Thank you for pointing that out. I will correct it.
My educated guess is that most of the CIM degradation is from the large amount of heat cycles they go through, and the insulation breaking down.
A CIM is only class B winding insulation (130deg C) vs a Mini-CIM with class H (180deg C).
We (971) started doing a basic check of motor resistance to gauge the health of a CIM, I’ll have to ping someone to get those numbers.
Entertainingly I still have access to a Dyno at work that can handle a CIM, but no one is paying me this time
(I’ll do it when I get bored, which probably won’t happen for awhile).
Usually when such insulation breaks down, it typically leads to catastrophic failure modes rather than small but observable decreases in performance.
It’s not uncommon to see it break down in the tightly wound corners w/ slightly higher resistance (and the bending potentially weakens the enamel some). This can cause a single winding to short out, just decreasing the amount of copper in play.
I am skeptical that this is the mechanism leading to reductions in output power since the techniques for preventing such failures are pretty well established and are common knowledge; i.e. it should be easy to not have such failures. If a motor can exhibit such failures in the relatively short time that an FRC motor is in use, it would mean that the manufacturer of those particular motors is not following standard industry practices and one should expect to see many more significant failures such as very low output power and no output power. The effect will not be consistent from unit to unit. To my knowledge, actual failures in CIM motors used in FRC applications is rare.
If a winding-to-winding short occurs on the same layer of windings, it would be between adjacent turns so only one turn is taken out of play and there is a minor reduction in output. When bent over a corner, it is more likely that the short is across layers of windings taking multiple turns out of play leading to a much larger reduction in output. In a high power motor like the CIM’s, Mini-CIMS’s etc. there will not be very many turns so shorting out even one turn would make a big difference and shorting out one layer of turns would be a major difference in output power and more people would be complaining about it. Has anyone taken a CIM apart to see how many turns are on each pole of the rotor?
Al posted a picture here. Looks like fewer than a dozen turns.
Thanks for posting the link to the photos, Gus.
It doesn’t look like the design of the CIM makes the armature wires go over sharp bends. It also looks like they have inserted insulation between the rotor core and the windings to prevent chafing through the insulation of the wires.
Reading through the thread that Gus provided the link for, it sounds like they were observing effects consistent with failure mode mentioned earlier in this thread of the friction increasing in the bushing.
The experiment that Mr. Forbes is suggesting would shed a lot of light on this issue.
We performed some testing a couple of years back with a torsional pendulum setup. We identified:
CIM: 0.0000775 kg m2
minCIM: 0.0000555 kg m2[/quote]
FTFY - more reasonable precision, and added units.
Here’s the thread.
I wonder if remounting the CIM after rotating it 180 degrees around its axis every once in a while would increase its lifespan.
I can’t speak for other teams, but typically on 973 we make robots that drive forwards and also backwards.
Some teams also drive left or right, but this is a waste of time.
And a select few have driven up and down too…
Just so we all remember: https://www.chiefdelphi.com/forums/showpost.php?p=1338899&postcount=38
Yes, ours drive in reverse as well, but most robots end up driving in one direction far more than the other. Also, for things like shooters and launchers and pickups and climbers, the heavy work is usually biased in one direction.
Another technique to reduce the bushing friction would be to use planetary gearing at least for the first stage.
Put their mechanisms on skid steer and it (could?) would do just as well.
I still have no clue how they were so good at picking up both game pieces.
Never heard of em
Measuring from the motor leads (I am assuming you are) also measures the brush-rotor interface resistance and whatever winding(s) the motor arbitrarily stopped on.
I’d be interested to see your data!