Yes, we have had problems. We followed the instructions posted by Banebots when assembling the CIMs, used a spacer to accomodate what we thought to be motor shafts that were too long. Encoders were installed and calibrated, the things ran well and pushed our proto bot nicely. However, after about 3 hours of running, one of them developed a lot of what was thought to be gear backlash, and so it was pulled apart.
The gears were fine, but what was damaged is the planet gear carrier that affixes to the output shaft. The output shaft has two milled flats that caused the planet carrier to become damaged, allowing almost 90 degrees of wheel rotation before engaging. This drive assembly also trashed an encoder. What happened to the encoder came about as a result of excessive axial end-play, moving the encoder wheel assembly into contact with one of the sensors, effectively grinding away the sensing element. The encoder wheel appears fine. When the encoder wheel was affixed to the shaft and glued thereon (following the packaged assembly instructions) there was no discernable end-play and the wheel was centered in the encoder sensor.
We were able to resurrect the drive plate and reassemble the transmission to find that the spacer that we needed between the motor and its mounting plate was not necessary. Apparently the assembly had about .040” end-play from the factory because of the output shaft not fitting snugly into the planet carrier.
Dr. Joe has a post in a separate thread that talks about this problem. We are investigating ways to improve the lifetime of the planet carrier, maybe it can be hardened. Is welding the shaft and carrier together the only solution?
C. B. Petrovic
Mentor, T166