Neo encoder skipping counts

We just redid our turret to use a Neo with 4:1/4:1 rev gearboxes instead of a Neo 550 with 4:1/4:1/5:1. Since the swap, we have found that we sometimes lose track of how the turret is positioned (the encoder counts jump).

We will have the mechanical guys go over the motor and gearbox stack to make sure nothing is slipping (the probable cause), but has anyone seen an issue with a Neo encoder reading jumping when subjected to a jolt?

Late reply, but I’d suspect a loose connection on the encoder cable. If it pops open during a jolt you can get an extra step or lose a step. Rev has been great about replacing our CAN cabling when it gets damaged. I’m not a fan of the JMC connectors they chose, but they have great support.

Another suggestion to look into: if you are referring to the Ultraplanetary, the names of the gear box ratios are nominal. See this website for details:

https://docs.revrobotics.com/ultraplanetary/ultraplanetary-gearbox/cartridge-details#actual-cartridge-gear-ratios

If you are computing the position of the turret based off of the encoder divided by the gear ratio, you need to use the numbers on the site, rather than the “4” or “5” in the ration name, otherwise you’ll be off by a bit. It doesn’t account for the “jumping” that you’re describing, but I could see this causing similar issues depending on how you’ve coded it

If you’re using the versaplanetary I highly recommend making sure the planetary pinion on the NEO 550 shaft is using “green” loctite in addition to making sure the set screw is very tight and that the cost is oriented properly. It appears some NEO550s may have slightly undersized shafts and we have had these pinions slip before. The green loctite is a retaining compound made for slip fits.

1 Like