1.2A on CPU fan header

[jvroig]

I have a 786GM-US2H that I just managed to kill last night while testing my brand-new Thermaltake Frio OCK to replace my dual-fan Hyper 212+ on an overclocked, overvolted 1090T.

The good news is that there was >10C difference between the two. The Frio OCK is much louder (still better than AMD's stock heatpipe cooler at full blast, though), but there's a knob to control it, and fan control software in both Windows and Linux are so efficient that it is not a real problem for me in daily usage.

The bad news, of course, is that a mere 15 minutes in P95 small fft ended up killing the board. This is not a new board - it housed an Athlon X2 7750 and a Phenom II 965 BE before the current 1090T. It could very well have died of abuse (or "natural enthusiast causes").

However, one thing that worries me is that it might be the CPU fan header that caused the catastrophic failure. I'm pretty sure most cpu fan headers are only rated for 1A, and the Frio OCK (to my surprise) is rated at 1.2A (2x 130mm fans on one header). I don't know why I missed that. If I remember correctly, most CPU fan headers can only handle 1A, and only the highest end mobos go up to 2A. I just can't seem to verify this information from motherboard manuals.

I'm not sure if this is a valid concern, so I'm throwing this out to the Cooling gurus here. How likely is it that it was the CPU fan header causing the catastrophic failure of the mobo due to having to power the Frio OCK's dual fans?

 

[artvscommerce]

Sounds reasonable. However I've checked this spec on (2) different boards in the past. One of them was rated @ 1.6 and one was one was 1.4 if I remember correctly.

 

[aigomorla]

and this is why i tell people:

NO HIGHPOWERED FANS ON ANY MOTHERBOARD HEADERS!!!


1.2amps is insane for a fan.... is it a nidac fan?
Typically they run no more then .58amps on even the ones which can leave a deep cut in your finger should it get in the blades.

 

[Modelworks]

It isn't the header that is the problem, it is the circuitry it connects to. They use FET for switching and they don't put a lot of room on the pc board copper to use for a heatsink. The way to solve this is use a pass transistor. The pass trans will handle the current load and you would attach the pass trans to the case to use that as the heatsink. Cost is around $5 in parts and not really a complex circuit.
http://talkingelectronics.com/projects/200TrCcts/200TrCcts.html

 

[aigomorla]

It isn't the header that is the problem, it is the circuitry it connects to. They use FET for switching and they don't put a lot of room on the pc board copper to use for a heatsink. The way to solve this is use a pass transistor. The pass trans will handle the current load and you would attach the pass trans to the case to use that as the heatsink. Cost is around $5 in parts and not really a complex circuit.
http://talkingelectronics.com/projects/200TrCcts/200TrCcts.html

or u can get a cheap fan controller (sunbeam) which can support up to 1.5amps per channel.

 

[jvroig]

Right, thanks all.

@artvscommerce: Where did you find such specs? I went around hunting for one in the manufacturer's website, product page, my own manual... nothing.

@aigo: yeah, I wouldn't have put high-powered fans (I didn't even have high-powered fans until this Frio OCK), but the product itself had twin fans AND only one cable for both of them. Each fan is rated at 0.6A, so it comes to a total of 1.2A on one header. Luckily, I scrounged around my "left-over parts" box, and found a 3-pin to IDE adapter, so now I'm just going to connect it straight to the PSU and use the built-in fan control knob to set the fan speed. This is way better, right?

@Modelworks: Thanks, will take a look at that. Aside from the pass trans, the 2 books actually look like an interesting educational read since I am not too keen on electronics but always wanted to learn more about it "when I get time" (isn't that always the excuse).


Cheers!