Fan header voltages?

duronboy

Gawd
Joined
Feb 1, 2003
Messages
549
Looking at GTX 600 replacement fans, they're listed as being 12V. I'm not getting any fan movement on mine, but the voltages are pretty low from the header, too. The highest voltage I get is about 3.3V.

I can't test the fan externally because I don't know the pinout. And, all the wires going into the fan motor are black. And of course because computer fans are mostly 3-phase jobbers behind an on-fan controller, I'm not sure spinning the fan by hand and testing the outputs will reveal what I need to know.
 
Looking at GTX 600 replacement fans, they're listed as being 12V. I'm not getting any fan movement on mine, but the voltages are pretty low from the header, too. The highest voltage I get is about 3.3V.

I can't test the fan externally because I don't know the pinout. And, all the wires going into the fan motor are black. And of course because computer fans are mostly 3-phase jobbers behind an on-fan controller, I'm not sure spinning the fan by hand and testing the outputs will reveal what I need to know.
My assumption would be the base, or lowest, voltage would be 3.3v or off. I would suspect the voltage is controlled by the card based on its fan curve and temperatures.
 
Yeah I did that and the highest setting can actually set is 85%, and that doesn't change the highest voltage. Some of the other voltages change and one of them even goes lower. But none are over 3.3V. The 3.3V(actually 3.287 or something like that) is unchanged when the fan control percentage is changed.
 
install afterburner and set the fan to 100% and then measure, see what you get.

Yeah I did that and the highest setting can actually set is 85%, and that doesn't change the highest voltage. Some of the other voltages change and one of them even goes lower. But none are over 3.3V. The 3.3V(actually 3.287 or something like that) is unchanged when the fan control percentage is changed.


Sometimes a card will default to a baseline voltage if it cannot detect a fan installed.
 
So I wonder if the fan is indeed dead and the card when the fan is disconnected is like, "well, you're at 0 RPM at 3.3V, might as well not waste any more voltage on ya, ya damn bum."
 
So no workie. I plugged the fan I suspected to be broken back in and used a hand drill with a little round grinding wheel attached to spin the fan housing a little faster than I was able to flick it to. Tried both directions and it stays pegged at reading 0. So my guess is the fan header is borked on this here puppy.
 
Ummm... Are you hoping to get voltage at the pins by SPINNING the fan?

I mean, an oscilloscope would give a reading but not in voltage. More of an on/off of a hall effect sensor.

You need the voltage to spin the fan, not the other way around.
 
Since a fan that works in another card doesn't work in this card, I was thinking it might be interesting to see if the non-working card could at least report spinning if the fan was spinning. As stated earlier, as computer fans are 3-phase, you can't use a voltmeter to check to see if a fan works by spinning the fan.

I'm going to remove the fan on this card with weird header voltages and see if I can plug it in the card that had working fans.
 
Wow what a nightmare. So removing the fan is proving to be more difficult that I believe it should effing be. I already had to dremel a slot in a screw for the frame, but now I had to slot another screw for the fan because of the thread locker. But now the fan appears glued on, too. Neat-0.
 
Wow what a nightmare. So removing the fan is proving to be more difficult that I believe it should effing be. I already had to dremel a slot in a screw for the frame, but now I had to slot another screw for the fan because of the thread locker. But now the fan appears glued on, too. Neat-0.
There may be a hidden screw under a removable plastic piece or sticker, or it may actually be glued on. Never know with these things...
 
Came across a post saying they were glued in. Only way to remove them is destroy them.

I was able to arrange the frame so I could plug the fan into the working card header. No go, so it's a bad fan. Noice. Or it somehow is a different pinout even though the connector is the same.

Back to the drawing board. The card works, it just won't power a fan. I'm going to have to come up with a cooling solution that is externally powered and is quiet enough that it's not annoying when the card is at low to no load. So... big-ass CPU heatsink, I guess.
 
A double stick tape pad is what the stock blower is held on with. A little heat would have made it easier to remove. Anyway, now it's gone.

I bolted an 80mm fan to the hole. Powered it with a mobo header that apparently responds to system temp.

Took the GPU 1:12 to reach 80c in Kombustor.

Found a new header that's probably 12V full-time and it took the GPU 1:38 to reach 80c.

Took the card out to the shop and made the fan hole larger. Took the card 2:05 to reach 80c.

I thought I'd try some apps.

CS GO 70c average peak of 78c
Sniper Ghost Warrior peaked at 60c.
Fallout New Vegas peaked at 47c.
Quake Live 2 or 3c above idle.
Sketchup 3c above idle.
Davinci Resolve barely utilizes the CPU or GPU for some reason. Can't immediately identify the bottleneck.

I was going to try some rendeders but I couldn't get them to install.
 
Back
Top