Mini Cim Drive Motors?

[Jonathan Norris]

Originally Posted by Ether

The meaning of the above is a bit unclear to me.

It sounds like it's saying "no gear compensation for the different free speeds". That's not the intent, is it?

Thats the intent that I remember from Paul (I might be totally wrong), but to be honest I haven't worked on motor dynamics in a couple years. Obviously at free speed and stall the mini Cim wouldn't match the CIM, but the dynamics of how the load is shared in a gearbox with motors of non-matching free speeds is escaping me right now. I'll shoot Paul an Email and see if he can correct me.

[Ether]

Originally Posted by Jonathan Norris

Thats the intent that I remember from Paul.

I had forgotten that the mini CIM free speed is 6200. I guess that's close enough to 5310 that 1:1 would work. The mini-CIM would be carrying just a bit more than its "fair share" of the load (relatively speaking).

[BenB]

Originally Posted by Jonathan Norris

Thats the intent that I remember from Paul (I might be totally wrong), but to be honest I haven't worked on motor dynamics in a couple years. Obviously at free speed and stall the mini Cim wouldn't match the CIM, but the dynamics of how the load is shared in a gearbox with motors of non-matching free speeds is escaping me right now. I'll shoot Paul an Email and see if he can correct me.

I wouldn't be surprised if this was the design intent. Joe Johnson explained why it doesn't matter if motor speeds match in this post. For packaging reasons, we geared 2 CIMs 1:1 with 2 mini-CIMs in our climbing mechanism this year and it seemed to work fine.

[Ether]

Originally Posted by BenB

Joe Johnson explained why it doesn't matter if motor speeds match in this post.

To be clear: Joe did not say "it doesn't matter". He said:

The question of whether to put multiple motors in parallel is not a simple one to answer. ... I don't not agree with those who say that you must always match free speeds of the motor

There are times when you do want to match free speeds. It depends on the application and the expected operating conditions.

Here's a simple chart which shows the difference between running a CIM and a mini 1:1 vs running them with matched free speeds. Column H "relative heat" shows the mini's heat generation relative to the CIM. It's an indication of how "hard" the mini is working relative to the CIM. The mini is smaller than the CIM and cannot be expected to sustain the same rate of heat dissipation as the CIM.

Does anyone have any data on the relative heat dissipation capability of the mini compared to the CIM?

[Paul Copioli]

Ok so the Mini CIM was designed specifically do mate 1:1 with the CIM motor. The design goals were:

(1) Same exact mounting as CIM motor

(2) Have current draw of approximately 40 amps at max power.

(3) Optimized (while keeping constraint #2) to mate 1:1 with the CIM motor to maximize the Mini CIM contribution at maximum power. We are about 10% off from this goal.

We utilized a Magtrol automated Dynamometer setup to test various conditions to make sure the contribution of the mini CIM when combined with a CIM motor was as expected.

My team ran mini CIMs along side CIM motors in our drive gearbox (4 CIMs and 2 Mini CIMs) and they performed as expected.

I hope this sheds some light on the mini CIM performance.

Paul

[Tom Line]

Originally Posted by Paul Copioli

Ok so the Mini CIM was designed specifically do mate 1:1 with the CIM motor. The design goals were:

(1) Same exact mounting as CIM motor

(2) Have current draw of approximately 40 amps at max power.

(3) Optimized (while keeping constraint #2) to mate 1:1 with the CIM motor to maximize the Mini CIM contribution at maximum power. We are about 10% off from this goal.

We utilized a Magtrol automated Dynamometer setup to test various conditions to make sure the contribution of the mini CIM when combined with a CIM motor was as expected.

My team ran mini CIMs along side CIM motors in our drive gearbox (4 CIMs and 2 Mini CIMs) and they performed as expected.

I hope this sheds some light on the mini CIM performance.

Paul

This is exactly how we ran our drivetrain as well. 2 cims + 1 minicim per side. It worked exceptionally well.

In answer to Ether's question: We had them 1:1 but the minicim was unpowered except when we got into pushing matches, so we never heated the minicim up.

[Richard Wallace]

Originally Posted by Ether

Does anyone have any data on the relative heat dissipation capability of the mini compared to the CIM?

Only a little data now, will try to get more later:

We fitted a CIM and a mini-CIM with thermocouples to measure temperature on their brush guides; this is the hottest non-moving internal part -- the moving armature windings get a few degrees hotter. On a Magtrol brake dynamometer, we ran the CIM at its normal load, which is about 0.45 N-m drawing about 27 Ampere, per the data sheet. We ran the mini-CIM at 2/3 of that load, 0.30 N-m drawing about 17 Ampere.

As has already been noted, the mini-CIM operating at 2/3 the torque load of a CIM is actually running more efficiently; thus its power losses (waste heat) are lower than the CIMs in both relative and absolute terms.

Our thermocouple measurements showed the CIM brush guide reaching 125 degrees Celsius after eight minutes, while the mini-CIM brush guide took twelve minutes to reach the same temperature.

[Ether]

Originally Posted by Richard Wallace

we ran the CIM at ... 0.45 N-m drawing about 27 Ampere... the CIM brush guide reaching 125 degrees Celsius after eight minutes

Did you record the CIM RPM under those test conditions?

Only a little data now, will try to get more later

This would be interesting: apply whatever load is required to pull down the mini (with 12V applied) to the same RPM as the CIM (1:1 mechanical pairing simulation) and see how many minutes it takes to reach 125C.

