I have seen several posts warning rookie teams not to use treads and use four, tank-driven wheels instead because treads are harder to design, debug, etc. What are some problems that you have encountered using treads in the past?
Right now, we are considering using 1.5 inch wide, double sided timing belts as treads, possibly self-tracking if they are made. Is this a good idea?
There slower, they drain more power, and they tend to track of corse when you try to drive it in a stright line.
I’m going to make a suggestion and say that a DOUBLE SIDED timing belt has a LOWER coefficient of friction, meaning that there is little resistance to movement. As for using a SINGLE sided timing belt (Can’t remember the company that sells these, but…) they type they sell actually has a HIGHER coefficient of friction. It is also EXTREMELY hard to find double sided belts. (I should know, our team is kida looking for them but may go with single sided).
The gearing…he’s right about that part.
Last year we used treads on TOBOR… yea… if you use treads make sure that they WILL NOT FALL OFF!!! We had problems with them falling off numerous times last year- it was not cool!
Treads can be a sticky subject and they’re not easily tackled, especially by rookies, but here are a few suggestion. Doubled sided timing belt is a very powerful way to build a quick set of treads. Pre made gears and cogs for these types of belts are readily available. I’d personally suggest looking at a company called Breco-Flex. They make a profiled belt which will more or less guarantee that they won’t track themselves off. This will most likely be your main problem. When you turn, especially at high speeds the belts will create a tremendous side load on themselves. This can easily walk them off your drive gear or walk them over your side belt guards. This will take some time to perfect.
Secondly…and by far the most common made mistake in my opinion is to try and have the belt make contact along it’s entire surface. Trying to skid stear a belt along a 36" length will most definately cause failure. Your best bet is to put an idler pully somewhere near the balance point of your drivetrain that is slighter lower than the others. If effect your machine will end up rocking onto one side of the treads while moving, and when turning will balance on the single idler creating much less friction and making turning easier.
Take a look at the gallery on this forum for pictures of bots with treads…Truck Town Terror’s robot from 1999 is a great example of how to do the job well, as is team #1’s machine from that same year. Good luck.
From the pain and agony of belts that my team suffered through last year, I would say that the belt is good. However, consider the following:
- What size pulleys and idlers are you going to be using? IIRC, the diameter of the pulley/idler has to be so much larger than the width of the belt.
- If you are going to use belts, order soon! Breco-Flex (as suggested above) takes 2 weeks to make belts and 1 week to make pulleys.
- Again, Breco-Flex isn’t cheap. Our belts (double sided, forgot the length, but it was BrecoFlex’s minimum) cost $147 ea and pulleys cost $67. The special machined ones we wanted costed $97 ea. Unless you have an ‘expert machinist’ handy, don’t attempt to make the pulleys yourself.
- You need more time to test a belt system. Leave at least one week for carpet tests. We didn’t have time last year to test on carpet, which, as previously stated, added a lot of pain and misery when it came time to actually run the robot.
- Buy a DC Clamp-on ammeter. Innovation FIRST let us borrow one for a while at the New York Regionals last year and it helped us debug the drive system greatly. As a note, had an Fluke 36 which costs $250, but there are cheaper alternatives.
I would advise you to reconsider treads. Much of the advantage of treads is lost when you have a flat floor.
You may want to spent the extra effort into finding a good material to make your wheels out of or to wrap around whatever wheels you are going to use.
If u make the threads out of metal make sure they are not sharp!..they will tear up the carpet…
We did lots of testing to make sure we didnt do this last year…
We used treads in '99 and '01, and if the game requires traction, treads are hard to beat.
Our '99 machine had some problems with chains rolling off of sprockets, mainly because our structure was too flexible. We made some changes during the season, and got to where we had few problems by nationals. Our '01 machine has a very stiff structure eliminating our '99 problems and has been very reliable.
As Clark mentioned, if you use metal treads as we have, you need to round off the edges and remove the burrs or they can damage carpet, but they have great traction, even ‘smoothed up’ to where they are ‘carpet safe.’
Still, for a rookie team, I would tend to recommend wheels. It is much easier to make a good reliable drive train with wheels, and with four wheel drive, the traction can be good. Just be careful to avoid the mistake of gearing too high. Going fast is nice, but going a little slower, but being reliable and, with this year’s game, being able to push well, is better.
Maybe the big compromise of good treads vs wheels if you are the ‘master of wheels’ as is Chief Delphi, is that with wheels you can do the ‘crab’ thing. That was really cool in the '99 game as demonstrated by Chief Delphi, Wildstang and others.