**With over 6000lbs of clamping force and 3 tons of press force!
Oh… wait this isn’t battlebots….**
All joking aside, after seeing how many folks were talking about 775pro drivetrains. We wanted to design and build a 775pro style WCD this fall season to test and develop to see if it was a viable idea moving in to build season.
I recently acquired limited access to some neat machines, so the design for the gearboxes ended up being one billet housing with an internal frame, and then a cover. It runs 3 775pros with two stages of gear reduction. The 1st stage is the WCP 32DP 12T pinion and 100T aluminum gear. I’m worried about the longevity of the aluminum but that’s something to find out. And the 2nd stage is a 20DP 22T to 56T gear.
This results in an overall ratio of 21.21 which at free speed with a 4in Colson we are running 15.4 Ft/sec. So I figure after losses we max will be in the 13s. A bit slow but I wanted to start somewhere on the safer side, and depending on the game acceleration is more useful than top end speed.
The chain in tube setup was done with a 1.5x2x.125in box tubing and 16T vex sprockets, it worked out pretty well. I did end up adding a bit to much to the C-C math so the chain is tight as heck which is why you hear the clicking. But it resulted in a very clean setup.
The frame + electronics + drive train parts all weighted in at 28lbs. Loaded up as seen, it weights in around 145lbs, what kurt vices are good for
Currently we are running very conservatively with the TalonSRX doing a 25amp current limit and an 30v/sec ramp. Definitively need to u the ramp rate as you can feel the lag a bit to much with the controller. But after driving around for 5-7 minutes the motors were barely warm. But a lot more testing will happen. If anyone is interested I will post results as we go.
Magnetic encoders are getting added next week. For vel measurements
The 22 to 56 is a nice ratio since I can swap in other gears for testing down the road.
We have a 3 axis CNC Bridgeport at the school shop that did the frame and some finishing ops. I did the main gearbox housing at a lab i have access to on a combination of a FADAL 15XT and HAAS VF2ss.
That gearbox is a work of art!
Do you know what the run time of the CNC was? (lol)
Will you still have access to it for your in season robot?
I can’t help but mention that the new AndyMark vent spacer was specifically designed to remove the need for the entire second op on the back, but I hesitate to mention them because of how cool it looks with the tooling marks
Also looking forward to your test data. These motors are really the next step in FRC drives.
Aside from being awesome, does the milled-from-a-block gearbox give any performance advantages over a more traditional plate-and-spacer gearbox? Or was the idea just to get a feel for what the new CNC can do?
Also just curious, what was the difference in weight between the block pre- and post-CNC? (i.e. how much metal did the CNC take out)
Really flipping cool, I’ve always wanted to do integrate spacers into our gearboxes but couldn’t justify the machine time cause our machine is just toooooo slow!
there’s about .09 of center drop in the frame. I haven’t traditionally run omnis in WCD but if it’s a defense light game it’s something we will consider.
I mostly just look at how much wattage we are putting into the motor, and what it’s rated for. The 25amps was mostly a back of the napkin guess and something we felt safe with. We are gonna be testing a rage of values.
After I got the program slightly more optimized to run the 2nd one it was under 30 minutes to do the entire main side, and then to do the back side it was 10ish minutes (mainly because i faced the entire thing small tool instead of doing a face mill and then a finishing tool (and honestly just to get the cool circle pattern)). Could have been faster but I didn’t have the balanced tool holders yet to push past 10,000 RPM on the v2fss.
Hopefully? unsure atm I am ordering some new roughing tools so doing feed and speed math, cutting it out on our in house 3 axis cnc bridgeport honestly wouldn’t take that much longer if we go with this during season. Which is an big if.
I saw that in the product released and chucked to myself a bit… I was thinking making something similar to that for this but then realized it would be more fun to billet everythinggggg.
The two main reasons why I wanted to do the billet gearbox was it heavily reduces parts and can be enclosed so that crap doesn’t get into it and grease stays inside. There’s literally no spacers in this entire drivetrain, which are always a bit of pain to make/buy and keep track of. Also to keep the gears happy i can dump a crap load of grease into it and don’t have to worry about it getting anywhere to dirt/etc getting onto the gears.
Also because I wanted to do something cool.
