Motor controller design?

[czwalga]

What I have:

Brushed 10 motor, max current 8amps. (2 pin, pos/neg)
Motors transfer function.
Potentiometer (0v 0degrees, 5v 90degees)



I need to build or subcontract out a motor controller to these specifications. The motor will only turn from 0 to 90 degrees and the position I would like to be controlled via a PWM signal because I already have my FPGA built and do not want to change it.

Whats my best route to take on this? Any help would be appreciated.

 

[Papabravo]

I think we are going to need a bit more information.

1. What is the voltage of the motor supply?
2. If the motor goes too far in one direction what do I need to do to reverse it?
3. Where will the motor be with duty cycles of 0%,50%, and 100%?

How many of them do you need?

 

[czwalga]

The motor is 12v. The motor is built into a housing where it will not let it turn passed 90 degree's or less than 0 degrees. It is also conneced to a very strong spring so if you open it all the way, then cut the power it will go back to 0 degrees.

Where the motor will be at each duty cycle? I'm not quite sure what you mean. My FPGA can output any settings (duty cycle, pulse width) I want, I just have to program the chip according to what the motor controller settings call for.


Simply... I have an adjustable PWM signal, a 12v motor, a poteniometer that varies from 0v (0 degees) to 5v (90 degrees). I just need to a controller to beable to vary the motors position.

I only need one for now, however i could need many in the future; just kinda going for a working prototype, everything is completed but the motor controller.

 

[Papabravo]

So if I apply 12V across the motor it will go to the 90 degree point, run up against the hard stop and sit there drawing 8 amps, or more and dissipating a bit less than 100 Watts. If I drop the current a bit the torque will be reduced and the spring will pull the motor back toward zero.

The spring helps you move in one direction, but you are fighting it in the other direction. Do I have this correct?

 

[czwalga]

So if I apply 12V across the motor it will go to the 90 degree point, run up against the hard stop and sit there drawing 8 amps, or more and dissipating a bit less than 100 Watts. If I drop the current a bit the torque will be reduced and the spring will pull the motor back toward zero.

The spring helps you move in one direction, but you are fighting it in the other direction. Do I have this correct?

Yep, thats correct. There really no need to be a reverse voltage to drive the motor backwards althought it could be done if needed.

 

[Papabravo]

So for starters we could have a comparator such as the LMV331
https://www.electro-tech-online.com/custompdfs/2007/06/LMV331.pdf
running off of 12V for Vcc comparing the 0-5V signal from the pot with a 0-5V command signal. The open collector output would indicate that the drive current was either on or off. I can't tell if hysteresis would be a good or a bad thing in this case because I don't have any feel for the spring action versus the torque produced by the motor.

I guess you could generate the 0-5V command signal from a PWM into an RC filter with a corner at 0.1 Hz.

I guess a 20V FET with a current rating of 12 to 15 Amps and a gate drive that would take the comparator output and drive the gate of the FET and you've got something to experiment with.

So what's it worth to ya'?

 

[czwalga]

Haha, could be worth a fair amount if it works. I'm trying to see how well a comparator and a speed control chip would work out together.

 

[Hero999]

Can't you just buy a servo?

 

[czwalga]

Can't you just buy a servo?

The original prototype was built with a servo. It's how I programmed my FPGA, however it did not have enough torque I needed.

 

[Papabravo]

The more I think about the Bang-Bang control that I suggested the less I like it. It is quit probable that it would limit cycle and depending on the strength of the spring cause great stresses on the motor and the control circuit. In a normal servo system there is an inertial deadband around the the setpoint such that if the system does not move the control voltage will allow the velocity to come to zero or near zero without rocking back and forth which I think your system will.

I think you need to look for a different solution.