Motor Turns Slowly in Reverse After Stopping

[Brent_C]

I have an ELF (with a mid-drive motor that goes to a rear freewheel). The motor is a XOFO DD45 and it runs off of a Grin BaseRunner v4 L10. Though I don't think it is relevant, there is also a second motor running on a separate controller, both running off of my CycleAnalyst v3. I recently pulled the motor out and reinstalled it - almost 100% sure I didn't do anything different from how it's always been (for years). It runs great and seemingly normally ... until I stop. When stopped, if the motor turns backward the slightest amount (such as if the vehicle rocks back on a hill at a stoplight), the motor will start to very slowly rotate in reverse with no throttle input.

By "slowly", I mean something 30 RPM. And, I can reach under there and stop the motor easily with one hand. When I do, it twitches a couple of times, then stops. It will sit motionless unless I turn it again ... even just like a degree.

It doesn't feel like it is actually trying to run in reverse. It feels more like it is doing some weird e-braking/regen thing. I don't have any real evidence for that, so take it for what it's worth. I have tried tweaking the braking and regen voltages to no avail. The input min/max voltages are the same as the other motor/controller which does not exhibit this same issue.

I think the best clue, here, is that if I turn off hall sensor start and run purely sensorless, the problem goes away. I would blame a hall sensor, but I tested each individually and can also see that they are firing in a logical order: Blue/Green -> Blue -> Blue/Yellow -> Yellow -> Yellow/Green -> Green -> Blue/Green.

In practice, this isn't really causing any problems. The motor runs like a champ and when it turns slowly in reverse, it really isn't hurting anything since it is on the freewheel. But, it makes me wonder if it is not a symptom of something I do need to pay attention to.

Any thoughts?

Cheers,
-B

 

[amberwolf]

If it didn't do it before, then something did change during the reinstallation. Presumably there were no settings changes within the controller or CA during any of this (if there were, you'll need to list specifically what was changed, what it had been and what it is now).

The usual problems are wiring, for things that were working, then disconnected, and reconnected:
--misconnection/wrong connection (swapped halls or phases)
--poor connection
----spread contacts so they don't mate properly
----a contact backing out of it's shell during plugin but looks normal when unplugged
----wire broken at the back of a contact or within a cable at a bend, etc., perhaps makes partial contact but insufficient

Depending on the specifics of the motor removal and reinstallation, (if anything was disassembled), there could be other mechanical issues.

 

[Brent_C]

Mystery solved. Obviously this is not a new issue. It's just that I didn't notice it until I got it up on the jack stand the other day. The issue is the controller was set for a throttle voltage range from 0.2V to 4.0V with a brake start voltage of 0.0V and brake max voltage of 0.2V. The CycleAnalyst was configured for a min throttle output of 0.0V.

So ... when at no throttle, the controller was interpreting this as a brake signal.

I'm not entirely sure why this would cause it to slowly/gently rotate backward, but it really did feel more like an e-braking sort of thing than a reverse throttle sort of thing. My guess is that e-brake applied to a motor with no resistance (i.e. on a freewheel) was subject to this sort of "drift"? I suppose?

In any case, changing the CA min throttle output to 0.9V did the trick. The trottle range is now well clear of the brake range.