HIS 7970 IceQ X2 fan behaviour?

[fizzy]

Hi,

I have a HIS 7970 IceQ X2 card, and while delving into my machine for another issue I noticed that on idle one of the fans is not spinning - the one nearest the end plate/video connectors. The other one is spinning as expected.

Running gfx benchmarks I noticed that the fan twitches occasionally while the other fan gradually spins faster as the card temp rises. Once it gets to a point the 2nd fan starts spinning.

I don't know whether this is a noise limiting issue or a temp fan control, or a sticky fan which is only spinning once it get's a bit more voltage through it. Can't recall if they both span on idle before when it was new.....?

Does anybody else have this same card, and how do your fans behave?

It's not overheating at the moment and temps seem reasonable - but if the 2nd fan fails completely it will obviously cause cooling issues, so may need to RMA it.

 

[evohicks]

Hi,

I have a HIS 7970 IceQ X2 card, and while delving into my machine for another issue I noticed that on idle one of the fans is not spinning - the one nearest the end plate/video connectors. The other one is spinning as expected.

Running gfx benchmarks I noticed that the fan twitches occasionally while the other fan gradually spins faster as the card temp rises. Once it gets to a point the 2nd fan starts spinning.

I don't know whether this is a noise limiting issue or a temp fan control, or a sticky fan which is only spinning once it get's a bit more voltage through it. Can't recall if they both span on idle before when it was new.....?

Does anybody else have this same card, and how do your fans behave?
I have 2 of these and both fans are spinning @idle on both cards.
Just out of interest do you run Afterburner with ULPS disabled?

It's not overheating at the moment and temps seem reasonable - but if the 2nd fan fails completely it will obviously cause cooling issues, so may need to RMA it.

I have 2 of these and after just checking, both cards have both fans spinning at idle.
Just out of interest do you have Afterburner installed with ULPS disabled?

 

[fizzy]

I have 2 of these and after just checking, both cards have both fans spinning at idle.
Just out of interest do you have Afterburner installed with ULPS disabled?

I have 2 of these and after just checking, both cards have both fans spinning at idle.
Just out of interest do you have Afterburner installed with ULPS disabled?

Thanks for checking - suspect mine is misbehaving then.

I've not got afterburner installed but manually disabled ULPS via registry editing.

Just had a check and looks like both fans are running off the same four pin fan header on the board so no obvious way it could differentiate between the two. Looks like the fan is getting weak then and not spinning at low volts or pwm setting.

Just been doing some tests via CCC overdrive - and the fan doesn't start spinning until I up the fan speed to at least 75%, but then stays running as I drop the speed until I drop it to 40% or below. Looks like the fan is on it's way out then.

I wonder how hard it would be to get a replacement fan and fit myself rather than have to go through the RMA process...

 

[fizzy]

Hmm - just had a close look at the card after removing it from the pc, and even though both fans are connected into the 4 pin header, the front (working normally) fan only uses 3 wires where as the end fan (the "problem" one) is actually using four...?

Fan seems to be spinning freely, doesn't feel jammed up or anything at all. Could possibly be an issue with the card or whatever signal is on the 4th wire (a green one)?

CCC and monitoring tools are giving a fan speed reading when used even when the rear fan has stopped, so doesn't look like the extra wire is a speed sensor one...

Edit: Right, seems to be temp controlled (or affected by temp). Set CCC to 40% fan and started a demo. When GPU temp hit around 69c, the rear fan started up, still at 40% reading in hwinfo/ccc. Quite demo and when GPU temp dropped to 44c, rear fan stopped. Repeated this a few times - seems too consistent to be a fault, but not something I've spotted before. Odd.