[KrazyCarl92]

At 12 volts, a CIM goes 3790 rpm at 40 amps and a miniCIM goes 3375 rpm at 40 amps. Therefore, designing a drive train to pull 40 amps per motor at the traction limit with CIMs and miniCIMs in parallel is near achievable with a 1:1 pairing of the motors.

Running the miniCIM at the same speed as the CIM at the traction limit (1:1 pairing) would mean running the CIM at 40 amps and the miniCIM at 34.3 amps, at 12 volts.

Based on these results, I would say it makes sense to run CIMs and miniCIMs together 1:1 in the drive train if design calls for it. So if you need extra power elsewhere, need weight savings that can't be found anywhere else, or have other design criteria, it should work well. It has worked well for many teams.

From what I've seen, miniCIMs are best suited of any of the 2013 legal motors to supplement or replace CIMs in the drive train since they are designed to take that type of abuse. Most other FRC motors simply are not.

[Richard Wallace]

Originally Posted by Ether

Did you record the CIM RPM under those test conditions?

Yes, but that data doesn't matter for the question you posed: compare rates of temperature rise. Waste heat dissipated by the CIM and mini-CIM depends almost entirely on torque loading.

My test was intended to check Paul's design goal for the mini-CIM, which he has stated was to share load with a CIM (at 1:1 gearing) while taking about 2/3 of the CIM's torque; i.e., the mini-CIM takes 2/3 N-m for each 1 N-m taken by the CIM, so the motors share the torque loading about 60:40.

My results indicate that Paul actually exceeded his design goal () since the mini-CIM's internal temperature is rising more slowly than the CIM's while it delivers 40% of the total load torque.

Conclusion: if your drivetrain needs more power than one CIM can deliver, but not as much as two CIMs, then one CIM and one mini-CIM is an excellent combination. You can run them at the same speed, and mini-CIM internal temperature rise will NOT be a limiting factor.

[Ether]

For the sake of the discussion below, let's agree that the mini-CIM meets the stated design goal for FRC competition use, but pursue the heating question just a bit more.

Given these two motors, how would one gear them if the goal were instead to share the load in such a way as to balance the motor temperatures as much as possible?

There is no single answer. It depends on the location of the temperature you are trying to balance, and it depends on the duty cycle.

Using the CIM motor curves, at 27A at 12V (Richard's test), the CIM would be spinning at 4320 RPM and generating 119 watts of waste heat. Only 66 of those watts, or 55%, is due torque-producing current (I²R heating). The other 45% (54 watts) is due to eddy currents, hysteresis, internal friction and windage, etc, which also cause motor heating.

If the mini-CIM is constrained to the same speed as the CIM (1:1), then at 12V and 4320 RPM it would be generating 133 watts of waste heat.

Under these conditions, the mini would be producing just slightly less torque than the CIM.

So it appears at first blush that the mini-CIM would run hotter than the CIM under these conditions.

However, these numbers obviously change as the motor heats up due to changing coil resistance and magnetic properties. And the temperature has a distribution within the motor due to internal structure and the different sources of heat.

It's possible (but quite a bit of work) to model all these parameters accurately. With the right equipment and setup, a test would yield the answer.

For sake of discussion it would be interesting to include non-competition duty cycles. Are there any teams out there who have a CIM+mini 1:1 on a practice/demo bot? Is either motor hotter than the other after a good hard workout?

[Richard Wallace]

Originally Posted by Ether

It's possible (but quite a bit of work) to model all these parameters accurately. With the right equipment and setup, a test would yield the answer.

One possible test set-up could be a brake dynamometer, a two-motor gearbox (e.g., AM CIMple Box included in recent FRC kits), thermocoupled CIM and mini-CIM motors, and a laboratory dc supply capable of feeding the motors at 12 Volts, with combined current draw up to 84 Ampere. I have all that on hand. The CIMple Box ratio is 4.67:1. Test loads of interest might be 4, 5, and 6 Newton-meter at the CIMple box output shaft -- loads beyond that would require more current than my supply can deliver. At each load, the test procedure would be to record temperature rise until the brush guide reaches 125 degrees Celsius and compare the plots, CIM vs. mini-CIM.

Please check my figures above and suggest changes.

[Ether]

Originally Posted by Richard Wallace

Please check my figures above and suggest changes.

Sounds like a good test. Thank you for doing this.


I would suggest only 2 changes:

1) record the outside case temperature of each motor (in addition to the brush guides), and

2) add a lower torque (higher speed) test point of 2.5 Nm (at the gearbox output shaft).


Rationale:

1) Since motor case temperature is the metric most familiar to FRC students, it would be enlightening to record that and show how it does (or doesn't) correlate with brush guide temperature.

2) Higher speeds are where the mini might be expected to fare worse than the CIM. Useful to have a data point there.

[Richard Wallace]

Originally Posted by Ether

I would suggest only 2 changes:

1) record the outside case temperature of each motor (in addition to the brush guides), and

2) add a lower torque (higher speed) test point of 2.5 Nm (at the gearbox output shaft).

Good suggestions. Our previous tests (see earlier post on CIM temperature rise at peak mechanical power) included thermocouples at four locations: brush guide, internal air, motor case, and lab ambient. And I agree a lighter load point will illustrate the effect of differing free speeds when the CIM and mini-CIM share load at 1:1 gearing.

PM me if you want to come by the lab when this test set up is ready.