Motor braking power spikes

[Infraviolet]

When you stop a fast moving motor quickly, any motor, brushed DC, brushless-DC/PMSM,stepper... The conversion of mechanical energy to electrical means a big back-emf spike gets produced whilst the rotor comes to a halt.

As the back-emf can never get above the motor's driving voltage, is it a rule of thumb that the maximum voltage seen in such a spike cannot ever get above 2*V_supply? Would this then mean that any risks of electrical damage from ast braking are avoided so long as the components involved in driving the motor, and anything else connected to the supply rail of any h-bridges involved... , are rated to cope with >2*the_voltage_actually_used_as_the_supply ?

Thanks

[Smokey]

You could get the motor velocity up a little higher with some tricks like field weakening or playing with the angle in an IPM. I guess in that case the peak backEMF on decel might be higher than 2x bus.
Can't just use a regen clamp?

[Terry Bites]

Motor back emf and transient spikes have different origins, they are not the same thing.

Transients arrise from stored energy in the motor whereas back emf is caused by induction.
A transient will be limited (or not) by the driver configuration. Eg In a mosfet bridge, transients will be limited by the internal body diodes or external freewheel diodes.
A bipolar driver needs external diodes or other snubber components.

[Siwastaja]

When you stop a fast moving motor quickly, any motor, brushed DC, brushless-DC/PMSM,stepper... The conversion of mechanical energy to electrical means a big back-emf spike gets produced whilst the rotor comes to a halt.

Wat, no? Of course not. Back-EMF voltage is nicely and linearly dependent on the RPM. There is no "spike" of any kind.

[Infraviolet]

"Can't just use a regen clamp?"
I've been looking up some more about this...
Would opening the low side switch of all bridges (both where brushed motors are concerned, 3 for BLDC, all 4 for a stepper) be an adequate solution instead? So long as the maximum current that can be flowing in the motor is not so large as to cause damage to those low side switches and/or damagingly excessive heating in the motor coils.

[pcprogrammer]

"Can't just use a regen clamp?"
I've been looking up some more about this...
Would opening the low side switch of all bridges (both where brushed motors are concerned, 3 for BLDC, all 4 for a stepper) be an adequate solution instead? So long as the maximum current that can be flowing in the motor is not so large as to cause damage to those low side switches and/or damagingly excessive heating in the motor coils.

That is what al lot of full H bridge motor driver IC's do when you apply break mode.

Both inputs zero is free running motor both sides of the bridge in high z.
One input high the motor turns one way, so one bridge high, the other low.
Only the other input high the motor turns the other way, so one bridge low, the other high.
Both inputs high, the motor breaks. Both bridges either low or high.

Handy for dead stopping a servo, not good when PWM'ing. Works bit gives jerky behavior.

[Siwastaja]

If you just don't drive any of the MOSFET gates, then the motor just freewheels and does not regenerate.

... except if it rotates to high enough speed so that BEMF exceeds DC link voltage, so that the body diodes conduct. But that only happens when the RPM is higher than what the DC link voltage is capable of making the motor run at, so this only happens if you power down the circuit, or if you apply external high-speed force to the motor.

[Siwastaja]

Would opening the low side switch of all bridges (both where brushed motors are concerned, 3 for BLDC, all 4 for a stepper) be an adequate solution instead? So long as the maximum current that can be flowing in the motor is not so large as to cause damage to those low side switches and/or damagingly excessive heating in the motor coils.

You avoid excessive current by measuring it and feedback your PWM not to ever exceed current limits. This is symmetrical, there is nothing special in regen. Current increases too far positive -> adjust duty cycle towards 0. Current increases too far negative -> adjust duty cycle towards maximum.

[Zero999]

When you stop a fast moving motor quickly, any motor, brushed DC, brushless-DC/PMSM,stepper... The conversion of mechanical energy to electrical means a big back-emf spike gets produced whilst the rotor comes to a halt.

Wat, no? Of course not. Back-EMF voltage is nicely and linearly dependent on the RPM. There is no "spike" of any kind.

That's not true. There is a voltage spike caused by the back-EMF from the motor's inductance. In the case of a simple transistor and diode driver, the current from the back-EMF will be clamped and circulate through the motor, during the off time. In the case of an H-bridge, it'll be injected into the power supply, via the diodes across the transistors, thus increase the power supply voltage, when the motor is slowing down i.e. braking.