My team and I have recently been challenged with building a motor controller. All it has to do is move in forward and have multiple speeds. We will soon have 48V 40amp Brushed motor. What all will we need to put together this speed controller? Are planning on buying this breadboard:
Adafruit Perma-Proto Raspberry Pi Breadboard PCB Kit
http://www.adafruit.com/products/1135#Technical_Details

 

Just don't expect the board to carry 40amps
There are plenty of designs out there, the simplest is a SCR bridge, there is no need for a separate DC power supply and variable speed is by simple potentiometer.
Max.

 

Just don't expect the board to carry 40amps

Just depends on how long you need to carry it before that trace turns into a fuse..

 

Just don't expect the board to carry 40amps
There are plenty of designs out there, the simplest is a SCR bridge, there is no need for a separate DC power supply and variable speed is by simple potentiometer.
Max.

It's for an EV. See his other thread. EV, NOT AUTOMOTIVE MODIFICATION. Shell eco marathon entry. He needs to BUILD a PWM controller. Off the shelf is violation of the rules.

 

A 40A brushed motor controller is not too hard. You only need a few main components that need to handle the 40A;
1. switching device (large NFET or a few NFETs in parallel)
2. freewheel diode (T0-220 pack power diodes, a few in parallel)
3. a 50A current sense shunt (a very large resistor with very low ohms)
4. the connectors and wires to the motor

Speed control is as simple as adjusting the PWM duty cycle that is driving the FET, that can be done with a 555 timer and a few parts.

For now a good idea is to google for PCBs or kits for a 40-50A PWM controller for DC motor, and yes i know you are not allowed to buy a controller but some googling will show you how large things are, the way they do the wiring and heatsinks etc on a 40-50A controller.

Kits are particularly good to search as some will have schematics and parts lists that may help you a lot.

 

If you have picked your motor those specs are needed. For example is 40 amps the running current or the stalled current?

 

This is all the specs we have as of now:
Specification MY1020
Rated output Power 750W
Rated Voltage 36/48/60V DC
Rated speed 2800RPM
No load speed 3500RPM
Full load Current ≤26.7/20/0/16.0A
No load Current ≤2.5/2.2/2.0A
Rated Torque 2.8N.m
Efficiency ≥78%
Application Light E.V./ E-scooter

How do we get energy from our battery and make it into a PWM?

 

Something like this might do it. Didn't ask about your batteries. Are they 12 volt lead acid in series? If you think you guys can make this and people review it we can pick specific components.

Have you thought about what you will use for a throttle?

 

Something like this might do it. Didn't ask about your batteries. Are they 12 volt lead acid in series? If you think you guys can make this and people review it we can pick specific components.

Have you thought about what you will use for a throttle?

Our battery is a LiPO4 48 V battery. Just 1 used at a time.

The motor and controller that we purchased last year, we received a twist throttle, we then were able to modify it to become a foot pedal. Since we have switched over to a brushed motor, it has not been discussed.

 

Something like this might do it. Didn't ask about your batteries. Are they 12 volt lead acid in series? If you think you guys can make this and people review it we can pick specific components.

Have you thought about what you will use for a throttle?

While this circuit will work, the frequency changes as the duty cycle is changed.

Wouldn't the addition of a LM393 to the circuit as Bill Marsden does in one of his blogs ( http://forum.allaboutcircuits.com/showthread.php?t=19075 )
Fig.#5.3 about half way down the page be better? With it you set the frequency on the 555, and change the PWM duty cycle with the LM393.

Most of the ready made PWM controllers use this or something very close to it.

 

It's worth including a current sense and current limit, especially where batteries and stall currents (vehicle startup) are concerned.

Some of the better controllers have graded current limit, so at low throttle (0% PWM) the current limit might be 10A and at full throttle (100% PWM) the current limit might be 30A. In between 0% PWM and 100% PWM the current limit ramps up linearly.

The result is a more natural vehicle "feel" and better safety, something like a gasoline engine where the torque is low at low revs (low throttle) and torque and power increase up to max throttle and max revs.

On the other hand, if this is to be drag raced for max acceleration, then you can run the max 30A current limit at all PWMs.

 

What kind of 555 timer should we look at?
In addition, What kind of LM393?

 

Or use the unused half of the LM393 as an oscillator and don't bother with the 555 at all.

Ian

 

Yea, I think Somian is right. The 555 is not good if we need to make 12 volts from the 48 volts because it will try to drive the FET before there is enough voltage to fully turn it on. So what I was working on last night before the power failed was just a 339 oscillator and under voltage detect. I'll redo that one today and see if I can't put in a current limit. Do you have a link to the motor and battery? I think the thru hole 339 is the lm339n.
Does your throttle have 2 or 3 wires?

 

Well, that got a lot more complicated real fast. Lets look at this one:

 

Ouch!

An 8-pin microcontroller like a $1 PIC 12F675 can read the analogue voltage from the throttle pot, and can directly drive the mosfet gate at low PWM frequencies. Even the software would not be too hard.

Sorry Ronv that was not meant to insult the obvious amount of effort you put into that circuit design. It's just that these days it often works out that the best solution is a micro and a handful of code.

 

I just found my old thread for my controller build. For some reason it was really hard to track down. could be of help to you. This thread should be especially helpful; in it, I believe I had a working torque-mode single transistor controller. The design was based off a design by SgtWookie here. I never got around to testing that circuit in real life, but it looked damn good in simulator.

 

Yea, A bit of stuff, but actually only the one IC. It always sounds easier until you actually get into it.

 

So we would be able to replace a lot of that circuitry with a PIC 12F675?

 

You could replace everything to the left of the driver, but would need to add a small voltage regulator and of course the PIC. You would need to buy a programmer(probably about $20) and learn how to program the PIC.