paper: Preliminary REV MAX and NEO Stall Torque Testing

Thread created automatically to discuss a document in CD-Media.

Preliminary REV MAX and NEO Stall Torque Testing
by: juchong

A summary of data collected on traditional FRC motors (CIMs) and the first brushless FRC motors (NEO).

Stall torque was tested by fixing the motor undergoing testing onto a steel plate screwed to a table. An 8mm to 1/2" Hex shaft adapter was used to simplify attaching a 1/2" wrench to the motor output. The motor under test was driven to stall and the force measured on the scale was recorded. Each test was repeated 5 times with reasonable “ramp up/ramp down” intervals and averaged. Note that the MAX motor controller is the only controller capable of driving the NEO motor to date, but the MAX is also capable of driving the brushed (CIM and Mini CIM) motors.

Prelim_FRC_Motor_Testing_JC_DH_Rev1p4_20181117.xlsx (11.9 KB)

Update 1.3:
Captured data on a brand new CIM. Numbers adjusted from previous version are highlighted in red.

Update 1.4:
Added compensation to account for the angle of the wrench during testing.

What time did you stay up until, or are you still there??

Today 01:34 AM

This is what passion looks like. Thanks for sharing the data!!!

Great work, good to see our results concur!

That’s 2:34am where Juan lives!

The motor controllers used are a Talon SR, Victor SP and Spark MAX, correct?

From talking with Juan, yes.

Personally, I’m curious as to why the CIM’s readings vary so much. I could see it going up from the older-design Talon SR to the more modern Victor SP, but the drop back on the SPARK MAX is curious. And doubly so when the Mini CIM sees none of this variation.

Please don’t take this as me criticizing the value of your efforts, because I very much appreciate what you’ve done to provide information in the absence of manufacturer data on the motor, so thank you for that.

My concern is the inconsistency in the difference between the torque data for the Cim and Mini Cim on the Max. If you compare the results to the motor vendor’s data one motor is ~11% greater than the value stated in Vex motor testing data and the other is ~11% less than the torque stated in Vex motor testing data.

I know this test is imperfect and has some downsides, I’m just wondering if you have any insight into that particular inconsistency? Any one else with insight, feel free to chime in.

I think its safe to assume that the SR is current limiting to around ~80A, and that the MAX is limiting to around ~110A. It would explain the differences in the data. Especially given that the stated max “burst” current of the MAX is 100A according to rev.

One is +11% the other is -11%. If they were both +11% the higher current would certainly explain that, but they’re skewed in opposite directions.

Sorry for the confusion, my original comment was a response to billfred’s observation.

Looking at your observation though, I believe that the current limiting can also explain the data. In the angle compensated testing, the cim is +6% and the minicim is +12% for the SP (which is assumed to be the controller that is not current limited) For both the SR and MAX, the stall torque should be under the vex data for the motors.

Yes, you’re absolutely right. I’ve got it now. The stall current of the Mini Cim is below the current limit of the MAX, so it’s actually reaching the limit of the motor, whereas the CIM is actually higher than the 100A limit of the MAX is cutting off the top of that motor curve. This actually makes the Neo that much more impressive. Brushless efficiency I suppose.


It was a rather late Friday, but I’m glad I got the data out there! :slight_smile:

Correct! I suspect the SR has an upper limit since the MOSFETs used in the motor controller are likely a generation or two behind what’s in the MAX or SRX.

I suspect that the Victor SP doesn’t have any upper current limiting, so given enough time pushing 100+ Amps through the controller it’ll eventually die. The Talon SR is probably limited due to the hardware and the MAX is limiting current to 100A in software.

No worries, I understand that the measurements seem a bit far from everyone’s expectations. I too thought that my data seemed far away from what VEX has distributed, but after thinking about it my test data seems to be more in-line with real-world testing. I suspect that VEX recorded data using a bench supply using ideal conditions where I performed all testing using more real-world conditions. I can’t fix the voltage at exactly 12V since I used a battery as my power source, so some of the variation could easily be attributed to this one difference. We’re also assuming that the CIM and MiniCIM are both manufactured to tight tolerances. Who knows what guard bands these Chinese manufactures have in place.

Exactly! I’m working to secure a calibrated force gauge to use for additional testing. This is 100% anecdotal, but at about 85~90% stall I observed an additional ~25% torque from the NEO before reaching full stall (which is the number I stated in the document).

So, when are we going to push one of these to failure? I think that is the big question that still needs to be answered in my mind to determine if this motor will truly be able to replace CIMs and mini-CIMs in all applications.

It seems like REV has provided a few of these for people to test. Right now they are probably too valuable to intentionally break them. But at some point, somebody is going to have to push them to the limit to see where the breaking point is so that we can program the controllers to stay away from that limit.