Running DC Motors in Series

[Richard McClellan]

I recently finished building a three-wheeled robot using several left over parts from build season. It has two large driven wheels and a caster in the back. Each driven wheel is powered by two CIM motors, each going through a Dewalt gearboxes, and then both linked through sprocket and chain to the wheel output shaft.

In order to control the 4 CIMs, I did not have any spare 12V Victor 884's lying around, but I did have two 24V Victor 885's. The CIM motors from each side are wired in series off of the Victor 885. This setup has worked fairly well, and I haven't had any major issues with it, but I am wondering if there are any disadvantages to running them in series as opposed to each motor having its own motor controller, such as a loss in total power output.

In a standard DC motor, the brushes are always in contact with either 1 or 2 sets of windings. So, in the case of each motor powering one of its windings, the resistance is equal, and each motor should be getting 12V. The same is true if each motor is powering two of their windings. However, if one motor is powering one winding and the other is powering two windings, one motor will get 16V and the other will get 8V. With this setup, would the voltage provided to each motor actually change in this way, and could it be at all damaging to the motor or speed controller?

[MrForbes]

It would be interesting to connect an oscilloscope to the connection between the motors, to see what happens.

For the underwater robot we're working on with surface batteries, we decided to use a 48v power system, to keep the tether size and tether voltage drop percentage to a minimum. My brother designed and built a couple 4 channel speed controllers that control 12v motors, but give them 48v for up to 25% duty cycle each. It seems to work fine (http://www.youtube.com/watch?v=BLfU7_iz4cs)....maybe you could do the same thing with one CIM per victor, running at 24v, or just try them at 12v and see how they work.

I didn't look at the specs for the 885, but does it give a minimum voltage? The Banebots speed controllers are spec'd for 6v to 24v, we used some last year at 14.4v.

[Al Skierkiewicz]

Richard,
You didn't specify whether your battery was a 12 volt or 24 volt source. If the battery is only 12 volts, the 885 doesn't do anything different as opposed to the 884. It merely uses FETs in the output that are a higher breakdown specification and that can handle 24 volts without damage. The current is what affects the motor speed. So as the motors turn and contact one or two windings as you have pointed out for the CIM motors, the current then steps four times as opposed to two. Since the current remains the same in a series circuit, the current through both motors is then affected. While one motor is contacting one winding, the other motor receives current dependent on how many windings are in contact with the brush assy. Minimum current would occur when each motor is only contacting one winding each. Max current would occur when both motors are contacting both windings each. However, if the source voltage is still 12 volts, the motors in series would only be able to reach half the maximum current. The motor specifications are taken with a power supply at 12 volts. So stall current for CIM motor of 133 amps would be halved to 66 amps which would flow through both motors. This may actually be advantageous in your application.
To remind everyone, not all DC motors are designed or operate alike nor are they electrically equivalent. They can have more or less than two windings in parallel.

[daltore]

Really the only thing I can think of that could be a problem in a series circuit is that you're having the Victor source vastly differing amounts of current that are fluctuating quickly. To lessen this effect a bit, you could add in a medium-sized capacitor in parallel with each motor to kind of even out the current flow a bit. Something around a couple thousand microfarads should probably take care of everything. I could try to do the math, but on the other hand, I haven't taken calculus.

[Al Skierkiewicz]

Aaron,
The caps you are suggesting are available in electrolytic varieties. Those types of capacitors do not like having reverse polarity. The PWM output of either type of controller does not work well into a capacitive load. The controller is asked to charge the cap and move the motor at the same time.

[daltore]

Right, good point. I was remembering the ceramic capacitors on small-scale R/C cars.