For the hardware folks - FET driving for BLDC motor

[InvaderZim]

Hi all, this is only vaguely related to a PICAXE. But I'd like to have a logic circuit drive a brushless dc (BLDC) motor, which means having FETs on 24VDC switching on and off, but with logic at a different voltage. This means level translation, which I've got no experience with.

Would the attached picture be a good start? I'm really worried about transients affecting the driving logic; don't want those FETs on at the same time! There will also be hall effect sensors in there somewhere, feeding the logic.

Also, here's another catch: the 24VDC for the motor may not be clean. In fact, it might just be rectified AC, peaking at 40+ Volts and diving down to 0 (or less...pesky motors). In the future, there may be PWM, but not yet (baby steps!) The driving logic will be fairly clean though. In theory that isn't a problem, but motor transients have me all worried.

Max current is about 1.5 amps; average current is much less.

I'm sure there's been things like this done before, but I'm having trouble finding good examples; they're all 15V max, or rely on a level translation or motor drive chip that is obsolete. I'm looking for a really generic example that I could use again in the future or teach with.

Thanks in advance!

 

[jglenn]

This is an extremely difficult task. There is a series of articles in recent
Servo Magazine on a scratchbuilt one. The software is mind boggling.

I fly Eplanes, can get cheap ESC electronic speed controls, but just had
to try to built one, 3 phase bridge with fets, and a raw pic chip, prog in assem.

Only worked at one speed, and the motor got hot. Later I found I did not
address what they call "timing". Unless you have a sensor, usually hall effect,
to tell where the rotor is, you need to do it by measuring back EMF. Oops.

 

[InvaderZim]

Just to clarify, the BLDC timing won't be controlled by a PICAXE; I was going to use the hall sensors and a couple discrete logic gates to handle all that timing. That way my AXE can do more fun things!

What I'm uncertain about is the eternal problem of how to drive those FETs safely with a logic circuit.

In fact I'm starting to question the wisdom of driving the low one directly; ground bounce there could cause a misfire. Ugh! See? I'm starting to chase my tail here.