Has anyone here successfully used set screws to reduce the backlash between hex hubs and hex shaft?
My team is fairly confident that most of our arm’s backlash comes from the slop between the hubs and shaft. Right now we have a functional robot, but the backlash is annoying and we’re looking for options to fix it.
Directly attaching our sprocket to the arm isn’t an option unfortunately.
This is a fairly high load (200 inch*lb arm moving fairly fast) and I’m not sure how best to reduce the backlash in our system.
You can always try to slide shim-stock in between the sprocket and shaft. Or you can bolt two hex-hubs together but offset them before you tighten them.
We have the same issue, but we haven’t had time to improve it and it doesn’t really affect the performance of our mechanism.
I prefer keyed shaft or even dead axle for high load applications like high-torque arms. I’ve always found that hex develops a pretty high amount of backlash quite quickly. We have a very high torque arm on our robot that is running on a 1" keyed 6061 axle and 40 chain. Standard FRC axle and 20/30 chain just won’t cut it for a certain threshold of rotary mechanisms. If possible, try to switch it out for 1/2" keyed shaft, and if for some reason you have access to a TIG welder, you could always try welding the key to the shaft, assuming you have some other way of removing it if it breaks.
How is it driven? Most hex interfaces from VEX have some slop in them; their shaft is undersized and hubs/gears/etc are oversized to ensure you always get a nice easy fit. I don’t doubt some of the slop is between the hex shaft and hub, but if you have some spur gear reductions before that, the slop in those can be compounding to the output as well.
This process sucks, but one solution is buying hex shaft from a different supplier that normally comes oversized and milling down each surface until you get a nice tight fit on the hubs and/or gears.
What is the length of the shaft between the arm and where the arm is driven?
We were concerned about torsional flex in our arm shaft so we tested it. Our original design would have around 8" of shaft between the sprocket and the arm. We estimated a max load on the shaft of 120 ft. lb. I put a piece of hex shaft in a vice with around 8" sticking up. I set a torque wrench to 50 ft. lb. and used it to twist the shaft. We got around 15 degrees of twist and the torque wrench had still not clicked so we ended the test.