[SOLVED] DC motor Controller Reverse enginering

[Electro nS]

Hi
i am an electronics students and i am doing a project on brushed motor controller ..

i have experimented with some circiuts some with sucess , some with flames unpleasent endings

anyway i am intresed in making a 60A H bridge and most important part is a the protection agianst overvoltage , overcurrent , spikes , sudden braking ......

I am not expecting magical answer and solutions . i just would appreaciate any help , An advise, a schematic , or even you could just post a good quality picture from your motor controller no matter what brand or motor type it is for (except kelly controllers, you know chinease products !! ) , this can help me read the IC names and see the protection (diode , snubber , TVS ..... ) and try to reverse engineer it ( hopefully )
i donot have the money to buy all controllers in the market and open them just to learn . the only information i have now is from OSMC project , It is great but would like to see more designs espically new ones .
thanks in advance guys

 

[BradtheRad]

Motor control via H-bridge often comes up at this board.

There is a list of similar threads at the bottom of this page. Clicking one may lead to others which will help.

 

[goldsmith]

Hi
i am an electronics students and i am doing a project on brushed motor controller ..

i have experimented with some circiuts some with sucess , some with flames unpleasent endings

anyway i am intresed in making a 60A H bridge and most important part is a the protection agianst overvoltage , overcurrent , spikes , sudden braking ......

I am not expecting magical answer and solutions . i just would appreaciate any help , An advise, a schematic , or even you could just post a good quality picture from your motor controller no matter what brand or motor type it is for (except kelly controllers, you know chinease products !! ) , this can help me read the IC names and see the protection (diode , snubber , TVS ..... ) and try to reverse engineer it ( hopefully )
i donot have the money to buy all controllers in the market and open them just to learn . the only information i have now is from OSMC project , It is great but would like to see more designs espically new ones .
thanks in advance guys

Hi Electro nS
A question : do you need to change the direction of your motor ? if yes as Bradtherad said before an H bridge would be the best solution .
Designing such a controller is so simple . but if you tell me where is your exact problem , i'll be able to help you as well .
Do you know about theorem of controlling ? or are you familiar with power switches ?

Best Wishes
Goldsmith

 

[Electro nS]

thanks for the reply guys .

offcourse it is an H bridge . I just want to know how can i make the bridge more robust . I have taken a look at OSMC project , it looks fine and have tried it with some sucess , but the transients "voltage" often kill the bridge after a very short time ( before even making the fets hot) , because of sudden stop (the motors have built in EM brake and i found it very useful)

how can i prevent that kind of spikes that is caused by the inductive load ?

i have heared about crowbar circuit , and RCD snubbers , but i donot know how to design such circuit and where to place it , over the motor leads ?? or across battery ?
for your information i am using a sign magnitude drive technique . is it better to use locked antiphase or another technique ??

 

[BradtheRad]

The system is probably okay at full speed, while 60A go through the motor continually. Each winding gets commutated into and out of the power loop in turn... but there's no spikes because the flux field is not asked to collapse suddenly.

But then try to stop it suddenly... that's when the spikes are generated. The flux field is asked to collapse suddenly, creating hundreds or thousands of volts at the coil ends, and causing heavy current to be pulled through switching devices.

Consider the moment when the switching devices are turned off suddenly. It would help if you were to connect a short across the motor terminals at that moment. Spikes would be greatly reduced. However it may not be possible to arrange components that would do this in both directions (since you want both forward and reverse motor spin).

Your project has similarities to a wheelchair motor controller. Some of the challenges are discussed in the threads below.

https://www.edaboard.com/threads/262762/

https://www.edaboard.com/threads/250866/

 

[chuckey]

There will be a small transient every time a brush goes across commutator slot. At 60A that a load of power and its difficult to find a snubber that is low enough impedance to soak up any power. You do not say what your voltage is. I presumes its 24 V? Have you fitted the four diodes between the motor terminals and the power supplies? make sure that the power supply ends of the diodes are interconnected with really good capacitors so the return current from the motor can return to the power supply. I would think 1 MF polycarbonates (bit expensive) might do. this should clamp both ends of the motor to the power lines. A small capacitor can be fitted (.01 MF) from the motor terminals directly to its case and earth, just to decouple any spurious transient voltages.
Frank

 

[FvM]

Besides clamping overvoltage, overcurrent shutdown seems urgently required considering the funny "sudden braking" problem. Although I imagine that the motor controller could be stopped simultaneously by suitably designed logic, the driver should be able to handle any kind of stall. Usually a shunt or current sensor is required.

There's a certain relation between overcurrent and overvoltage, because overvoltage will increase with load current in PWM operation. In case of an overcurrent event, it might be hard to decide if the output transistors have been killed by overvoltage or overcurrent. But considering the avalanche capability of modern MOSFETs, I would expect overcurrent induced "desaturation" (transistor operation in linear mode with high power dissipation) as the more likely mechanism.

 

[Electro nS]

There will be a small transient every time a brush goes across commutator slot. At 60A that a load of power and its difficult to find a snubber that is low enough impedance to soak up any power. You do not say what your voltage is. I presumes its 24 V? Have you fitted the four diodes between the motor terminals and the power supplies? make sure that the power supply ends of the diodes are interconnected with really good capacitors so the return current from the motor can return to the power supply. I would think 1 MF polycarbonates (bit expensive) might do. this should clamp both ends of the motor to the power lines. A small capacitor can be fitted (.01 MF) from the motor terminals directly to its case and earth, just to decouple any spurious transient voltages.
Frank

Yes the voltage is 24v , regarding the diodes (if you mean freewheeling diodes across each mosfet ) i didn't use any and relied on the internal diodes .
regarding the capacitors i didnot get the point , first what is MF (milli farad ?? or MEGA !!! ). second how are these connected ? from VCC to GND or along with the diodes , can you clarify , with a scetch if possible .
thanks

 

[chuckey]

Microfarad. If the motor starts to generate back emf then one end of it will go + and the other -, so you want this voltage to be conducted away to charge up the PSU capacitors. The problem is that of the wiring inductance, so for the first microsecond or so you get a voltage overswing at the PSU end of the diodes that are conducting, so ideally you want two capacitors connected diagonally across from the K of the +PSU diode to the A of the - PSU diode, so there is a circuit foor the back emf, running motor -> diode -> new cap -> diode -> other motor terminal, and the same duplicted for the emf when the motor is running in the other direction. Of course if your PSU leads are really thick and short to the reservoir capacitor, then just one cap across it will do. Electrolytics do have a series inductance, which the new cap will decouple.
You might be over cooking the internal diodes, on forward conduction, or if there is too much lead inductance, over volting them? have you tried to run your kit at half motor current for a prolonged period (2 hrs)? just to see if it gets too warm, which would just be accelerated by constant speed change. If it survives this test, then increase the speed to max for an hour, if still OK, try to constantly alter the speed for say 10 Minutes. So by now you will have proved that it must be the change of direction that the problem and not some lurking over dissipation problem, which is some sort of start.
Frank
Frank