trouble with CUI fan

achrn

I'm trying to use a CUI 40mm 4-wire fan and it's not responding right to PWM. It's on OUT1 on a toolboard on a Mini5. The Mini5 is on 3.4.5 and the toolboard is on 3.4.4 (which is latest for expansion boards, I believe).

Fan is this: https://www.digikey.co.uk/en/products/detail/cui-devices/CFM-4010-13-22/6194965

The datasheet 'detail' about PWM is a joke: "A speed control lead can be provided that will accept a PWM signal from the customer circuit to vary the speed of the fan. The change in speed is linear by changing the Duty-Cycle of the PWM. Open collector type and pull-up voltage is changed by maximum operating voltage and sink current by consuming current. (See Figure 4 below)."

All four wires are connected - gnd black, vout red, tach yellow, pwm blue.

If I configure as

M950 F0 C"!121.out1+out1.tach" Q25000        ; create fan 0 on pin 121.out1 and set its frequency
M106 P0 S0 B1 H-1 C"hotend"                  ; set fan 0 value. Thermostatic control is turned off

I get no fan movement (except during the blip time) for any PWM other than 100%, when the fan runs at about 96% of its datasheet max rpm. Reported rpm figure looks credible.

If I drop Q to10kHz get some control over the 0.5 to 1.0 range, but a sudden jump at full power. At lower frequencies I get better control, but the fan squawks (quite badly at 4kHz, very badly at 2kHz).

I get this response (all believing the rpm, but as noted it seems credible - it correlates with wind noise). This is rpm against PWM, for various frequencies:

If I carry on dropping frequences the trend continues - better control, but various degress of squawking. Around 500Hz the squawking is quieter but particularly annoying for some reason. When I get down to 200Hz or lower I think I can hear the fan pulsing at lower duty cycles, and there's a squawk just floating around at the limits of my hearing.

~~I suspect it's just a not-very-good fan (at least for PWM), however~~ [edit: I was wrong] can anyone suggest how to make it controllable but not squawky [yes: see followup].

achrn

(talking to myself again)
This turns out to be a pullup issue. If I add a 4 kOhm pullup between the +ve power line and the PWM wire on the fan, I get proper behaviour - nice linear response to the PWM and no squawking. This is 10kHz PWM with and without an external pullup:

On the oscilloscope it's fairly clear what's going on.
Without pullup:

With pullup:

dc42

@achrn I'm glad you solved it.

RRF works to the Intel PWM fan specification at https://www.intel.com/content/dam/support/us/en/documents/intel-nuc/intel-4wire-pwm-fans-specs.pdf. According to section 2.1.4 the PWM input should he pulled up to 3.3V or 5V inside the fan. It appears that your fan either doesn't have a pullup or it has a very weak one.