Motor efficiency

So here I was, wandering around the VEX motor testings site, and I noticed that the free current that the CIM takes is lower than the Mini CIM:


I was under the impression that Mini CIMs are supposed to be more efficient than the CIMs, since they use bearings instead of bushings to hold the shaft in place, and that that’s the reason teams used 6 Mini CIM drivetrains instead of 6 CIMs (or so I’ve been told).

Is the bearing thing a hoax? Is there another reason that teams used 6 Mini CIMs instead of 6 CIMs?

Something to do with higher free RPM? Id assume that the bearings wouldn’t make that much of a difference to efficiency, more longevity.

That’s interesting. I was also under the impression that Mini CIMs were more efficient, but according to Vex’s motor testing data they peak out at ~55% whereas the CIM peaks at ~65% efficiency.

I do want to point out that generally 6 Mini CIMs are compared to 4 CIMs, not 6. These combinations offer a similar amount of power but the load is spread across more motors, which is (I think) the reason that people say 6 Mini CIMs are more efficient than 4 CIMs. Adding more motors to the system spreads out the heat that they’re producing. When electric motors heat up their efficiency drops, so I think it’s plausible that this factor causes the 6 Mini CIM system to be more efficient as the higher temperatures in the 4 CIM system causes the higher efficiency of the CIM to matter less.

Disclaimer: that’s all a somewhat educated guess and I’m not sure if my explanation is correct or entirely accurate but I believe it’s kinda close at least. If anyone would care to correct me I’d appreciate it.

3 Likes

Another advantage to using 6 smaller motors rather than 4, is that each one will draw quite a bit less current, for the same overall power. That means that you can get more power out of the drivetrain before breakers start tripping.

Worrying about a few tenths of an amp under unloaded conditions…well, that’s not very important.

Paul Copioli was an early advocate of six-motor drivetrains for spreading the load out; the Mini CIM choice comes from not wanting to take as much of a weight penalty to do so.

While mathematical efficiency is important, so is understanding the power they can supply over time. After 30 seconds or so, the CIM loses most of its advantage over the Mini CIM. Six Mini CIMs would be putting out 1270W after 30 seconds of peak power, while four CIMs at the same point are right at 1000W. (I know robots don’t run consistently at peak power, but it’s the closest analog I’ve got for hard driving.)

We’ve taken that 2-or-3-pound weight penalty the last two years, and combined with a better understanding of calculating time to target (using the ILITE calculator) our drivetrains have been radically better for it while we wait for a little more of the dust to settle between the competing brushless ecosystems.

3 Likes

I’m not worrying about it for competition usage - I get why using 6 Mini CIMs over 4 CIMs is an upgrade, I was mainly talking about the theory of it.

I think there are enough variables in the design and construction of the motors, that the number probably has little relevance.

Kind of like how we don’t really worry about idle quality in racing engines.

Semi-related, but has anyone tested a number of motors to find how the parameters vary among batches or in general irrespective of batches? It’s unclear whether the Vex data is an aggregate or the results of testing a single motor of each type.

There’s a couple of competing effects here. The torque constant of the Mini CIM is about 10% less than the CIM, so it needs to draw more current for the same drag load. But the Mini CIM is spinning faster, so the required torque to maintain speed will be higher.

If we plug in our unloaded values for K_t i = b\omega (equilibrium torque balance) for each motor, we find that the CIM has a drag constant of b = 1.04\cdot10^{-4} \text{N-m/(rad/s)} while the Mini CIM has a drag constant of b = 9.625\cdot10^{-5} \text{N-m/(rad/s)} , about a 7% decrease in drag, assuming b is a constant.

Obviously from the rest of the data we can see that this effect doesn’t lead to a higher overall peak efficiency.

So I looked again in the Mini CIM page on motors.vex.com, and I saw the theoretical specs of it for the first time:


Why would there be such a decrease in performance between published motor specs and experimental data? 6200RPM vs 5840RPM is a pretty big decrease IMO, also the 1.5A delta in free current measurements. My rookie year was in 2018 so I wasn’t around for the Mini CIM launch, but was this a known thing back then? Is the reason for the decrease in performance known?

6200 +/-10% Free speed falls within that number. My money is the +/-10% comes from the permanent magnet tolerance. The weaker the magnet the FASTER its free speed and visa versa. I wouldn’t look too closely at the free current number. A slower free speed with a higher free current are related but there are other things like brush and bearing drag that "breaks in " after some time.

There is much more to motor design then meets the eye which always blows my mind since there is only 1 moving part!

2 Likes