Page 1 of 1
Problem with duetwifi's cpu fan
Posted: Wed Jul 19, 2017 12:28 pm
by chuyuanver
Hi all,
I've been conducting the duetwifi converting recently. There are quite a few problems I had during this time but wisely using the internet solves these. However, there is one thing left unsolved (which is not too urgent either). Here is a picture stolen from mhackney's blog.
The circled wire I believe is the cpu fan and the orange one next to it is the layer fan. So the problem is upon turning on the layer fan up to 50%, I can feel the CPU fan becomes significantly slower, causing a temperature increase in CPU. With CPU fan on, the temperature is about 30 degree C while with decreased rate of CPU fan, the temperature could go up to 37 or higher. I am not sure how high the CPU temp can go so I am a little worried about this.
Yue
Re: Problem with duetwifi's cpu fan
Posted: Wed Jul 19, 2017 1:44 pm
by Xenocrates
These sorts of microprocessors are easily able to exceed 50C. the specific chip is rated with an absolute max of 105C, so you are well in the safe zone. However, that seems to be a sign of weak 12V power regulation, which is another serious issue. Please check your power supply output while at idle, and under the load where you see that condition, and let us know what power supply style you have.
Re: Problem with duetwifi's cpu fan
Posted: Wed Jul 19, 2017 1:52 pm
by chuyuanver
OK, thanks Xenocrates. I have the stock MAX V3 PSU and I'll check the output after I get back from work. Just to make it clear, do you mean I should check the voltage between the pins of the fan or between the pins of the actual power supply?
Re: Problem with duetwifi's cpu fan
Posted: Wed Jul 19, 2017 2:24 pm
by Xenocrates
I would go with the actual power supply, as that will tell you if it's a regulation problem, instead of the fans, where it could just be a 'getting warm and experiencing some voltage sag' problem with the onboard traces.
Re: Problem with duetwifi's cpu fan
Posted: Wed Jul 19, 2017 8:39 pm
by chuyuanver
Xenocrates, I've tried to measure the voltage between the power supply pins, which is indicated in the picture shown below (again stolen from mhackney's blog)
I get the same voltage 12.4 V with turning on the layer fan and without. I also measured the voltage in the cpu fan, 0.8 V with layer fans on and 10V off. Another thing I have to admit is that I accidentally shorted the fan pins once and at that time there is a spark and some smell. I read this doc:
https://duet3d.com/wiki/How_to_destroy_ ... t_Ethernet and they claimed that they have a self-resetting fuse so it won't affect the performance of the board.
It seems that my layer fan and the CPU (always on) fan are now sharing the voltage from the power supply since one goes up with the other goes down at the same time.
Re: Problem with duetwifi's cpu fan
Posted: Fri Jul 21, 2017 10:58 am
by Xenocrates
That sounds somewhat nuts. The always on connectors should always output the same voltage as the input. I would connect the CPU fan directly to the power supply if you're particularly concerned (And potentially send Dave Crocker a message if he doesn't jump in here, given the weird behavior).
Re: Problem with duetwifi's cpu fan
Posted: Thu Jul 27, 2017 12:10 pm
by dc42
The self resetting fuse doesn't protect the fan circuits.
If you shorted out an always on fan connector, this typically fuses the trace between the VIN terminal block and the VIN pin on the fan voltage selector jumper block. This trace can easily be bypassed using a piece of wire on the back of the board. There is a thread about this on the Duet3D forum.
HTH David
Re: Problem with duetwifi's cpu fan
Posted: Thu Aug 03, 2017 11:33 am
by chuyuanver
Thanks David, I connected the V_in voltage selector and the V_in Power in and it works just as before!