1-1/4" x 6" x 20" Aluminum Billet: $25
Several hours of machine time: it was free to me?
Not ever having to worry about losing a standoff when swapping out a fried FP motor between elimination matches: Priceless
1-1/4" x 6" x 20" Aluminum Billet: $25
That billet of aluminum for $25? Where was it that cheap? They look amazing, and resemble some old KOP gearboxes we have sitting around.
The aluminum was from Yarde Metals “Drop Zone”.
This is pretty much the only place I’ll buy metal from now, since I hardly ever need full size sheets, and their prices are usually pretty competitive. Also, unlike some other online surplus metal stores, their stuff is usually free of nicks and gouges.
Their prices look good, but $18 handling fee if you don’t order $75 worth? that is kinda steep. Ive used online metal supply, and fortal on ebay, and fully recommend both!
NOTE:
Online metal supply is not Online metals, the latter I find to be very high, 2x the price on some titanium I needed vs Online metal supply
http://www.onlinemetalsupply.com/
Online Metal Supply is good, but their quantities tend to be more limited in the aluminum department.
As for the $18 fee, I just avoid it by using it as an excuse to buy more 1/8" and 1/4" aluminum plate until I reach $75.
Very nice gearboxes, compact and simple. I’m a little confused though… Isn’t your robot finished already? Are these upgrade parts, or perhaps for another team?
A robot is never truly finished - there’s always something that can be upgraded, something that to be designed better, faster, stronger. 
In this case, these transmissions are going to replace two Globe motors currently powering our polycord elevator which brings balls up to the shooter. Don’t get me wrong, the Globe motor powered elevator works just fine right now without any problems, but we can only fire about 2.5 balls/sec right now. With these new transmissions powered by a single FP motor, we’ll be able to get 8-10 balls per second to our shooter, with a net change in robot weight of -0.9 pounds.
The three gearboxes include one for competition robot, one for practice robot, and one for a spare that will probably be assembled with a Lexan (as opposed to aluminum) cover to have in the pits to show people. The gears used in the transmission are from the first two stages in the KoP FP gearbox. I couldn’t think of an elegant way to mount one of the KoP FP transmissions while outputting through the second cluster shaft with a cantilevered load, hence the design of this.
Previous revision: (on almost completed practice robot)
http://www.team228.org/gallery/103/slideshow/build-season-week-six_bc812-15eb1.jpg](http://www.team228.org/media/pictures/view/4956)
New revision:
http://img26.imageshack.us/img26/2698/newgearboxdesign.png
And the outer plate (the pockets are on the inside of final assembly)
http://img23.imageshack.us/img23/4996/newgearboxouterplate.png
Great Job Art, really innovative. Thinking about making some plastic version of these.
What was the run time for one of those?
Just curious.
What’s the weight difference to an optimized design with standoffs?
I imagine at this point it doesn’t matter, and the reduced number of parts outweighs it as you were already underweight (iirc).
Sadly, no where near as fast as I originally wanted. I had to run these on a TM-1, as the VM-3 and VF-4 were both already being used. It ended up being a little over an hour for each one on the TM-1, mainly limited by me manually slowing down the program to 4000rpm and 40ipm to avoid throwing chips and coolant everywhere (which it still did). Also, if I was going to do this again I’d optimize the CAM toolpaths, as my software added some dumb things in there that just wasted about 15 minutes doing nearly nothing. I would have loved to do this on the VF-4 at well into a triple digit ipm feed rate, with much faster rapids, and nearly instantaneous tool changes.
Edit: Also, if I make these again I won’t be as lazy. The only two three end mills I used were a 1/2" flat carbide, 1/2" ball carbide, and 3/8" chamfer. It would have been nice to also add in there a 1" carbide flat end mill to hog out the pockets faster, but that a: would have been pretty scary at full force in a TM-1, and b: I would have had to walk across campus to the other machine shop to borrow another tool holder. 
The unibody transmission bodies shown here weigh 5.44 ounces. Before I designed this, I did a plate and standoff design using 1/8" plate and 3/8" aluminum hex standoffs, which actually ended up being 1.6 ounces lighter. However, when I added 1/16" polycarbonate side shields to it along with a few #4/40 screws to hold it in place, the difference was essentially nothing.
The side shields were somewhat necessary in that this transmission is mounted right near one of the handles added onto to the robot to act as a safe lifting point, and also only a few inches away from the cRIO. Adding an ounce or two of shields is way easier than filling out the mountains of paperwork for that one impatient person to stick their arm into the robot reaching for who knows what only to have their fingers meet up close and personal with a gearbox.
First, let me say your gearbox is really cool! If we had that kind of machining capabilities in-house we’d probably do the same. Since we don’t, and value our sponsor’s precious machine time, this is what we came up with (which is why I asked this) - not sure you’d consider it elegant, but it sure worked
http://www.team383.com.br/temp/imgs%20ChiefDelphi/coroa%20FP.jpg
The last stage of the gearbox was removed, the case was reduced to acomodate only two stages and dremeled to pass the chain through. That way, the shaft which supports the second stage gears (and now this sprocket as well) is fully supported, not cantilevered.