Help me not burn up a $100 DC PWM fan

[LCA078]

I have a large'ish 4-wire DC  cooling fan being delivered early next week for an air flow test bench.  Fan spec sheet is attached.  I'll be supplying 48V with a good lab grade power supply so power is not an issue.

My question is understanding the PWM signal schematics.  I plan to use a cheap Amazon PWM square wave generator to signal the motor and vary the rpm but I'm having trouble understanding the limitations of the PWM circuit.  Screenshots from the fan spec sheet are thumbnailed.

Here's the PWM generator: https://www.amazon.com/Icstation-Frequency-1Hz-150kHz-Adjustable-Rectangular/dp/B0814MKZVW/ref=sr_1_16?crid=2QCC19NMQHK4R&keywords=pwm%2Bgenerator&qid=1670106124&sprefix=pwm%2Bg%2Caps%2C100&sr=8-16&th=1
No spec sheets on the generator...just the link I posted.

I'll keep the Vin to the PWM generator under 20v as the PWM generator Vout = Vin (will probably stay around 12v) but I'm confused on the Vfg and Vce in the first screen shot.  I'll hook I can't believe a cheap PWM generator would fry the fan's IC but I'm just being safe by posting here for verification.  Any worries about supplying ~12v to the PWM generator and using it hooked straight to the fan without protection?

[Marco]

The blue wire is RPM feedback, it just needs a pull up resistor and a reference voltage which if Vfg. You measure below the pull up resistor, the voltage goes from Vce to Vfg.

The PWM generator seems fine from a functional point of view ... but the fact that their image explaining the connector is exactly reversed from the sticker they put on it is a bit concerning.

[moffy]

The Vcesat is just the transistor saturation voltage or the minimum output voltage i.e. 0.5v but that might be a little high for the 0.4V low voltage of the fan. Try the PWM generator first but you only need a 5V supply for it. If that doesn't work use a 5V CMOS buffer like a 74HC14 to condition the signal it will provide a 0.1V to 4.9V output for a 5V supply. Just drive it with a suitable 0-5V signal i.e. use 5V for the PWM generator to drive the 74HC14.

[wraper]

This is not a PWM input cirquit. It's an outpout for speed feedback. Vpg is basically a VDD of a cirquit that will accept this signal. PWM cirquit and specs are on page 7 of the datasheet. PWM voltage must not exceed 20V. But safest would be running it from 5V or 3.3V.

[LCA078]

Ahhh- I missed the rpm/tach output signal.    But that's why I'm posting here...

Okay, this is straightforward then.  I agree about staying low Vin for the pulse generator.  Will let everyone know if I have issues when all parts arrive.

Thanks again.

[NiHaoMike]

Add to that proper supply sequencing, that is only turn on the PWM signal after turning on the 48V supply. That said, I'm pretty sure it's protected against incorrect sequencing.

[LCA078]

I'm sure it's protected against incorrect sequencing since the spec states that if you float the PWM input wire, the motor runs at full speed.   I would assume if there was a hint of sequencing dependence, the motor would not run without a robust signal.  It should be a brick if the input signal was grounded or floating.  Probably even shouldn't run if the input was a steady voltage instead of a true PWM...but who knows how that driver was designed with safety features.  But yes, thanks for the reminder about proper sequencing.

[Geoff-AU]

The voltage spec on the pin will tell you if sequencing matters.  If it's something like -0.3V to Vdd+0.3V then in the case of no voltage rail present, don't take the input above 0.3V.

The PWM input schematic on page 7 shows it's just an NMOS gate with some high value resistors around it, and the voltage rating at the top of the page is -0.8V to 20V.  Doesn't say anything about Vdd, so sequencing doesn't matter.