Here’s something I’ve been thinking about for a little while.
After seeing several teams use 775 pros for their drivetrain this year with varying degrees of success I threw this together.
This is still missing all the hex bearings and sprockets, live axles, and collars.
I drew this up with the intent of making it easy and quick to manufacture for most people, my personal team has a plasma machine that could easily cut the plates in a few minutes.
it can also be made pretty easy on a mill or drill press if you wanted.
cantilevered “west coast drive”.
3 775 pros in 15:1 versa-planetaries.
sprocket reduction from 16t-18t between vp’s and live axles.
setup for 4" wheels, can be modified for 6"
the standoffs are 1.4" right now but the intent is to support the end of the vp shaft in a bearing, as well as support the live wheel axles in two places. so that spacing will have to be adjusted slightly once i get all the sprockets and stuff in the drawings.
I’m pretty confident the vp’s can take the load like this.
Assuming that you have a center-dropped 6wd, won’t only 2 motors be engaging the carpet on each side? That seems like a big waste in power, and with only 4 of these motors powering the entire movement of the robot, I would be afraid of the motors stalling and burning out. With 4 775pros driving the robot with 4" wheels you would have 75ish amps being drawn :ahh:. (also some of my math might not be fully right as I’m just plugging numbers into JVN’s calculator. I still don’t think having what essentially is 4 775pros driving the robot would end up too nicely).
Overall I like the idea of having this simple way of powering wheels! Everyone loves versaplanetaries!
the center wheel is not dropped, I planned on omnis in one corner
Also I saw one specific example at champs where a team had 4 775pros in their swerve drive(which i imagine is basically the most stressful use of the motors) and they managed to move a full weight robot pretty easily.
EDIT: all the wheels will be chained together, sorry if that wasnt clear
That makes a LOT more sense! I think it would still have to be traction limited even if all 3 motors are powering each side of the drivetrain. It seems like most teams who went with a 775pro drivetrain this year ended up with 4 per side so they wouldn’t burn out. This design would be interesting to see come together because of how easy it is compared to designing and machining a custom gearbox!
First, why bother making a chain reduction WCD? The shaft from the VersaPlanetary should be long enough that you could easily change the VP reduction a bit and direct drive a 1in wheel with it, which would both remove a point of failure (chain) and make you wheels non-cantilevered, and thus more ridged, and substantially simplify the build process (it would also make for a lighter, more compact drive system with easier bumper mounting).
Secondly, does the drive have a drop-center? If so, have you considered that at least two of your six motors will essentially be free-spinning 100% of the time? If it’s not drop-centered, is the intent to use omni-wheels in the front and/or back?
Beyond that it looks like a solid drive, though personally I’m terrified of using a 775Pro on a drive for fear it might burn out. CIMs may be heavy but in 12 years of FRC I’ve never seen one burn out, and your drive system is the one thing on the robot that MUST work EVERY match. This does at least seem like it would have some redundancy in the sense that if one motor burned out the other two would still run (assuming no drop-center), but the additional load might also damage the remaining motors and cause you even more issues. The small pinion gears of the VP gearboxes also scare me a little. I feel like if a team were to go this route they would be checking their gearboxes for wear between each event (or even more frequently). You also have to be VERY careful with your gear reductions to make sure the drive is traction limited, cause if it’s not, you’ll be seeing a lot of magic smoke.
By the way, were you intending to run these on 40A breakers, 30A breakers, or some combination of both?
Holy ninja’d by like 3 people. That’s what I get for creating the thread and then taking too long to reply… >_>
the chain is needed to achieve the ~18:1 reduction that I wanted. and to chain all three wheels on each side together for encoder purposes
direct driving a wheel with the vp’s I thought would not be a good idea as I have found them to fail more easily on a direct drive system.
Especially if the wheels are going to be unprotected on the exterior of the frame perimeter with only bumpers to protect them.
1" wheels??? that would be something i would want someone else to try first lol
no drop center, going with omni’s in the corner as i want to integrate this idea with a “perpendicular drop drive”-“slide drive”-“Hdrive”
6.the maintenance with this is rather simple, just remove the wheels and take the nut off the standoffs and remove exterior plate if you need to change a vp out, also the new v2 versa-planetaries allow you to remove the motor from the back if it burns out.
anecdote: I saw 16 burn out their 775’s on the swerve drive they had a couple times at houston champs but they were only using 4 on the whole machine as far as i could tell.
The 18:1 reduction could easily be achieved with a 6:1 and a 3:1 reduction in the gearbox itself, no need for sprocket reduction. The encoder part is a fair point, though you could use integrated encoders on the gearboxes or an idle drag-caster (which might be ideal since you’re considering using it for an omnidirectional drive anyways).
My suggestion was to NOT have the wheels on the exterior of the frame, but between the two plates of the frame, that way it’s both protected and supported (not cantilevered).
1", as in 1" wide wheels, I should have been more specific. The idea was with narrow enough wheels you could fit the wheels between the frame plates rather than having to mount them externally (see #2). Obviously no one is going to run an FRC robot on 1" diameter wheels. :rolleyes:
Interesting concept, using these motors+gearboxes certainly gives you the real-estate to do that too.
Just watch your current draws, 775s will not last a full FRC match at a constant 40A draw according to Vex Motor Testing.
Simple, perhaps, but you have to make sure the rest of the robot allows everything to be easily accessible. I personally would rather have a drive system I do zero maintenance on all season over one that I have to but it’s easy to do. There’s enough other things that can go wrong on a robot that I don’t need the drive to be one of them.
