Stupid question: Is there a method for determining good vs. bad CIM-class motors using a multimeter?
I wouldn’t think so. You might get a slightly higher resistance reading if the brushes have deposited enough material (Are these copper impregnated or just carbon? Can’t remember) on the commutator, but multimeters aren’t great at reading low resistnace. It’s also not going to tell you how the bushings have held up. My guess is you’d have a better idea on the status of an old CIM if you used a screwdriver rather than a multimeter. If there’s little to no buildup on the commutator, it might mean it’s fine…might.
I’ve used a torque wrench readout on the stalled rotor to compare older CIMs with newer ones while measuring current draw.
Usually it’s just to sort out better CIMs for non-competition or practice robot uses.
Not really. Your best bet to test old CIMs is a test board with a RoboRio, PDP, and motor controller (preferably a good one like a Talon SRX.) We keep one around specifically to test motors.
We’ve tried. You’re not going to be able to measure the change in an already-low resistance with meters you can get.
The best way we’ve found to verify old motors is a bit complicated, but it works:
- Back drive it to a known speed. You can do this with a drill and an optical tachometer, or with another motor if they’re installed in a multi-motor gearbox
- Measure the voltage generated by the motor (or the current generated across a known resistor) at that speed. Compare the reading to a known good motor. A bad motor will also be a bad generator. Good motors should all agree within ~10%.
If you were testing a large number of them, you could build a simple rig using VP gearboxes where you back drove the CIM motor with a NEO. The gearbox would allow you to attach the CIM and the other motor to the gearboxes fairly easily and using a hex to hex coupler, you could take the output shaft from one gearbox to drive the output shaft of the other to back drive the motor that you are testing. I suggest using a NEO because it has the built in encoder so you would know the speed and given the simple interface through the USB port into the SPARK MAX controller, you could drive the whole rig directly from a laptop without needing any special programming. You could then back drive the CIM motors at various speeds up to their max rated speed to characterize the motor. I don’t know if that would uncover issues that a single speed test wouldn’t but it would give you the flexibility and repeatability to test them very thoroughly.
Is there even a reason to use a whole VP setup instead of just a piece of C channel and an 8mm shaft coupler? Are you thinking of putting a gear reduction between the NEO and CIM? I would think the faster you run the test the more accurate results you’ll get because the higher output voltage will mean measurement noise is less pronounced. I don’t know a ton about DC generators though, so I might be missing something.
+1 for judging good/bad motors by their back-EMF, especially if you have a way to confirm the speed while measuring the voltage.
Our preferred method is to drive them at a known voltage, and measure the free current. We built a tester for that purpose a few years ago. It also allows us to run-in new motor-gearbox assemblies.
Free current draw (12V on the terminals, nothing on the motor shaft) should should match the data sheet value; lower is better. It is basically a measure of the motor’s internal drag caused by several factors, including friction of the bearings and brushes, and iron losses due to alternating magnetic flux coupling the stationary magnets with the armature core.
It is also a good idea to run the motor CW and CCW, checking for similar free current draw and sound both ways. CIMs have witness marks on the black case and the lead-end plate which should be aligned, indicating that the brush holder is correctly clocked.
As I read @Billfred’s OP, I was starting to phrase @Richard_Wallace’s “generator” method until I’d read that far; the bottom line is that a good motor will make a good generator. If you already have a bunch of FRC hardware hanging around, the easiest way to do this is definitely to use a 2+ CIM gearbox with an encoder, and spin up to a known speed approaching free speed but definitely below (say 3000 RPM for CIMs); the actual speed isn’t important as long as it’s consistent. The problem with an 8mm shaft coupler is that you still have to keep the motors from rotating with respect to each other.
Well, I guess I was thinking that to get the best test results, you would want to be able to secure the motor (by bolting it to a rig) and since the VP motor collars are easy to mount to the shaft, it would be an easy setup to attach a motor to. But there is really nothing special about the gearboxes other than the mounting holes are already there to use. Also, the mounting holes on the gearbox make it easy to secure to the workbench.
Given the speed similarities between the NEO and the CIMs, I wasn’t thinking of any sort of speed reduction. I was just thinking of a 1:1 straight from one motor to the other.
So really, it was just an easy set of tinker toys to use to build a rig.
You could easily build a coupling rig with C channel and an 8mm shaft coupler as you suggest. It is a bit more fiddly as the bolts for the two motors would be in a tight space in the C channel which would make them a little harder to tighten down. You could just slot the bolts from the test motor into slots in the C channel rather than holes, but the alignment might be harder to get straight. But it would probably be good enough for the test.
So, yes, the rig could be done in many different ways. I was mostly suggesting to use the NEO motor to drive the motor you were testing rather than using a hand drill and optical tach. Both would work, but if you have a large number of motors to test, building a dedicated rig might be worth it however you construct it.
Piling on to these “how to attach two CIM-class motors to eachother”, there’s a very easy method almost any team can do with parts on hand. The CIM’s keyway is close enough to the end of the shaft that inserting a key into each motor and using a standard pinion between the two (with zip ties or tape to prevent it from sliding) should work fine for a relatively low load test.