Team 1658 flying inverse differential swerve Video+CAD files
 

As previously seen here: http://www.chiefdelphi.com/forums/sh...d.php?t=140181

Here is a video of a quick test run we did. We don't have full functionality yet but serves well for initial demonstration purposes of this system.

https://www.youtube.com/watch?v=WfxK...ature=youtu.be

Here are CAD(autodesk inventor) files of the system. I'm fairly confident all the parts and dimensions are accurate but as a disclaimer: Accuracy is not guaranteed.

https://drive.google.com/file/d/0B4f...ew?usp=sharing

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

I still love this idea. The well-controlled slow driving would have been helpful in 2015 to not knock over stacks when placing your stack. With some pretty basic (or not so basic if you choose) programming, you could make an 'auto-shifter' from differential mode at low joystick input values for better control to normal mode at higher joystick values for faster driving. One less thing to for the driver to worry about. I look forward to hearing more about this project!

BTW anyway we can get the CAD either in STEP format or on GrabCAD so Solidworks users can take a look?

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

On the higher values differential mode is kind of like normal mode since the motors are driving in the same direction, just not same speed. But at full tilt they are the same so our high end is more or less identical.

As far as the CAD goes, I don't have inventor on my home pc so I wouldn't be able to do any conversion until after the season starts but pretty much all the CAD software packages have the ability to convert files from other formats so i'm sure you'll find a way to import the files into your cad package or at worst use an online converter.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

Interesting, on the video it looked like it went faster in normal mode. Maybe the driver just wasn't pushing the stick as far on differential mode.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

Ye it was the first time that driver picked up the controls. It can go both faster and slower in a more controller manor but he wasn't familiar enough with it yet to properly show what it's capable of doing. I had a bit more practice with it and could have showed off the various aspects a bit better for comparison but I had to do the recording so can't do both xD we attempted to have someone else record while I drove prior to this but i'm pretty sure that video would make a lot of people puke.... so I ended up having to take the video myself. Certainly not a shining achievement from what can be seen in this video but I kind of like that too so others don't get a full proper demonstration and while it can be seen working the finer details are left to their own imaginations =p Until competition that is then this system will shine in all it's glory xD (if we end up deciding to use it anyway) or they can attempt making one themselves to find out.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

Can you guys tested pushing stuff with it yet?

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

Why are motors making so much noise even when it's not moving? Feels like a car on neutral :yikes:

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

We did some small tests but no video of it. Nothing is properly setup(the robot is 80 lbs short so traction is very low, we're using clamps to hold some parts on, the chain doesn't have proper guides/tensioners, need more time to refine and implement more code, etc...) so there would be little point in even attempting it. This prototype was a proof of concept and it succeeded at that task for our needs. Real testing will be after we see what the game is to see if we want to continue on this route at which point we will attempt a competition rated prototype.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

when it's not moving it doesn't make any noise. However when we go at low speeds in Differential mode then indeed it is a noisy machine. That's because the top motors are going full speed and the bottom motors are going nearly full speed the opposite direction. So there are a few parts spinning at very high speeds. That's kind of the whole point of the differential system, we can have high power at very low speeds.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

Have you thought about trying to optimize this so that the CIM is running at the RPM at which it has the highest power? Or does running them like you describe automatically result in that?

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

it all depends how you want to run the cims. There are various modes of operation which is one of the nice things about the design. We can certainly chose what reference speed the cims run at for varied effects such as most efficiency, or most available power(all out) or anything in between. We will certainly be playing with various methods but as far as which ones we end up using,who knows.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

Is it a correct statement that it is impossible to stall the top motor? If so it seems like it'd be a good candidate for a lighter and more efficient open-air motor (new 775, e.g.) so long it's geared correctly.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

If the top motors are going "full speed" (which I read as free speed, or nearly so), they are generating no torque and thus producing no mechanical power. The bottom motors are going "nearly full speed", they are generating low torque, and are operating well down on the power curve. Where does this "high power" originate?

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

I just want to double check, the transmission you are using is similar to this, correct?

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

you can still stall out the motors and while it would work to a degree the total capable speed and power is still reliant on both of the motors so if you used a 775 you will lose in both regards compared to a cim.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

I'm curious about that worm gear that you've got. It looks 3D printed - is that right? Can you give us some specs and tell us how you made it? Does it seem like it will be sturdy enough to hold up well?

