Speed Control according to DT0001

[r239955_stm1_stmicro]

Posted on March 11, 2015 at 10:54

Hi,

the application note DT0001 suggests controlling the motor speed by using the FWD/REV pin with a duty cycle. Though this sounds very promising especially for feed-back controller design I was wondering what happens if the motor is running in one direction and I force the motor to drive the other way. Isn't the motor than used in generator mode acting as a voltage source?

Has anyone had experience with this usage and how is the L6235 reacting to it - where is the power burned?

Thanks,

Rasmus

#brushless-speed-control #l6235 #dt0001

[mikemike9141]

Posted on March 13, 2015 at 12:15

Hi Rasmus,

in your scenario, the inertial energy stored in the spinning rotor will be ''dumped'' in the inverter's DC bus capacitance, raising the bus voltage. In other words, you are right, the kinetic energy is transformed into electrical energy by the inverter (any three-phase BLDC controller has this property inherently). If that one is large enough and there are no other protection provisions, you will see nice green fire coming from the power stage (from the vaporized copper :)

Fortunately, you have several options:

- add a brake chopper, a resistor and a control loop that limits bus voltage (already done in ST's FOC library)

- connect a battery in parallel, which then will be charged instead (no rectifiers please :)

- add a strong TVS on the DC bus, but make sure the maximum kinetic energy never exceeds its rating

- increase capacitance, reducing the amplitude of the voltage peak

- use power MOSFETs with UIS (unclamped inductive switching) rating. These do accept repeated overvoltage-induced Avalance breakdown, making the transistor work like a TVS. The caveat although is the chip's relavitely small thermal capacitance, allowing an ''instantaneous'' energy input of about one Joule or so for ''normal'' sized transistors, until energy starts being transferred to the heat sink.

- use dissipative braking mechanisms. Hobbyist's BLDC controllers achieve this by interrupting normal PWM operation, and shorting the motor leads to ground for braking by enabling all three low side's (also PWM'ing this in order to control braking torque). This way the energy is dissipated in the motor's winding and the power stage's ohmic resistances.

The last mechanism would be a *very* nice and *very* appreciated feature for the ST FOC library...

Regards

Frank

[r239955_stm1_stmicro]

Posted on March 30, 2015 at 10:25

Hi Frank,

thank you for your thorough answer on this topic.

> - add a brake chopper, a resistor and a control loop that limits bus voltage (already

> done in ST's FOC library)

I assume that the FOC library works in a very general way and not with a motor controller like the L6235. However, I think that using a brake chopper seems to be the solution to my problem - thanks for the advise.

> - add a strong TVS on the DC bus, but make sure the maximum kinetic energy never

> exceeds its rating

I would still raise the DC voltage behind the TVS in this case. And of course the energy would still need to go somewhere. So this seems to me like a risky solution - or I would need to add quite a few capacitors into the circuit to store the energy.

> - use power MOSFETs with UIS (unclamped inductive switching) rating. These do

> accept repeated overvoltage-induced Avalance breakdown, making the transistor work

> like a TVS. The caveat although is the chip's relavitely small thermal capacitance,

> allowing an ''instantaneous'' energy input of about one Joule or so for ''normal'' sized

> transistors, until energy starts being transferred to the heat sink.

I am not using MOSFETs directly instead I am using the L6235 which has these included. However, I did not find any information on the MOSFETs in the data sheet of the L6235. Since ST has an application note DT0001 which suggests driving the motor like I have described I am assuming that the MOSFETs have UIS - however I am not sure.

> - use dissipative braking mechanisms. Hobbyist's BLDC controllers achieve this by

> interrupting normal PWM operation, and shorting the motor leads to ground for

> braking by enabling all three low side's (also PWM'ing this in order to control braking

> torque). This way the energy is dissipated in the motor's winding and the power stage's

> ohmic resistances.

I would rather not like to do this, since I am using the motor in a feed-back loop control. Using a brake resistor is included in the L6235, however this always kills the feed-back controller since you are changing the system model.

Of course if I would be able to control the braking torque with PWMs this would change, but the L6235 does not support such feature.

So putting it all together my remaining question would be: On the L6235 are the MOSFETs able to handle the current coming back from the motor. If that is the case I assume your first solution - brake chopper - would be the way to go.

Thanks,

Rasmus

[Gigi]

Posted on April 02, 2015 at 08:51

Ciao

> So putting it all together my remaining question would be: On the L6235 are the MOSFETs able to handle the current coming back from the motor. If that is the case I assume your first solution - brake chopper - would be the way to go.

I suggest to ask

https://my.st.com/public/STe2ecommunities/motordriver_ics/Lists/Motor%20Driver%20ICs%20and%20Boards/AllItems.aspx

for question regarding ICs.

Ciao

Gigi

[andrej]

Posted on June 08, 2015 at 18:07

Hello,

reading this case I have to admit, that I am totally confused: what is now with DT0001 - can it be used as it is written and also mentioned in AN1625, page 33?

Thank you for your answer,

Andrej