I built something like this for my cnc mill and may do the same for my printer:a set of relays on the lines between the drivers and the stepper motor. (by relays I mean fets, because actual relays are big and expensive.) That way when the power is off the drivers are disconnected from the steppers, and I can run things by hand without cooking driver hardware.
Do you use matched nFET/pFET pairs to allow current in both directions?
It’s a design I got from a power directing switch matrix at work, and yeah it uses a pair of fets. I made it years ago and don’t remember the schematic off the top of my head.
The only difficulty is finding a negative voltage to feed to the gate on the pFET. I think that most 3d printing boards don’t generate one. Did you have a negative supply for the version on your mill? Or did you generate it on the board?
I have a negative voltage available… I’ll go dig up that schematic.
Yeah, since most boards don’t have a negative voltage, generating it is probably sensible for 3d printers. I don’t even remember what oscillator I used the last time I had to do this…
[I moved this thread to a new topic so we’re not hijacking @HalfNormal’s new printer thread… ]
Since it is common to move 3d printer heads around with the printer turned on but the steppers disabled, I’d think that on a printer you would want to use
~ENA as the signal, not just connecting the motors to the drivers only when power is off.
Feels like this could be an easy simple interposer board. Might as well make it five or six channels since the parts per channel are probably pretty cheap. Power nFETs and pFETs that would work here are about 30¢ in onesies and you would need 4/motor channel. If you need to invert
~ENA hex inverters are common and would support six channels. A dual schmitt trigger as the oscillator to drive the negative voltage supply. Some small signal nFETs as gate drivers for the power pFETs.
Of course, my solution has been to move the steppers slowly.
I’m sorry to be so belated: I’ve had a lot going on. What I used was a bidirectional power switch based on this paper:
He goes through which different topologies have advantages in different situations.