Results of NEO motor testing

#1

After 33 competition-like simulation matches using 4 rev Sparks and Neos on the drivetrain of our off season bot, our team found no major functional problems with the sparks at high battery charges. Each simulation was a 2m 30s match, followed by a 6 minute downtime, where the robot was powered off. During these tests, we found that we faced brownout problems at batteries with lower than ~100% charge. The motor controllers also run slightly warmer than talons, but not enough to make a major difference, however other teams have reported them to melt. (this is incorrect) We also noted that the motors a have a large EMF field of about 10~12in. To test this, we ziptied our radio to one of the motors, and ran for 10 minutes, however we encountered no radio problems during that or any of our other testing. Other than that, the motors performed excellently, and had no problems.

1 Like

#2

Have you encountered the problem with encoder values returning 0? If so, has it been fixed for you in the update from today?

0 Likes

#3

Could you please provide a Cite regarding the melting of motor controllers? Test conditions etc? Saying that a motor controller melted would be very worrying to the manufacturer, and I’m sure they would like to investigate the root cause of an incident like that.

2 Likes

#4

Would you be able to provide a bit more information on your testing setup? Wire length, (motor and battery), battery age, gearbox reduction, wheel size, and drivetrain configuration all play a part in this.

0 Likes

#5

We haven’t tested that yet. We only wanted to test reliability and consistency of the motors. The tests were performed last week so we haven’t used the latest firmware

0 Likes

#6

Temperature rising too high (melting may be a bit too much in hindsight) was discussed in this Chief Delphi thread: Rev Spark Max - Mounting Solutions and REV already has a Trello card for temperature monitoring https://trello.com/c/0TD3BRhv/51-motor-temperature-throttle-and-cutoff-functions
I personally wasn’t the one to conduct the tests, it was another team member and he wrote the summary I posted. Although I do recall at one point seeing something about melting, I can’t find it at this time so it may be a bit of an exaggeration. However, we still found in our testing that the Spark’s got a bit hot (not to unreasonable levels) and since the bottom of the Spark is aluminium we are using thermal paste while mounting.

0 Likes

#7

Each 2 minute 30 second “match” used a different, fully charged battery (around 130% charge according to the driver station, which is around ~12.5v). We found that on lower charges, at around ~12v, the Spark’s start to brown out and jitter. Variety of different batteries were used at different ages.
The robot had 4 NEO’s + Spark’s, each using the wire lengths out of the box (with Anderson’s connected of course). Additionally, we had some Talon SRX’s and Victor SPX’s on the same CAN Bus running a manipulator in parallel to test whether the Spark’s can coexist peacefully with other motor controllers on the same bus. During our tests, we ran the mechanism and drove around to simulate the most load as possible on the whole system.
There were 4 6" omni wheels on our drive base, tank drive. First stage was 7 to 36 and second stage was 2 to 1. Frame perimeter was 28" x 31.5".

1 Like

#8

We have received zero reports of a SPARK MAX melting. If someone has experienced an issue with SPARK MAX or NEO, minor or major, please reach out to us at support@revrobotics.com so that we can help troubleshoot the issue.

A newer version of the firmware, 1.0.384, was release last night that includes improved brownout behavior. From my experience, the batteries will be around 13V or even a bit higher when coming fully charged off of the charger. I would consider 12.5V and especially 12V significantly drained and I suspect this is the major cause of the brownouts you experienced.

Brushless motors inherently produce more electrical noise than brushed motors, so you can be expected to see more noise with the SPARK MAX running a NEO. Even though your experiments with the radio attached to the NEO were successful, it is always best practice to keep your radio away from any noisy electrical components (brushed and brusheless motors, switching power regulators, etc.).

0 Likes

#9

That thread didn’t seem to have any test results that said it got too hot, just people who wanted to take heat dissipation into account when they mounted. Do you have another reference?

0 Likes

#10

I guess we just misread or misheard something somewhere. Regardless of that statement, we found that our Spark’s did not get excessively hot to the point where it was an issue. I am sorry that we may have misreported what we heard and I will edit the post accordingly.

0 Likes

#11

I’m surprised it lasted that long! Our brushed motor drivetrains (775pro) start browning us out at around 12.2V on the battery, and in general I consider anything with a surface voltage lower than 12.3V dead.

Thanks for doing these tests!

5 Likes

#12

In your tests did you do any pushing where you stalled motors?

0 Likes

#13

No, we simply wanted to test the reliability and consistency over many matches. We haven’t tested stalling the motors.

0 Likes

#14

I’m one of the mentors on the same team liavt is on.
Just to clarify, we did do a wall push, but when you’re using Omni’s it’s kind of a moot point. We did make a nice friction burn a nice tire pattern in the carpet though. :smile:

1 Like

#15

How do these motors do with stall? In other words, can they handle being stalled like the cim or are they more delicate aka 775?

0 Likes

#16

Only a partial answer:

We haven’t tried stalling these at full power, but they stall just fine at around 3V. When your battery voltage starts dropping (running one battery for 30-50 minutes during programming or practice) they seem to lose a disproportionate amount of power.

1 Like

#17

They’re much better than a 775, but seem to not be as good as a CIM, from what I’ve seen. Granted, I haven’t done destructive stall testing on the motors, this is all anecdotal.

I’ve seen a number of teams use them successfully on their drivetrain over the past two weeks, and I’ve also seen one of them burn out on a drivetrain. But just one.

1 Like

#18

We did some quick stall testing this build. It’s hard to compare apples to apples because the Spark Max’s “smart current limiting” behaves a little differently than the Talon’s. However in general it looked like:

The Neo / Spark Max can* be stalled for 30 seconds safely at 12V, 15 or 16A “in”, which corresponds to a smart current limit of about 60 -65A “out” in our testing.
(*I am not providing any guarantee here. Get something in writing from Rev or conduct your own testing)

For reference, our testing this summer suggested a 775 pro could only be stalled at 12V, 9A safely. The CIM has been known to stall at 30+ amps without consequence, as long as it eventually gets some time to cool down. So yeah, in terms of stall safety, the NEO is somewhere between a 775 pro and a CIM (but closer to the 775 pro on the scale)

1 Like

#19

We just released some locked-rotor testing data here: NEO Brushless Motor - Locked-rotor Testing

The intent of the data is to show an approximate time-to-failure, not the current-torque relationship of the NEO. The time-to-failure depends on many different variables and the times you’ll see in the graphs linked above are not guaranteed.

Please take a minute to read through the information as the data is in a different format than what most are used to. In short, the data was taken at constant current limits instead of constant voltage limits.

1 Like