The final housing was only .48lbs finished. The block started off around 4lbs.
Thanks! And thanks for making this stuff possible with the cool things WCP puts out.
Just wondering what are your thoughts on how those 32dp 100T alumium gears will hold up in an DT?
This project is super cool!! Any chance you plan on sharing the cad? What was your decision process for using 3 instead of the common 4 775’s? Could you provide some more documentation on your setup/experience with chain in tube? Also, I’ve been hearing rumblings about steel 32dp gears to address the exact issues you’re talking about. Lets hope those come out before season!
I’m shocked to see a Warlords bot with no carbon fiber but those gearboxes cover the coolness factor anyways.
Extra props on the billet gearbox… Been wanting to do one for some time now… We just don’t have the same sponsorship like we used to and with a router we have gone more flat plate.
Quick question. I couldn’t quite see it in the video…On the cover plate how close was the tolerance? A tight fit? Or did you have alignment pins.? One of the things I never liked in our flat plate gearbox was the torsional stability wasn’t quite where I liked it.
3647 actually ran the exact same thing for our competition robot last year. When we made ours we actually messed up a bit and did the opposite of what Mark did and made the chain c-c too close to the actual one. You are supposed to add about 0.018" to the c-c, while we did the exact c-c, yet despite this, it still lasted the entire season (lots of practice on practice robot and 5 events for competition robot) without breaking and only required the occasional drop of chain lube. Towards the end of the season the chain did sag a bit inside the tube (if you do the proper c-c this won’t be an issue), but there was no danger of it falling off as it was chain in tube.
As for the 2x1.5, we found it really easy to use and route the chain with, although we did have to make a couple pvc spacers as there is extra room when you use the 25 16t sprokets. My favorite part about the 2x1.5 is that you can have 2 rows of 10 holes spaced 0.5" from each side which lets you mount stuff to the tube even though there is chain inside the tube.
I’ll see about cleaning up the CAD and getting it out there. There’s some things that aren’t great about the bellypan setup since it was savaged from an old robot. There’s not much to the chain in tube, C-C + .02in, bearing holes, and either a spacer or collar clamp to hold the 16T sprocket in place.
If the aluminum 32dp gears died to fast I was gonna just order some steel ones and broach/mill them out.
We haven’t made CF sicne our 2016 robot just due to the amount of time and hassle it takes. And with our new CNC mill our machining has been coming out a lot better.
Thanks!
Ah yes, this is one of those things I disliked about a lot of traditional FRC gearboxes. The cover plate that holds the bearings are a tight fit by design, you can tap it in with your hand. The photo in the video of the optical comparator was me doing some maths on this.
Thanks!
Yep, where did you get your 2x1.5? i couldn’t find it at IMS so i just ordered from MCM. Also thanks to 3647 for lending us some sprockets so we didn’t have to wait for our vex order!
We will have a 100T in steel coming out shortly, please use that guy. We’ve had great success using 1-4 motors around an aluminum 32DP cluster in NON drivetrain applications. But we’ve had teams report that in a drivetrain application the steel gear would be much better.
We went just 2 weeks ago to replenish our metal supply and were quite lucky to stumbled upon an 8’ section of it in the scrap. If you can’t find any in the scrap or on the shelf, you’ll have to go to the checkout counter and ask the guy to get you some from the warehouse. Depending on the cashier, they may or may not charge you for a cut as they come in 16’ lengths. Funnily enough though, when we asked for some 2x1 thin wall, they said that they didn’t have any in any aluminum alloy and have never had it even though we insisted that we had gotten tons of it from them in the past.
That’s well past conservative. According to the Vex 4V locked rotor tests, a 775 pro can take 30+ A for five minutes with no airflow apart from convection. You should be perfectly safe running a 30A current limit and skipping the voltage ramp, unless your driver decides to push on a wall for longer than five minutes (and the voltage ramp wouldn’t help in that case anyway).
Edit: and yes, agree with all above about the design and tooling!
We built a very similar enclosed 3-775 gearbox with a 32dp aluminum cluster gear in 2016. After one regional, the inside of the gearbox looked very similar to the picture of your mill after finishing a housing from a solid billet but cast in grease. The cluster gear was shredded to a smooth disk. I would highly recommend using the steel gear.
