Protection of H-bridge Motor Driver against induced voltage without supply present

[M_Schmidt]

Hello ST Community,

I am using a VNHD7008AY ICs to control a brushed motor (XDW043003-01, LowSide-Mosfets: SIR606DP-T1-GE3CT)

The PCB is supplied with 24VDC. The application has the problem, that the motor can be moved by the user both in and against the turning direction.

To protect against short overvoltage cases, where the user moves the motor in turning direction, I am using a zener diode (BZG04-22-M) connected to Motor_VCC and Ground. With Software (simple PID Controller) I can detect the user moving the motor faster than expected and reversing the polarity of the motor, which slows down the movement until the actual position is again equal to the expected position. Therefor the application has no problem with overvoltages during normal operation.

The problem arises, when the PCB does not have a supply voltage. When the motor is moved, a voltage is induced due to the nature of brushed DC-Motors through the body diodes of NMOS and PMOS from Motor_VCC to Ground.

Motor characteristic values:

• short circuit current when connecting OutA to OutB of the motor: <10A
• maximum induced voltage when OutA and OutB are open Line/not connected to anything: <50V

I am able to destroy the H-bridge Motor driver by moving it with force due to the voltage exceeding VCCmax of 38V. The zener diode junction is not rated for the amount of power to be dissipated that would be necessary for the motor and it also is destroyed.

I am looking for a solution how to dissipate the power or block the induced voltage if there is no supply voltage on the PCB. Is there some easy way to protect the Motor Driver against this error case? Or is the Motor Driver not suitable for the motor?

If you need any more information, I am glad to provide it!

[Peter BENSCH · Accepted solution]

Welcome @M_Schmidt, to the community!

A similar problem was discussed here, which essentially has the same causes, although it is a stepper motor there and a brushed motor in your case. As you have already recognised yourself, the small BZG04-22-M cannot absorb the generated energy, because the max. 3W power loss, if on a heat sink with <25K/W, is guaranteed to be too little with your powerful motor.

Either you charge any existing battery with the reverse energy, or you convert it into heat in a suitable way.

Hope that helps?

Regards
/Peter

[M_Schmidt]

Hi Peter,

Thank you for your quick answer.

In my case the problem is willful mishandling by the user. I need to convert the generated power into energy and ideally short MotorA to MotorB to add braking force to the motor. I could implement this by using a relais, another possible solution is a "self-shorting" NMOS by tying the gate to the drain over a pull both up to VCC. By connecting the gate over a logic mosfet to ground I can disable the self shorting mosfet. (See sketch in image). 

I will test out the design and post if it works while still being fairly energy and cost efficient compared to a relais.

My thought process was, that maybe there was a better solution that just might not be documented in the reference manual/data sheet of the motor driver IC that I couldn't think of.

Many thanks 

Markus

[Peter BENSCH]

Well, that could be an idea in theory, but it has a few problems:

• if you switch on the supply but forget to activate the Enable NMOS, you immediately short-circuit the supply voltage
• however, the short circuit is not a hard short circuit, but a limitation to approximately the threshold voltage Vgs of the right NMOS
• this limitation puts the NMOS into linear mode, for which it must be designed both technologically (e.g. as a planar type that is hardly ever produced) and thermally, i.e. it must have sufficient cooling

Another approach could be to simply short-circuit the brushed DC via the NC contact of a relay, but you have to take into account the response and release time of the relay when switching on and off.

Regards
/Peter

[M_Schmidt]

Hello Peter,

once again, thank you very much!
I designed a new PCB with a IRLR3410TR as Power Mosfet for shorting the current. Its not a cheap but a very robust solution for an acceptable time of the motor being moved externally (thermal limitation after >>1 minute, which is acceptable in my case). 
The solution with the relais has the disadvantage of being energy inefficient because in the application the PCB will be on in most times. The approximation of the threshhold voltage is also ideal, because with a full short circuit the motor would be slowed heavily, which is not fitting for the application.

Thank you for your tip with using a planar type mosfet, I was not aware of those being ideal for the application.

Regards

Markus