Originally Posted by Kevin Ainsworth
The failure point of all the 775Pros (non-drive train related!!! for all the trolls out there) we have seen this year is on the brush arms. See the link below for pictures. The brush arm seems to overheat where its cross sectional area is reduced, where the brush is inserted into the brush spring arm. Either this or there isn't great contact between the brush and brush arm and the heat is generated at the contact point which therefore melts the brush arm. Older style RS775 motors had brushes with braided copper wire and torsion springs which then lead to failure at the windings, not the brushes. I would recommend adding the cooling into the brush end cap vents where the wires enter the motor, this would be most likely to cool the brush/spring arms which is the main point of failure.
I think that time would be better spent limiting the current/voltage based on motor speed or some safety controls scheme. You can run full battery voltage when the motors are spinning freely and the system browns out without it at low speeds, so you will need this anyways. And on the flip side, a fully stalled motor will live a very short life no matter what cooling you add.
If one zip tie, rivet, nut, etc kicked up from the field gets into your gearbox, no cooling feature will protect the system.
Who else has observed a 775 failure and opened it up afterward? Did it fail in the same place?
If the failure mode is in the brush holder arm, blowing (more) air through the motor might not be sufficient to prevent the failure. If the cross section of the brush arm is as small as Kevin is describing, it's ability to transfer heat to the surrounding air will be very limited and moving more air over it probably won't increase the heat removal by much. Does anyone have a photo of the end cap of the motor with the brushes attached? It is likely that the brush arms are right up against the plastic end cap so it will be difficult to get significant air flow over the brush arms without (illegal) modifications to the motor.
The thermal mass of the brush arm would also be so small (and the thermal time constant so short) that it would likely be futile to try to use thermal sensors on the outside of the motor to protect it.
The only protection scheme that would have any certainty of working would be a current limiting scheme, as Kevin and others have described.