Bomb Squad only used four 775pro motors for traction. There are others elsewhere on the robot. They are limited by the number of slip ring connections available in their latest swerve module.
A better reference for 775pro motors in a tank drive system would be PWNAGE. Their great idea is to ‘derate’ the 12V motors to a smaller maximum voltage, so they don’t smoke on stall. Their resulting drive train never gets the motors warm.
1.They don’t sell a 6:1 or a 2:1 for the versaplanterary at this time so the 18:1 is impossible to achieve in two stages. Hence the need for the chain, I could go with a larger reduction but then we are starting to get too slow of an fps than I desired.
Also minor corrections the chain reduction is an 18 tooth on the VP and a 22 tooth on the axle.
2-3 oh ok that makes sense lol.
5-6. We’ll have to see. I don’t think it will be pulling all 40 Amps from each slot during a match but it is going to depend on the game I guess, also there are some other electrical tricks we can do to prevent brownout and burnouts
But you do have the ability to make the drive plates shown in the OP, right?
So you could use the main idea of PWNAGE’s tank drive – running several 775pro motors at reduced voltage – without making a custom gearbox as complex as theirs. They ended up using four 775pro motors per side, but I think three would give most of the benefit. You could do that with Versaplanetaries, and chains or belts, using the same construction method as your picture in the OP.
That is an interesting design. Are you going to build it as an off-season project? What are the anticipated torque loads on the shafts? It will be interesting to see if the VP’s hold up to the shock-loads.
In addition, I would suggest adding several plates oriented horizontally and several vertically to turn this into more of a rigid box structure to resist the forces that will be applied to the shafts by the wheels and to the side plates from robot to robot contact. These plates can replace some or all of the standoffs. Standoffs do not resist racking forces very well.
If you don’t feel that the plasma cutter is accurate enough for making a gear box, it is not accurate enough to make your side plates since you need similar precision. It may be faster overall to cut it all on your mill, when you take into account the time to make the parts fit and work. As designed now, there are 11 holes that must line up fairly precisely. Otherwise, you will have difficulty assembling it all or there will be binding in your shafts.
Your design shares a lot of features with the octanum modules that Discobots made in 2014. They were absolutely miserable to assemble and repair afterward.
It would be best to consider ways to swap out your VP’s without having to disassemble one side of your robot. There will probably be bumper mountings and other structures attached to the side outer plates. In cases like this, it may make sense to add a part such as an intermediate mounting plate for the VP that then screws into one of your side plates. Do not rely on using hex cap screws to hold your VP’s on and then using an L-shaped Allen key to remove the screw. It will take a long time to get the screws out and longer to get them back in, assuming you have not installed something above that blocks access to the screws.
Maintainability is something that is designed in from the beginning. It will be next to impossible to put back into your design in the 20 minutes you have between matches.
Interesting ideas, im going to add the VP pattern to the outside plate so we can use a long t-handle Allen key to remove the vp mounting screws.
The plasma machine is accurate enough to do the bearing holes, I just wouldn’t want to do anything that requires gears to be meshed correctly.
Also with the vp’s being supported at both ends of the shaft im not too worried about shock loads assuming we get a relatively flat floor next year, and seeing how 1296 used them in 2016 with some slight modification I imagine the gearboxes should hold up if they survived stronghold.
I am thinking about doing this in the offseason as we already have alot of the parts we need in our shop, and we want to play with an h-drive like 624 has been doing for awhile.
Right now the only thing I’m worried about is maybe our gear ratio is too low. But that is easy to change.
As for the standoffs, I’ll have to see what the stiffness is like once we put it together but there are easy ways to fix that with a super structure or cross beams. Maybe I’ll put in a c-channel end cap.
It would make sense to use the same pattern for the inner and outer plates to simplify manufacturing. You can drill out these holes in the outer plate afterward to clear the Allen key. You would still have to remove a wheel but that is much better than having to take the chassis apart.
You will still want the bearing holes on the inner and outer plates to line up well for the shaft to fit through properly. This is where it might be better to cut the four plates and screw them to each other and drill/cut all the holes in your mill.
I was more worried about the torque shock-loads on the gearboxes. In a more typical multi-motor gearbox, the gears are larger than those in the VP’s. If you hit one of the wheels, all the shock force will go into the gears of just one of the gearboxes, not all 3 on one side. Whether it will or will not survive can really only be determined by building it and beating on it.
It occurred to me after making my first post in this thread that using a piece of C-channel that is commercially available would make this much easier. Cutting a piece of C-channel will probably take less time than cutting two pieces of churro accurately and trying to get the ends square and will most likely give you better results.
How does the space used by the motors and VP’s compare to a typical bunch of CIM’s and gearbox? How much space is left between them once you add the other side? This might be a good candidate for a bottom mount design where the PDP and motor controllers are installed and accessed from the bottom. This would give you a flat top for installing scoring mechanisms.
I’m seeing this late. I just wanted to close this up.
We use 4 775s for drive motors (one on each module). We lost 2 of these motors in our practice match…I believe they were already damaged from issues at Midwest that caused them to bind and eventually fail. We didn’t lose any during qualification matches.
In QF3-2, we lost a master link from one of our steering chains. This caused the drive motor on that module to fail and damage the other three drive motors. We weren’t able to replace that drive motor before SF2-1, so we ran with three damaged 775s, all of which eventually failed (the field smelled terrible :D)
We replaced all 4 drive motors between SF2-1 and SF2-2.
They are a great drive motor, but are definitely more fragile than a CIM.