# How do I troubleshoot this high-frequency PWM fan control circuit?

I have a 230 Volts fan from EBM papst. It is a fan that can be controlled through a PWM signal (between 1 and 10 kHz). I have built the circuit that is in the manual (see image)

I did get this working using a 5.0 Volt arduino clone. I have been able to change one of the timers, so I am able to get a clean 4 kHz pwm signal on digitalOut 4. The fan responds nicely.

But I would like to get it working on another arduino clone that runs on 3.3 volt, currently I am getting no response at all from the 3.3V circuit.

The only thing I can think of that is different is the PWM voltage. The 3.3 volt arduino clone only makes 3.3 volt peak-to-peak, whereas the other arduino does to at 5.0 volt.

My question, can this be solved by choosing a different transistor? And if so, what specifications should I look for?

And otherwise, what other information is useful to know? I have some basic measurement equipment including a simple oscilloscope, but I don't know what to look for.

Page 4, not the exact same fan, but the same schematic, could not find the correct one at the moment.

• Can you add a pointer to where you got that schematic? It looks like it is going into a control circuit that looks similar to an Intel-style PWM fan control, but it is a little different. Commented Feb 24, 2015 at 15:53
• How did you pick the value of 15K ohm for R4? Commented Feb 24, 2015 at 16:07
• A friend of mine helped me a bit with that, if think he calculated it based on the voltages, but this was based on the 15 volt supply and 5 volt mcu. I'll ask again to see what formula he used and post it here. Commented Feb 24, 2015 at 19:48

According to the above drawing, it looks like the fan wants to see a 0-10 volt swing on the PWM input (yellow wire) and will supply 10 volts DC on the Voltage output (red wire).

simulate this circuit – Schematic created using CircuitLab

If you simulate the above circuit, it looks like a 3.3 volt input to R1 should be enough to saturate the transistor and provide almost a full rail to rail swing on the fan's PWM input.

Can you verify that your connections match the above schematic?

Some tests to try...

1. Disconnect the pin from the input to R1 and use a wire to connect the resistor's input directly to the 3.3 volt power supply. Does the Fan come on full speed?

2. Now connect that same wire to ground. Does the fan turn off?

3. Attach a volt meter to to the collector of the transistor. You should see 10+ volts there when the input to R1 is connected to ground, and 0 volts there when the input to R1 is connected to +3.3 volts.

• The fan also accepts a 0-10 voltage on red, but I'm trying to get the own working. I definitely have it connected to the yellow wire. I'll try your proposed tests tomorrow and post the results! Commented Feb 24, 2015 at 19:43
• It should not matter if the 10/15V source comes from the fan or you provide it, but the drawing from the fan manufacturer does seem to say that the PWM input should be 0-10 volts (either analog or generated by PWM). If so, you would want the top of R4 to be connected to +10V DC. Commented Feb 24, 2015 at 19:48
• I mistyped, own should have been PWM. I've read the drawing from the fan manufacturer in a different way. There are 4 distinct rectangles, the left rectangle is a PWM signal, between 1 and 10 kHz, should be above 10 volts. The right rectangle shows a 1-10 voltage control. The 5 volt version I've got working is only connected to yellow and blue. I've assumed the PWM was doing its work, I'll check tomorrow if that's is actually the case or I've been controlling from 1-10 volt by mistake. Commented Feb 24, 2015 at 20:18