Anyone experience neo motor failures? We had 4 fail in the last week. We first noticed that our robot didn’t quite drive straight (neo’s on the drive train). After taking the drive train apart, we noticed a difference in resistance when hand turning one of the motors. We found other suspect motors with similar issues. We did some current testing of good and 2 bad motors in an unloaded setup. The bad motors showed the following symptoms
#1. It spun very slowly for about 30 seconds. When we went to unplug it, the motor wires were hot to the touch and the Spark Max was getting hot too. #2 spun up fast, but screeched and started smoking. Then the breaker on the PDP popped. Again, wires were getting hot.
2 more motors were found to have much higher than normal current draw in another drive train.
Our drive train has two motors on each side connected mechanically with one motor in “follower” mode. I’m wondering if this is somehow related to our issue.
We had this happen with a Neo 550 this weekend, but worse. It started smoking. I called Rev, and Mac was incredibly helpful. He told me that when the encoder cable is not plugged in, the motor controller defaults to max current. The controllers ship from the factory with 80 Amps of max current. This is enough to hurt the 550 in about 2 seconds. The regular Neo takes longer, but still will not function correctly (We went through this repeatedly last year before we learned our lesson, although, TBH, we are still learning it again this year. He recommended we use the software to set the max current at 40 amps for the Neo and around 20-30 for the 550 during prototyping to prevent damage. Which we are doing now.
We are aware this is a real danger, so we were careful in setting it up. However, that doesn’t mean that they aren’t in fact fighting each other. In one case, only one of the motors seemed to have an issue. I would think if they were fighting, they would both suffer and there wouldn’t be a clear winner.
When the encoder cable is not plugged in the motor is essentially at stall. The NEO and NEO 550 both have locked rotor testing data that have time to failure summaries for each.
80 amps is the default Smart Current Limit in the SPARK MAX. It is highly recommended to adjust the Smart Current Limit when driving the NEO 550.
My personal recommendation for the NEO 550 is to adjust down to 20 - 30 amps as there is a longer time to failure when at stall, however this is mainly driven by a conservative design philosophy. NEO is a bit more forgiving and starting around 40 -50 amps and adjusting from there is usually alright.
Your team should consult the locked rotor testing data and motor curves for your motor selection and choose a current limit accordingly to provide the best performance for your design.
It sounds like some of the motors were fighting the others and that lead to higher load which could have damaged them. A few questions:
How are the Idles Modes configured on your controllers: Brake or Coast?
Did you adjust the current limit on your SPARK MAXs? @Michael_Kaurich linked to our Locked Rotor testing for the NEOs in his post above.
Approximately how long was the robot driving before you noticed it veering to one side and how long was it driving in that state before you took it apart?
Did the veering happen after a specific event, like pushing up against a wall, or driving hard for a long period of time?
Describe your drivetrain. Gear ratios, wheel size, etc.
This is definitely a possibility if the followers were not set up correctly. The leaders would be fighting the followers.
We experienced the same thing last year and were unable to resolve it. Our robot always drove a bit crooked. We limited the current to one side of the drivebase to even it out though.
How are teams even experiencing this? Our spark Max’s display the encoder error light code when they are unplugged, and don’t operate. I can’t imagine the spark would continue to try to drive the motor while it knows the encoder is non responsive.
Also, @dyanoshak , any plans for a sensorless backup mode for when encoder wires get unplugged?
It depends. The MAX tries to detect these scenarios, but a disconnected sensor sometimes can look like a properly connected sensor on a stalled motor.
As far as a sensorless backup, we’re looking at how different implementations would work on the MAX, but I can’t commit to a timeline at this time. Definitely not for this current season.
Basically any motor is going to give you that due to manufacturing variations. Some teams resolve this similarly to how you did, but some choose other solutions like using a gyro and having a dedicated straight driving function or PID tuning the drivetrain.
We made a NEO blow magic smoke just a couple days ago after I was told to test the supposedly broken NEO. The wires and motor got hot to the touch when we tried to run them in our drivetrain and then immediately blew smoke when I plugged it into a Spark Max and tried to run it at 100% power via the Spark Max application.