Free backward motor rotation stopped by snubber diode. Workaround?

[pirvcb]

 Hello there. I'm looking for help to solve this problem. A PWM controlled motor, normally works in only one direction. When i stop PWM driving i can't move it in reverse rotation because it's stopped by the Snubber diode. So, the motor, acts as a generator and the diode prevent it to rotate in the other way. I'm asking some solutions or tip to solve this problem. I need move the motor, NOT DRIVE THE MOTOR powering it, in the opposite rotation way. Thanks in advance for any help.

[Checksum8]

Can you post a schematic of the output, showing the transistor/mosfet, snubber diode and motor connections? Otherwise it is impossible for anyone to help you.

[pirvcb]

Yes, sorry. Nothing much more this. A brush motor permanent magnet 200W at 12V. The diode is a 30A schotty and in to-220 becomes warm but ok. The transistor is a N mosfet at around 12Khz of pwm drive. My need is freely rotate backward the motor when it is not powered.

[Checksum8]

All I can think of is disconnecting one side of the motor, using a relay, switch or second mosfet. Is this a vehicle or something like a machine spindle?

[pirvcb]

Are baby cars and them needs to be parked pushing in back direction. The only solution that i thinked was instead of diode put a zener to admit a more high inverse voltage but in this way i get down the diode efficacy. And yes, a relay that disconnect  the diode seems the only solution. For this reason i asked in the forum


Posted on: March 16, 2018, 06:46:04 18:46 - Automerged
But i can't see no way to do it with a semiconductor instead a relay

[pickit2]

Can you not move the diode to across the driver.

[pirvcb]

Can you not move the diode to across the driver.

Yes, i can, but this mean protect only the driver leaving unprotect the other side of the circuitry.

[bigtoy]

How about changing out the snubber diode for something else? One idea is back-to-back zener diodes (or a bidirectional transzorb).

[pirvcb]

Could you explain me some more details about your solution? I'm sorry but i can't understand. Anyway i go to google about your "keywords"... I mean... ok for back to back zeners, but it's hard find 30A zeners, or maybe i'm wrong? And, bidirectional tranzorb: what is this? I see i see, should be a good workaround, thank for the input. I'll try in this way. Thanks again.

[bigtoy]

Sure. To clarify, due to its inductance, the motor can generate voltages substantially above its rated voltage. For example a 12V motor can generate spikes much greater than 12V. These spikes can (or will) damage any attached circuitry (like MOSFETs or other transistors) which are attached to the motor. The purpose of the snubber diode is to "snub" those spikes. The diode is simple, cheap & effective, but only works in one direction. For something that works in either direction, a device like a bidirectional transzorb would work. For example, with a 12V motor and a 12V power supply, selecting a bidirectional transzorb with a 15V stand-off voltage means that spikes which go much above 15V will be clamped. But our example transzorb won't conduct at 12V, so it won't cause problems during normal operation.

Note there are unidirectional transzorbs (much like a zener diode) and bidirectional transzorbs (much like two back-to-back zeners) so make sure you choose a bidirectional part, with a standoff voltage suitable to withstand your normal motor operating voltage.

[pirvcb]

Sure. To clarify, due to its inductance, the motor can generate voltages substantially above its rated voltage. For example a 12V motor can generate spikes much greater than 12V. These spikes can (or will) damage any attached circuitry (like MOSFETs or other transistors) which are attached to the motor. The purpose of the snubber diode is to "snub" those spikes. The diode is simple, cheap & effective, but only works in one direction. For something that works in either direction, a device like a bidirectional transzorb would work. For example, with a 12V motor and a 12V power supply, selecting a bidirectional transzorb with a 15V stand-off voltage means that spikes which go much above 15V will be clamped. But our example transzorb won't conduct at 12V, so it won't cause problems during normal operation.

Note there are unidirectional transzorbs (much like a zener diode) and bidirectional transzorbs (much like two back-to-back zeners) so make sure you choose a bidirectional part, with a standoff voltage suitable to withstand your normal motor operating voltage.

Yes. This is exactly what i realized after your value input. Again a big thank you!

[solutions]

Why don't you drive the motor with a bridge?

That way you get regen for free (as much or little as you want) and can freewheel whenever you shut all legs of the bridge off....

[pirvcb]

Why don't you drive the motor with a bridge?

That way you get regen for free (as much or little as you want) and can freewheel whenever you shut all legs of the bridge off....

I'm still evalutating if i can use a chinese pre made module for PWM. Because i'm not so expert in motor drive issues, i would like skip to go in deep in this. Anyway i'm at first look. So i have to establish if PWM frequency of that module will be ok for my motors. I read somewhere that an enough good method to investigate this parameter is put the motor with a medium load at half rotation speed and then change the PWM frequency locking for the minor current.
These premade modules are configurated as previous said; no half bridge, just aparallel of mosfets some filter on gates and a micro that vary the impulse wide at changing voltage at its adc input. Advanteges? No made PCB nor cost for protype (i will make only 10 pcs). Fast substitution due its modular concept. Very low cost; the mosfet price is equivalent to the whole module. Still thinking about and waiting for some tools to determine the motors specific as i told before. I see that your SOLUTION is much more elegant. And also, zeners and transil, seems not so good for this issue: i sohuld prefere, after all, to have a relay that switch on off the Schotty diode. Anyway thanks for your suggestions!

[pirvcb]

Reply to myself. Added an Nmosfet that can be polarized by a 40 eurocents chinese stepup to let the motor a free reverse movement. Also added a Pmosfet PWM drived to brake.

[pirvcb]

Sure. To clarify, due to its inductance, the motor can generate voltages substantially above its rated voltage. For example a 12V motor can generate spikes much greater than 12V. These spikes can (or will) damage any attached circuitry (like MOSFETs or other transistors) which are attached to the motor. The purpose of the snubber diode is to "snub" those spikes. The diode is simple, cheap & effective, but only works in one direction. For something that works in either direction, a device like a bidirectional transzorb would work. For example, with a 12V motor and a 12V power supply, selecting a bidirectional transzorb with a 15V stand-off voltage means that spikes which go much above 15V will be clamped. But our example transzorb won't conduct at 12V, so it won't cause problems during normal operation.

Note there are unidirectional transzorbs (much like a zener diode) and bidirectional transzorbs (much like two back-to-back zeners) so make sure you choose a bidirectional part, with a standoff voltage suitable to withstand your normal motor operating voltage.

Just to give correct infos, actually the snubber diode or freewheel diode, can't be swapped by a tvs. As in H bridge, during inactive state, you have to close both transistor on the same level (up or bottom) to let the current countinue flowing. This also clamps spikes but first of all provide a good working cycle for motor: I sperimented as keeping off snubber diode in a chopper configuration increase noise and not least, slows down motor RPM with the same current rate.