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

In that situation there wouldn't be much as you mentioned. If we drive the system like that and we try to move, as we increase the load on the system the speeds of both motors will be loaded down and slowed down until it hits a point where it has sufficient torque to move. Or we can try to keep it in the higher power band but that would arbitrarily limit our top speed when that amount of torque isn't required. With a desired output speed, the faster we run the motors the faster total speed differential we can get making the top speed of the robot higher. But indeed there isn't much power available there but as in a regular control system when you apply a load the speed will be reduced to the point where torque is high enough for the task. In our case both motors will be slowed down so our torque will vary while our total speed output remains the same.

Or at least that's what we think, could be wrong. The whole point of building the prototype was/is to find out if we're just kind of wrong or completely wrong XD

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

they are NOT 3d printer. They are plastic though(nylon?)they are actually very common garage door opener gear sets you can buy at sears. Though they are twice as cheap online. In fact, andymark uses the exact same set(but they hex broach theirs) on one of their 90 degree transmissions. We used two sets last year to drive our elevator mechanism and they held up just fine.

As far as specs/how it was made take a look at the CAD files.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

Maybe? From what he said it indeed sounds like he just made an overly complicated inverse differential but he seems to hide the fact that he has another motor somewhere counter driving one of the gears and doesn't mention that both motors need to be able to provide the torque you want out of the system. This seems like one of those try to scam investors out of money schemes but since he seems to be hiding information can't really say.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

I really like the idea behind this - I think it may have applications beyond just drive trains.

Something you may want to look into is gearing the upper and lower motors at different speeds, with a non-backdrivable reduction for the high speed motor. This would let you get a wider range of gear ratios without having to run motors at really high speeds.

This paper, if it is ever released, describes the exact same idea in a planetary differential with some equations/control ideas that might be useful.

https://ras.papercept.net/conference...=103&Number=70

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

Second. Would love to take a look.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

isn't solidworks able to import inventor files? inventor can import solidwork files it would be odd if the opposite isn't also true. As mentioned in the other reply I don't have inventor at home nor will have access to the cad pc until after the season starts so I can't do much on this end but i'm sure you can find converters online at worst case but i'm fairly certain solidworks should allow you to be able to import inventor formats.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

It depends on the version from my experience. Where is the Inventor CAD? I want to try importing it, so I'll post back in a second.

EDIT: Nope, turns out we need "Inventor 11" to open inventor assemblies. https://gyazo.com/085d9a7b724d8e8d3cae02e00aaa5ca8

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

what about the .ipt files or the drawings? ye not sure how assemblies would be handled even if they were exported, i guess the entire assembly would just become 1 part? only thing i can suggest is maybe look for an online converter there's nothing I can do until season starts.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

Do you have access to Inventor? If so, you can open it and just use Save As to save the assembly as a STEP file.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

You can just upload the files to GrabCAD and we will be able to look at them online without any software at all.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

So this is gonna be a bit noobish, but reading through the thread and watching some videos, I still can't grasp the idea behind the inverse differential. I understand that 2 motors run independently and control the output velocity, but how do they manage to do that? And did you make or buy those parts?

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

A differential gear system is standard equipment on automobiles, easiest to understand on old rear-wheel drives. A drum is turned in the direction that the wheels should turn as an average (using bevel and crown gears in most cases), and another bevel gear is mounted to the inside of the drum to turn the wheel axles. If the wheels both move at the same speed as the drum, this later bevel gear does not rotate about its axis. If one of the wheels has to move a bit faster (e.g. in a turn), the bevel gear will rotate to let one move a bit faster than the other. The extreme case is when you're stuck in the mud and one wheel spins freely while the other does not turn at all. The net result is that average speed of the wheels must equal the speed of the drum. Inverse differential is if you run the differential backwards - put motors where the wheels normally are and use the common shaft as an output.

I'm not sure what flying means in this context.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

I have rebuilt more differentials than I want to think about and taught automotive for a handful of years. I had a tough time following the last description, so I'll point you to an old but good video.

https://youtu.be/K4JhruinbWc

Start around 3:20 for how the differential works, then just replace wheels with motors and the concept should be pretty clear.

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

flying in the sense that at top speed they are moving REALLY fast... it's rather scary especially when we had it mounted to a bench spinning at top speed unloaded....

Re: Team 1658 flying inverse differential swerve Video+CAD files
 

watch the video mrbasse posted.

the gears and sprockets we bought, everything else we made.