Motor Speed Control Problems
Greg Norton
Posts: 70
My brother and I are working on a project. Our goal is to control the speed of a large DC motor (starter motor from a car). We have a commercial motor speed controller to do this for us (AXE-2444) and I built a circuit to supply control signals using a BS2 and a DS1868 digital pot configured as a voltage divider. Everything works fine when my circuit it used without the speed controller. However, when we hooked it together with the commercial controller we found that the output signal jumps around wildly (as much as 2 volts on either side of where I think it should be) when the motor is running, and the speed of the motor jumps around as a result. We found that this happens even when using an independent power supply for my circuit, and also when my circuit is disconnected (but still in close proximity) and we start the motor using another method of control. We plan to do another experiment by moving my circuit farther away from the motor controller and the motor itself to see if the problem will resolve itself this way, but the whole family is traveling for the holidays and I won't have the answer on this until after the new year. Eventually we would like to be able to mount all 3 (motor, controller, and my circuit) in close proximity to each other.
My theory is that the motor speed controller is using PWM and thus creating large magnetic fields around the motor when it turns on and off the large currents required to run the motor. This might be affecting my circuit by either a) inducing the voltage changes on the wires coming from my circuit to the speed controller, or b) causing noise in my circuit that in turn causes the digital pot to jump around.
In the "a" case above, the only solution I can think of is to twist the wires between my circuit and the controller box. Given the magnitude of the voltage fluctuations we are seeing, this doesn't seem likely to work. If it is "b", I don't really have a good solution for that case. My whole circuit is mounted inside an aluminum box, and we have tried connecting the box itself to the negative terminal of the car battery that supplies power to the whole setup.
So here are the questions, any and all suggestions are much appreciated.
1. Does anyone have any alternate theories on what might be going wrong?
2. If the problem is one of my theories above, are there any other solutions I might use?
3. The idea of grounding the aluminum box was to try and achieve a Faraday cage effect. Do I need to use an actual "stake in the ground" type ground or is connection to the negative battery terminal sufficient to get what I was looking for?
Thanks.
Greg
My theory is that the motor speed controller is using PWM and thus creating large magnetic fields around the motor when it turns on and off the large currents required to run the motor. This might be affecting my circuit by either a) inducing the voltage changes on the wires coming from my circuit to the speed controller, or b) causing noise in my circuit that in turn causes the digital pot to jump around.
In the "a" case above, the only solution I can think of is to twist the wires between my circuit and the controller box. Given the magnitude of the voltage fluctuations we are seeing, this doesn't seem likely to work. If it is "b", I don't really have a good solution for that case. My whole circuit is mounted inside an aluminum box, and we have tried connecting the box itself to the negative terminal of the car battery that supplies power to the whole setup.
So here are the questions, any and all suggestions are much appreciated.
1. Does anyone have any alternate theories on what might be going wrong?
2. If the problem is one of my theories above, are there any other solutions I might use?
3. The idea of grounding the aluminum box was to try and achieve a Faraday cage effect. Do I need to use an actual "stake in the ground" type ground or is connection to the negative battery terminal sufficient to get what I was looking for?
Thanks.
Greg
Comments
DS1868 digital pot.
You should consider placing a capacitor from the DS1868 output
pin to ground.
A high output impedance makes it very easy for noise to be
coupled to the controller input pin.
You'll probably have to experiment, but I'd start with something
like 0.1 uf.
If this is the problem, you could add a non-inverting, unity gain
op amp at the digital pot output instead of a capacitor. This will
make the noise reduction more effective over a wider frequency
range.
phil kenny
Post Edited (phil kenny) : 12/21/2007 5:01:29 AM GMT
Thanks for the reply. The output circuit is just the digital pot itself. One end connected to +5v regulated output from an LM7805. Other end is grounded, the wiper is connected to the controller box. The BS2 sends serial commands to control the position of the wiper. However, I understand what you are saying. If anyone is interested, the complete schematic is attached. Output pin to the throttle is named throt_pos.
Thanks.
Greg
PS - this was in answer to a question in Phil's original post. We were posting at the same time, not rambling here...
Post Edited (Greg Norton) : 12/21/2007 5:37:44 AM GMT
Have you got the motor controller set to the right throttle configuration? I see there are several possibilities and if you've got it set wrong, you'll really load down the digital pot output. I think the 0-5 volt setting is the one to try, along with a bit of filtering.
Cheers,
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tom Sisk
http://www.siskconsult.com
·
Yes, we verified that we have the controller set for 0-5v mode.· Thanks for the input.· I'll try adding the filtering as suggested, but it will be a while before I have a chance to try it out.
Greg