I remember I disassembled a brushed motor years ago to clean it. I also had plenty of brushless fans die on me on GPUs when I mined E T H c r y p t o, some even had rotor completely pop off. In my first job I also worked with MCUs and Hall Effect sensors in the past to measure high currents (inrush and geomagnetically-induced, up to thousands of Amps) in conductors in a contactless way - was a product for electric power utilities for power transformers. Fun times.

Hi there! So the Phanteks T30 is a 3 phase, 12 pole fan? My god, that accuracy at 3000rpms is ridiculous! Thank you for testing that. Have you heard of the Super Flower Megacool? I wonder who the OEM is. Also 3 phase, 12 poles, 30mm thickness, and made with LCP just like the T30s but with a different fan blade design. I didn't like them though due to the slight motor hum, but great otherwise. The T30s don't have any hum from my experience.

Thanks for another very informative video. I've had a question about computer cooling that I haven't been able to find an answer to for years. Basically, why don't computers ever make use of the Bernoulli principle for cooling? Wouldnt it multiply airflow?

12 coils will not result in 12 magnetic poles. It will most likely be 12 Slots and 8 poles so four north and four south poles. With a magnetic film you should easily be able to view it. The hall sensor is really not needed for accurate speed sensing since the motor frequency is directly proportional to the speed since it's a synchronous motor. Speed accuracy is really a software and control thing, not dependent on the motor technology itself. Single or three phase does not make a difference at all. Most of the times the PWM input will just result to a PWM of the motor, as the input voltage changes the speed changes as well. A closed loop speed control so like one specific PWM duty cycle results in one specific speed is very rare.

Had to replace the hall effect sensor(SMD) on a fan I somehow accidentally killed, symptoms; not spinning (obviously). Seems to also trigger stall protection; so it cogs and locks the rotor and so little wobble, and Im assuming it pulsing the N+P Fets to get the thing started but fails and goes into fail state then tries again. So did some measuring around the with multimeter, not ideal but works, suspected the hall effect randomly, but saw the pulses to Nfets from 0v to +12v and +12v to 0v on the pfets, the fets where good, also looked as the drive signals from the controller, looked fine, aka no shorts and good voltages. Then measured the voltage to the hall effect 4pin ones, had around 0.6v, removed the hall effect voltage was now 2v, so I'm thinking internally shorted stole a 4pin hall effect sensor for 3phase optical disk drive, very similar drivers btw would be cool if had look at them, how to maybe to get them spinning reliably...etc anyway swapped the hall sensor measuring 1v on the same pins, and put the fan together again and spun up perfectly! side notes; found the Delta fans pretty easy to disassemble, no glue or anything just some small retention ring(not the the one on the fan shaft) holding on bearing housing the stator and stator legs soldered to+ holding the PCB if that makes sense, and you twist back and forth gently and firmly, while pulling up away from the fan shroud but being careful not to pinch the wires too badly, the ones I dissembled only 1 had glue which you break by rotating force which was 120mm delta but still plenty friction and it is keyed and 92/90mm which had no glue was very strong friction fit anyway sorry the rammably comment XD

I keep wondering about one thing though. Can you replace the controller entirely and use a brushless motor controller that RC vehicles use? And if you can indeed do this, how much can you push their speed over the rated original?

I'm curious. Three-phase motors should not need hall-effect sensors to control them. You vary speed by changing the frequency of the pulses and change direction by switching phases. A hall effect sensor would only be needed if you want speed measurement and you'd only need one, not three. If the motor has three hall-effect sensors, then the circuit is probably three single-phase motor controllers designed to operate in tandem. Is this correct?