I noticed this year that the teams who did well didn’t have a very fancy design or anything they had a robot that drove well and had good speed and torque. I thought about it and decided that the best way to maximize traction and driveabiliy is to have all four wheels spinning at the same rate. This means four seperate gearboxes and a shaft encoder on each wheel. Then a program that keeps thmspinning the same speed. I was wondering if anyone else has had any experience with a system like this and if it woked or didn’t work ect. 
This year we used a simple system with the drill motors using the FIRST supplied gearbox, and made I made a grearbox for the CIM motors that had a relatively simular output speed. All we had to do in the end was chain the two ouput shafts together, and it made a simple and reliable system that satisfied basically all we needed/wanted it to do.
We place the CIM motors on the back wheels, and the Drill motors on the front wheels, thus making out 4 wheel drive.
Simple is good.
how about just 4 wheel drive? have one or two motors for each side, run a chain of that spoket to the front whhels and on another sproket to another whell in the back… ill look for photos… o i rememebr www.gosparx.org 2004 photos look and you can see what i am tlaking about.
Wheels driven by chain or gears are spinning at the same rate.
found one laready… go to picture galleries,2004,regionals,buckeye… you can find many of my bot and see the chians good…
If you are going to atlanta, seek out team 138. We’ll be happy to help provided we arent in a rush to get to a match
For simplicity, like its been said, you can just link each side with chain and sprockets or gears, etc
For driveability, i would use a gyro to give you steering feedback and have it adjust ur drives(not that i have any experience with them, but that seems to be the general trend around here)
For traction, maybe you could use a non powered wheel with an encoder on each driving side. That way, if a wheel slips, you can slow it down to the speed of the non powered wheel, (or just above it, depending) until the two speeds match, then you know that you are getting grip, and are good to go. Let me know what you think
-Josh
My team had a 4- wheel drive system that involved linking the front tires with the rear ones with chains. I did notice, though that when the robot went to turn, it “hopped” like a 4-wheel drive truck on a dry day. I myself was going to post a qeustion that asked if anyone had a somution without loosing to much traction?
This is not because of the chains linking. It is because the wheelbase is longer than the track width is wide and there is too much traction.
We’ve developed an excellent solution for this. I’m fairly sure it’s not an original idea, however I don’t know who to credit it too.
Our drive train consists of 4 drive modules. Two are powered by the Chips, and two are powered by the drills. The motors are mounted horizontally, making their polarity/wheel direction the same. Worm gears are used to transfer the power horizontally to vertically (to the wheel). With a little programming and gearing the motors spin at about the same RPM.
Even with using the sticky pneumatic tires we’ve never blown a fuse. We also rarely lost a pushing match. Our only problem is that we had turning problems, IE: hopping. I think we’re going to solve that via some omni-wheels. With some tweaks in the off-season, we are going to make this our standard drive train.
We’ve tried using 1 motor on each side to get 4 wheel drive, and we’ve never had a successful drive train. What we used was the drill motor in the middle, and had a long chain running from the front sprocket to the back sprocket and tensioners on both sides of the drill motor.
The problem with that is: since your not running the chain from front to rear, you don’t have enough wrap around your sprocket. An ideal drive train probably has around 140-150 degrees of wrap. Because our drive train did not have enough wrap, the sprocket and chain would engage well and the sprocket would always slip when we drive, and it would make that weird buckling noise because of the slippage.
We have since tried 2 wheel drive and it works better than before… next year we’ll go back two 4 wheel drive but maybe this time with a double sprocket or 4 motors?
As far as I can tell, the best solutions to your problem are to either use a double sprocket on the output shaft of your gearbox, as you said, or to snake the chain around the output shaft using idle sprockets. In 2001, team 254 did this. The picture should be nicely indicative of what I mean.
You could use idle sprockets to guide the chain but you could also use a rounded piece of HDPE or other low friction material bolted to the frame to save weight and machining/shipping time. I don’t know if this has been done in FIRST but I remember it on some battlebots… if you need a picture I could find one.
EDIT:
picture
close up
detatched
this isn’t exactly what I was talking about but you see how the yellow plastic bit is being used.
The past few years, we have used a small adjustable sprocket both as an idler to assure 180 degrees of wrap on all driven or driving sprocket, and as a chain tensioner. It works quite well.
You can see a picture of our 2k3 design here (this picture was taken during repairs, so there is slack in the chain). The bolt on the right hand side of the trapeziodal piece is what moves the idler up and down, and the sprocket at the bottom is connected to the motor. The is a stationary idler on the other side near the wheel that isn’t quite visible.
Also, by placing the idlers above the driven sprocket on one side and below the driven sprocket on the other side, we could have both our motors running the same direction (which is much more efficient than the worm gear solution posted above, since worm gear efficiency is around 70% at best, while sprocket efficiency is around 97%).
The problem with the 2 wheels move together idea is that to have good tunability your front wheels have to have less traction than the back wheels and slide when u turn. This means that your traction limit is lessened in the 2 wheels move together idea. WHat I really menat was to have each wheel move independantly from eachother so each wheel is at it’s traction limit. Anyone know what I’m talkin about.
Any help appreciatedI think the problem may lie in the pneumatic tires. We have a very simular 4wd drive system on out last years robot, and when we tested those tires on it, it jumped, and had horrible turning ability. I’m not sure exactly what we have…but we put some 9 inch wheels on, and that solved the problem.
Whoa! I may just be the programmer, but I know a complex drive train when I see one (besides, all our guys are seniors but me and Dan).
We just had a CIM and a drill for each side (CIMs facing opposite, drills same). They were linked together via chain and connect to both wheels directly. All chain. in the end: 2 motors, 4 axles, 4 chains, and 2 wheels. The only problem was that we kept breaking plastic gears (6 of them in all). The steel ones we had we’re just too heavy.
Also, by placing the idlers above the driven sprocket on one side and below the driven sprocket on the other side, we could have both our motors running the same direction (which is much more efficient than the worm gear solution posted above, since worm gear efficiency is around 70% at best, while sprocket efficiency is around 97%).
With worm gears the efficiency really depends on the maker (if commercially purchased or not…we used a Boston Gear worm reducers this year and the had about 80% efficiency. What would work best if you can find them is planetary gear sets. They have about a 97%-99% efficiency (depending on what you get).
You should hear Biff about last year’s drive train. Drill motors through the drill gearbox and out a pair of worms TO EACH OTHER. All said and done, we had 100 mechanical watts. It did not drive well.
.Are you talking about “crab” or “swerve” drive? There are drawings and info under those headings. And some teams that have people posting here can answer as to the complexity of the mech and programing. Most that I’ve seen have modules that swivel, truned with one chain per side. When looking at a swerve drive at Grand Valley, the team (sorry I forget their number) liked it but said it was a little over geared and tended to eat Fisher Price motors. I made a suggestion to put a spring loaded Z in the chain, on the outside of each “turning chain” and a piston or something on the inside so they could get the modules to form a O and do a tank style turn with out having to have the masses unballanced or be really geared down. I hope someday to be able to work with Students that have the machining (sp) skills to attempt a crab or swerve, for now I planning to stick with 4 wheel dirve and over power it to be able to turn.