This is my latest gearbox that we want to make as an off-season project. It is made for an 8 wheel West Coast, direct driving one of the middle wheels, and belts go out to the others, I will upload a picture of the frame when it is finished. The gearbox is a 2 speed ball shifter, with final ratios of 3.66:1 in high, and 7.35:1 in low, for projected speeds of about 18-20 fps in high and 7-10 fps in low on a 4" wheel. The PTO is driven off the idler gear between the 2nd and 3rd motor. Engaged to the hex shaft with a dog tooth with a final ratio of 4.73:1. according to solidworks, the assembly weighs about 11.5 pounds with motors and pneumatics. Please post any questions, comments or concerns. Suggestions are more than welcome.
Looks nice! How does it mount to the chassis?
The one thing I would caution against is that if you gear 6 cims that high, you’re liable to brownout the RoboRIO.
Very nice, I like it!
Oh brownouts… 
I’m not sure what is going on with beta testing, but for the 2016 year, is there going to be a fix for this?
Yeah, we have the bownouts in mind, thats why we want to experiment with this much power now so that we dont have to troubleshoot during build season. We want to try to implement some auto shift code to try to eliminate some human error. So if anyone has any input on auto shifting, please share. As of right now, I dont know if it would be best to put the encoder on the motor side of the gearbox, or the wheel side. (reading the speed of the motor vs. reading the speed of the wheel because it will change with shifting). Also, the gearbox mounts with 2 1/4-20 screws going through the 1/8 wall of the tubing on the frame, with the hex shaft mounted permanently to the frame, and just going into a female hex on the ball shifter, allowing the gearbox to come off without removing any axles.
I really like this design - having a quick swap gearbox is useful.
We did the same thing in 2014 with a 3/8" female hex made from a steel hex sleeve from SDP. There are two things to consider with this type of design: the gearbox flexing away from the frame and the alignment of the output shaft with the wheel shaft.
The fit of our female hex was sloppy, so we definitely needed two bearings on the gearbox shaft and two bearings on the wheel shaft. Getting the alignment of these four bearings is crucial for this design to work. The same applies for the mounting holes, but these could always be slotted if alignment is an issue. We had issues at first when we were using Vex hex shaft that was bent. As of 2015, the shafts are MUCH better, but there are still two very different versions of the hex shaft sold by Vex.
Your setup to hold CIMs will help decrease the moment on your center plate, but there will still be a moment on the 1/4-20 bolts and the wall of your tube. Be aware that different CIMs have different OD’s. The wall of the frame tube is pretty thin, and I’d be worried about the holes being enlarged or the wall being bent, though I have heard of teams getting away with a setup like you have. We welded inserts into the frame rail that were connected to both walls of the frame tube, but this may have been overkill.
You might have some trouble shifting that PTO depending on what it’s powering. We coupled a 1.5" or 2" bore (can’t quite remember) to a ball shifting shaft on our winch in 2014 to engage/disengage a ratchet and it often had problems shifting as the cylinder induces a bending moment on the shifting rod. No idea if you’ll see the same problems with the smaller cylinder, but making sure the coupler is as stiff as possible will help.
Do you think the extra plate for supporting the cims is really necessary? Adds a lot of parts and tolerances.
You said you guys are going to test for brownouts. Are you going toramp up the speed until you achieve the brownout or up to a speed conducive with an event type environmnent?
The brownout behavior is a fix for the roboRIO and other electronics rebooting if voltage gets too low. It is a safety precaution that triggers when you are asking more our of your system than it can reasonably provide. The brownout is there to keep the user from doing something they shouldn’t to the electronics. Generally, a brownout is better than radio reboots or rio reboots.
Eh, only 3-5 extra parts.
It looks like the third plate is for mounting and giving the cylinder for the PTO some wiggle room.
we haven’t had any problems with alignment between the plates in the gearbox, but I do like the idea of slotting the holes, we will consider that if we have problems later on.
We have the piston attached to a 1/4 plate that’s only an inch long, so I don’t believe flexing will be much of an issue. We did this same type of shifting on our lift this year and didn’t run into any problems.
The extra plate is only one extra part, and it keeps the motors from sagging outwards, if we find that it isn’t necessary and becomes a hassle, we can just simply remove it and make the final hex shaft shorter.
We might experiment with ramping up speeds when we get the drivetrain built, but until then, we are focusing on actually getting it built.
The cylinder isn’t mounted on this plate, I didn’t want to attatch anything to that plate in case we end up taking it out. that 3rd plate is just simply extra support to prevent sagging.
You can see it better in this picture http://www.chiefdelphi.com/media/photos/42422?
Sorry, should’ve worded that differently. I realize that the cylinder is mounted to the second plate. When I said mounting, I meant gearbox to chassis.
Would it be at all possible for you to post the CAD for this gearbox? I myself am in the process of designing a custom gearbox with ball shifters for both a PTO and for shifting between two speeds, rather than using a dog tooth for the PTO as you seem to have done.
Oh man, I never thought this thread would come back up. I’d love to post the CAD, the only problem is that it’s so old I’m not sure if I have it anymore. I’ll dig around and see if I can find it and post at some point