Fan Spinning at 0

flo1990

hello, i have some problems with the Fan0.

I've rebuild my printer and before it worked already with an 4 wire Fan. Now i use an sanyo 9gax0412p3s001 fan.

plus and GND connected to an Stepdown and PWM to FAN[n]- and Tacho to PB6

Code:
M950 F0 C"!fan0+^exp.pb6" Q250
M106 P0 S0 L0 X1 B0.1

The Fan is working and controllable but is spinning at 0 already. I checked the Forum and someone sad the fan could be with the wrong label.

Would an 10K pull up help?

dc42

@flo1990 the Intel PWM fan specification, which is what most of these small fans confirm to, allows a choice of behaviour when it is given zero PWM. It may either turn off, or it may spin at up to 30% of full speed. So it's likely that it is behaving correctly for that particular type of fan.

Noctua fans that I have used turn off when given zero PWM.

flo1990

@dc42
do you have some recommandation for an 4028 fan witch is working?

is it possible to switch it off via the gnd like descriped on the rat ric site for klipper? (sry cant post an link)

dc42

@flo1990 I've not used 4028 fans, I've only used 4010 or 4020. The 4028 4-wire fans that I found either didn't specify whether they turned off completely at zero PWM, or specified that they rotated at a low rpm.

Does it need to be 28mm thick? Noctua has 12V 4020 and 24V 4010 PWM fans.

achrn

@flo1990 said in Fan Spinning at 0:

sanyo 9gax0412p3s001

That fan is expecting a 25kHz PWM and you're only giving it 250Hz, might be relevant.

My reading of the data sheet is that it should stop at zero PWM (the closely related 9gax0412p3s003 does not stop). I would try a higher frequency and at the higher frequency you may well need a pullup.

Note that the 'Intel spec' for 4-wire PWM is non-specific on the pullup - it specifies that the fan should have a pullup, but not how strong it needs to be. It seems to me that many fan manufacturers read it as requiring a pullup that's strong enough only to force an unconnected PWM wire high (i.e. very weak indeed). Such a weak pullup is not useful at 25kHz.

Duet seem to interpret the requirement as meaning that the fan should have a pull-up that's strong enough to be relevant at 25kHz, so the duet boards just pull the PWM line to 0V and don't drive it high in between.

Older Noctua fans seem to have a fairly strong pullup and work fine on Duets. I've had several other brands that do not (CUI, Delta) - see e.g. https://forum.duet3d.com/topic/35576/trouble-with-delta-fan. These days I add a pullup to any 4-wire fan - generally somewhere in the vicinity of 4kohm to 1kohm.

dc42

@achrn said in Fan Spinning at 0:

My reading of the data sheet is that it should stop at zero PWM (the closely related 9gax0412p3s003 does not stop).

I agree, it says:

Note: 9GAX0412P3S001，9GAX0412P3K001 does not rotate when PWM duty cycle is 0％ .

@flo1990 are you sure that you have the -S001 variant, not the -S003 variant? Check the label on the fan.

The pullup resistor is irrelevant at zero PWM because that is equivalent to connecting the PWM input to ground (i.e. brown wire to black wire).

flo1990

@achrn thanks for that info. i will try at 25khz and also the pullup resistor

@dc42 the label on the fan is the right one but someone said that rat ric also had an batch with the wrong label