Weird resonance DC Fan PWM

[made2hack]

Hi all,

Below is my test setup, I want to control the fan speed using 555 pwm.



A few issues I'm encountering.

I am limited to a 5K pot and, using a 470Ohm resistor for R6, I can get duty cycle as low as 9-10% and max of 99% which is perfectly acceptable. C3 is 10nF.

Now, when C4 is also 10nF, all works well, such that there is no resonance / buzzing sound coming from the DC Fan. When I change C4 to 0.1uF (100nF), there is a rather annoying high pitch ringing coming from the fan and it persists until around 92% duty cycle. When I change C4 to 1uF, there is a lower frequency buzzing (like a whine) as opposed to a ringing and it too persists until around 92% duty cycle.

With C4 = 10nF, at 10% duty, the fan draws 60mA @ 4.5V (around 270mW)
With C4 = 100nF, at 10% duty, the fan draws 36mA @ 5.7V (around 205mW)
With C4 = 1uF, at 10% duty, the fan draws 18mA @ 2.7V (around 50mW)
At 99% duty cycle, it draws 150mA @ 12V (or close to 1.9W) with All of the capacitors

First off, I use a Fluke 21 (original) that is not True RMS so I'm not sure whether I am getting accurate readings of amps. Does having True RMS matter for reading DC Amps?

So, the 2 Big questions are:
1 - Am I getting accurate amp readings with my meter? Should 10% duty cycle translate to 10% power consumption? IE with amperage and voltage taken into consideration? For C4 = 10nF and C4 = 100nF, it seems to be the case where a 10% duty cycle is close to 10% Power usage. But not for C4 = 1uF.

2 - Why the ringing / buzzing / resonance when the caps are 100nF and 1uF respectively?

Thanks,

[Gyro]

You might find this of interest...

https://www.eevblog.com/forum/repair/cured-my-gps2032-psu-rattle/msg931131/#msg931131

[Neilm]

Are you sure the fan does not have any electronics in it? I recently went to use a fan in a similar manner and had several problems. Checked the datasheet for the fan and found that it had some anti-stall electronics in it - there was a small note down the bottom not to try to drive it by putting a PWM signal on the supply.

[made2hack]

I'm pretty sure it does have some electronics in it, granted maybe not anti-stall, however the datasheet does specify operating 4.5V - 13.8V? Maybe this is the reason?

It is the Sunon HAC0251S4-000U-999. 2 phase 4 pole brushless dc motor.

[bktemp]

It is the Sunon HAC0251S4-000U-999. 2 phase 4 pole brushless dc motor.

Did you read page 11? It clearly says do not use PWM for speed control.

[Gyro]

Any brushless motor is going to have electronics in it by default. The problem in my case (the GW Instek PSU) was that the PWM was beating with the motor's electronic commutation at some speeds. The PWM edges were far too sharp. Maybe time for a PWM controlled (4 wire) PC fan, you could probably simplify your drive circuit then too (logic input).

[made2hack]

  Glossed over all of those warnings.

What are the potential problems of using PWM (other than the annoying noise) ?

[bktemp]

Some fans contain a capacitor at the dc input to smooth the commutation ripple generated by the fan. If you apply a PWM, the capacitor gets charged and discharged in every cycle. This generates a high inrush current in every PWM cycle.

[Neilm]

Some fans contain a capacitor at the dc input to smooth the commutation ripple generated by the fan. If you apply a PWM, the capacitor gets charged and discharged in every cycle. This generates a high inrush current in every PWM cycle.

Depending on the capacitor, it might start getting hot leading to premature failure

[Zero999]

Glossed over all of those warnings.

What are the potential problems of using PWM (other than the annoying noise) ?

What about adding an inductor and increasing the frequency, forming a simple, unregulated, open loop, buck switching power supply?

[bktemp]

Glossed over all of those warnings.

What are the potential problems of using PWM (other than the annoying noise) ?

What about adding an inductor and increasing the frequency, forming a simple, unregulated, open loop, buck switching power supply?

It works fine (if the generated voltage is high enough for the control ic to operate correctly).
If you want to use the the fan speed sensor you need a highside switch, inductor, diode, capacitor. This increases the BOM considerably.
The best solution is a 4 wire fan, because it uses the motor coils for smoothing the PWM signal, similiar to applying PWM to a brushed motor. Because the control ic always gets the full supply voltage, the speed control range is better.

[matkar]

In your case I also suggest using a 4 wire fan. The recommended PWM frequency is 25kHz. I have used 62,5kHz with several brands of 4 wire fans and pumps and had no problems. I don't know what are the consequences of using lower frequency than 25kHz.