DC Motor Control

[Svart]

The pulsing problem is for much lower switching frequencies, the parasitic resistance period is proportional to the dv/dt and the switching freq, so the longer the rise time vs the frequency, the worse your heat will be. I'll get into that later. The motor regen period is proportional to the number of brushes and the degrees apart the commutator contacts are. Add all these together and factor in the mass of the motor assembly's rotor and your motor could very well move smoothly but yet be very inefficient during the regeneration periods causing high current draws during coil charging. It's during this period that the current rating of the FET is most important. Because you are running such a small motor I'm not sure that a lower RDSon will help here though.

Forgive me if I am rehashing anything you already know, I'm going to dig a little deeper into the theory for the folks watching at home. Maybe they can learn something they didn't know.

I^2*R=P

is to figure out your power using current and resistance, so saying that you are pulling 2 amps at full draw and your RDSon is .5 ohms, your max power is 2 watts at most. A lot of SMPS FETs are now down into the .05R or even .005R range! Remember that these RDSon ratings will change with junction (Tj) temperature as well, usually rising with temperature semi-logarithmically. Remember that this is all rated for a fully ON device. When a MOSFET is partially ON it's RDS can be MUCH higher depending on the gate charge. It could be so high that your FET will heat up to the point of failure. It's not quite a logarithmic ratio but it rises exponentially with lower gate charges. So if the gate is slewing slowly enough to allow the FET to be partially ON but still within a variable resistance state (transconductance) then you could have a big resistor for substantial periods of time.

But I digress.

I'd say that since you have what you want out of it, just increase the size of your heatsink and you should be OK, die size of the FET isn't going to factor into the efficiency equation here, it would only change your gross power handling capability. It may not be an elegant design but if it works, it works.

But as a rule of thumb, you want that gate to turn on/off as fast as you can possibly slew it. And by that I mean fully ON, fully OFF. That's no easy task at all.

And congratulations, a lot of folks don't even get as far as you have.

I'd still like to see what solution you chose. If you get a chance to draw it up and post it, please do.

 

[BYacey]

I've experimented with various ways of powering a coil winding machine, and I think if i was to do it again I would use a stepper motor. The problem with a series wound motor is the current usage is very high at startup, and at slow speeds it doesn't develop much torque. If you can find a shunt wound motor, you apply a constant current through the field, and vary the voltage on the armature from a variable or PWM supply to control your speed. This will develop large amounts of torque at slow speeds ideal for coil winding.

 

[adamasd]

-Forgive me if I am rehashing anything you already know

Rehash away!, I am always up for refreshers, not to mention there are alot of odd little gaps in my knowledge.

-I'd say that since you have what you want out of it, just increase the size of your heatsink and you should be OK, die size of the FET isn't going to factor into the efficiency equation here, it would only change your gross power handling capability. It may not be an elegant design but if it works, it works.

Well I am far from your 10times current load recommendation that you mentioned, which is my main reason for going to a sturdier transistor. I am pretty much flying on best guesses here. I spend most of my time in tubes and audio, so DC and motors is just weird for me. I am sure my design is far from elegant, but I never tried for elegant, I just tried for working, and I got it!

-And congratulations, a lot of folks don't even get as far as you have.

Thanks, it was a fun project and nice to leave familiar grounds for awhile and do something new.

-I'd still like to see what solution you chose. If you get a chance to draw it up and post it, please do.

After I get done moving and all I will take some pictures and do some drawings, any bugs should have popped up and been sorted out by then aswell. For now I need to jam a few years worth of woodworking and metalworking into the next couple weeks, not going to have a woodshop anymore.

- The problem with a series wound motor is the current usage is very high at startup, and at slow speeds it doesn't develop much torque. If you can find a shunt wound motor, you apply a constant current through the field, and vary the voltage on the armature from a variable or PWM supply to control your speed. This will develop large amounts of torque at slow speeds ideal for coil winding.

I have no idea from series to shunt wound, or what I have, I know it takes a DC voltage up to 12Volts and is happy at 2amps. What you described sounds pretty much like what I did for the controller and it has plenty of torque for my needs at this time. It had no problem breaking 36AWG wire at a fairly slow speed. Will have to get a bigger motor and improve the controller so I can try out winding some big power transformers or outputs in the future.

Thanks for all the help, probably would not have succeeded at this without all your help.