So the only explanaton would be that the second led is warning for overheating VRM's?

[Willhub]

Well I'd put heatsinks on everything. You got a 92mm fan or something on the S1 too?

[babalouj]

I have the same problem. I have my 4870 watercooled and every sinked but sometimes the 2nd led comes on and then everything is unstable.

[painkilla]

Well I'd put heatsinks on everything. You got a 92mm fan or something on the S1 too?

got 2 120mm fans on it with tie raps

[dennismagic]

i have the same problem too. watercooled the core and sinked every ram and mosoft.
only crashed on furmark but stable again if aa is enabled!

[dnottis]

I have the same issue as well. Stock cooler, duo orb and water cooled. So I emailed my buddy Terry Makedon at ATI and got a response about this. Looks like Furmark is causing the cards to heat up more than any other app will, however it looks like it's more of a driver issue with the black screen and not a hardware issue. They have been able to duplicate the problem, however no application will get it that hot in a real world scenario. I think its trying to kick in VPU recover and its hanging up. However that still doesnt explain why Furmark is killing cards, in fact Nvidia has the same issue with the black screen and Furmark - also a few cards have been killed while running Furmark.

For now, use another GPU stress test and not Furmark.

[RejZoR]

Well FurMark really stresses the shader engine hard resulting in quiet higher temperature and GPU load. If you are looking for 100% stability, FurMark is ideal tool for testing.
Otherwise use RTHDRIBL tool and crank up some FSAA to increase GPU load in a linear way (1x - 16x FSAA).

[painkilla]

video cards shouldn't just die from load, that's totally ridiculous. They are made for it. Besides the 4800 series is able to endure up to 100C? So its not possible to kill it with only 80/90 C or am i wrong.

[erocker]

Check your VRM temperatures! They are circled in the GPU-Z pic:

[babalouj]

my VRM temps never get above 50C but the 2nd LED still comes on and then everything is unstable.

[tiro_uspsss]

i have the same problem too. watercooled the core and sinked every ram and mosoft.
only crashed on furmark but stable again if aa is enabled!

would u like some 'post content' with ur sig sir?

[purecain]

i wonder if the clock states are changing mid bench... causing this issue...

[xMrBunglex]

what we need is a list telling us what each LED code means.

[NKD]

I dont think not testing for stability with furmark means that your card is not stable, My card crashes at 840 with furmark and sometimes it runs for 20 mins and sometimes crashes right away, and this benchmark does test for stablility but hours of gaming at 840 never gave me any artifacts or crashes, I believe it is either a driver issue or in reality furmark is just too harsh on these cards becuase the temps never go above 60c under load while gaming with fan at 42% and with furmark no matter what the fan speed is the temp shoots up to 70-80 and even high 80's under some instances. I ain't complaining about furmark, but I don't think game stability has anything to do with furmark stability. it is like prime95 of gpu's.

[Rickpatbrown]

I have the same issue as well. Stock cooler, duo orb and water cooled. So I emailed my buddy Terry Makedon at ATI and got a response about this. Looks like Furmark is causing the cards to heat up more than any other app will, however it looks like it's more of a driver issue with the black screen and not a hardware issue. They have been able to duplicate the problem, however no application will get it that hot in a real world scenario. I think its trying to kick in VPU recover and its hanging up. However that still doesnt explain why Furmark is killing cards, in fact Nvidia has the same issue with the black screen and Furmark - also a few cards have been killed while running Furmark.

For now, use another GPU stress test and not Furmark.

Sounds like dnottis knows what's up.

[Brave758]

OK, for the HD4870 cards .. the LEDs are as follows :-

D1601 - Red LED On, shows critical temperature fault

D1602 - Red LED On, shows External power connector A was removed

D1603 - Red LED On, shows External power connector B was removed

D601 - Red LED On, shows critical Core power fault

[cantankerous]

great, thanks for this info!

[babalouj]

Thats great info! My card hasn't had this problem since last week for some reason.

[deathman20]

Check your VRM temperatures! They are circled in the GPU-Z pic:

Man those temps look so cold compared to my 4850

OK, for the HD4870 cards .. the LEDs are as follows :-
D1601 - Red LED On, shows critical temperature fault
D1602 - Red LED On, shows External power connector A was removed
D1603 - Red LED On, shows External power connector B was removed
D601 - Red LED On, shows critical Core power fault

Would that be similar to whats on the 4850 then?

[msarp]

here are my temps in a P182. my new 4870

[erocker]

here are my temps in a P182. my new 4870

Do you have your air condtioner blowing in it?! They are entirely too low to be in normal room temperature!

[SPQQKY]

Do you have your air condtioner blowing in it?! They are entirely too low to be in normal room temperature!

Exactly what I was thinking.

[deathman20]

Indeed considering NY isn't the coldest of places excpecially at this time of the year.