A late-night drivetrain brainstorm
 

Yes, this means I've been thinking again.

This time, the thought has been around the DeWalt XRP transmissions, namely the fact that you can only stick one CIM into each. And I'm trying to figure out how to make that either less of an annoyance, or even a good thing.

Imagine a kitbot frame. Now imagine a CIM/XRP combo at each corner, driving wheels via sprocket reduction.

Now, this would seem to have the advantages of having less chain running about, and a degree of robustness to it (if one of the CIMs releases the magic smoke, a switch could theoretically cut the other motor to allow you to drive 2WD for the rest of the match). Of course, you're still limited by some of those pesky aspects of 4WD, such as turning issues if you don't build it right.

It makes a sense in my head. However, you are dealing with a business student. Am I remotely close on this?

Re: A late-night drivetrain brainstorm
 

You mean drive each wheel with one transmission/motor combination right?

If so, if two motors do burn out, then the other remaining motors need to back drive the dead ones, which won't be so easy since they're going to be geared quite a lot. Even if you open the circuit, the resistance given by the motor plus the gear/chain reductions will put a whole lot more load on the remaining motors.

Correct me if I'm wrong though.

Re: A late-night drivetrain brainstorm
 

It's not that bad of an Idea. But you pointed out that you are going away from the hassle/weight of chain/sprockets. Now let's think about it here for a minute. adding two XRP/transmission/motor, you are adding weight. Losing 4 ft. of chain and 6 sprockets (there are other configurations), you are losing weight. Will the gain and loss of weight balance out? Question is the what are the pros and cons? Noodleknight does have a point. At the same time, it can be done. This is when personal preference comes in. I rather stick with one transmission per side and run a chain from the output to the wheel unless I was to do a swerve drive.

Re: A late-night drivetrain brainstorm
 

Amazing how great minds think alike! ;)

What you describe is almost exactly what 1213 did for '05.
Cad here:
RealDeal Here:


We did not use chains. Direct drive with a 4 inch wheel produced a zippy 7-1/2 FPS The drive platforms have 60+ matches on them and mucho hours of practice without a single malfunction - other than a drive shaft snapped in practice right where it's machined down from 1/2" to the 0.3 x 0.25 double-D to mate with the tranny output. That was solved way before the first competition by heat treating them.

You're right about the turning issue. We solved that by moving the front wheels 5" back to shorten the wheel base, and by using 95A durometer wheels. Wheelie bars 1/4" off the floor at the corners kept it from tipping.

The design worked just fine for Triple Play. OTOH, it would never work for a game like Stack Attack - with the ramp and all that pushing.

Re: A late-night drivetrain brainstorm
 

This is true. However, doesn't the XRP transmission have the ability to shift to a point between gears? If that causes the wheel to freewheel (which would make sense--you're not engaging any gears, correct?), that would effectively stop the backdriving, yes?

Re: A late-night drivetrain brainstorm
 

I'm wondering if this even needs to be considered. What are the chances that the motors will burn out anyway? Much less in a critical match. Barring Murphy's Law, I don't see much need to expect a motor failure in a FIRST robot. Now if you'll excuse me, I'm scheduled for an electrocution for questioning Murphy's Law...

Now on a side note, I've been trying to figure out how to get 4 separate motors without a significant weight increase. Thanks. I've been working on an omnidrive system similar to this. My wheels and motors are on the way right now. The little version is cheap, the FIRST scale version could get expensive. In the mean time, I've been working on a speed controller. In my frugalness, I've decided to spend $30 and 2 weeks to design and build something rather than $200 to just buy it. It's able to read a pot, now all I need to do is figure out how to get it to read hobby PWM. Then after a trip to Radio Shack for boards and a few hours of soldering, I'll have 4 little speed controllers for my project. Sorry about getting a little off topic there...

Re: A late-night drivetrain brainstorm
 

Just to twist your mind a little bit more.

Mount a Mechanum wheel on each gearbox and you have a three-speed holonomic drive system.

Sweeeeeeeet.

Re: A late-night drivetrain brainstorm
 

One thing to consider when doing FWD with 1 motor per wheel rather then 2 motors + 1 chain per two wheels is that in the 1 motor per 1 wheel case you are depending on both wheels being on the ground to get both motors to do useful work.

If, for example, someone rocks your robot back on its rear wheels, you will not be able to get the benefit of having the motor driving the wheel that is in the air. This is not the case when you have 2 motor + 1 chain driving 2 wheels, in this case, the power is automatically transferred to the wheel that remains on the ground.

Is this a definitive answer. No, of course not. There are 100's or reasons to go this way or that with a FIRST robot design. You have to balance what you save in weight, complexity, and (often times more important) package space against the things you like but give up.

Welcome to the imperfect world of design!

Joe J.