What are the benefits of using Colson wheels?

Are there any specific benefits of using colson wheels over, for example, AndyMark plaction wheels?

Also, is there any specific reason to use 4" wheels over 6" wheels?

1 Like

I can speak to 4" instead of 6"… the smaller wheels allow you to have a lower center of gravity, as well as less of a gear reduction required to get you to a reasonable ground speed.

Relatively inexpensive. Good driving characteristics. Good wear characteristics. Same traction when worn. Durable.

I have a hard time finding any difference in grip between blue roughtop wheels and Colson wheels but they wear far far better. I try to run Colsons as much as I can.

We had some weird experiences with Colson wheels this year (and in hindsight, last year too). We found that wheel odometry measured different distances depending on the orientation of the robot relative to the weave of the carpet. Replacing the Colsons with blue nitrile tread completely resolved that problem for us. Not sure if that is only an issue with our practice field carpet (which was pretty heavily worn this year) or all low pile carpet in general, but the effect was pretty pronounced…up to 5% differences in measured distance.

2 Likes

A smaller diameter wheels allows for a smaller reduction in your gearboxes which means saved weight, space, money in gears, and manufacturing time.

Thats really interesting, at what point in the season did 254 make this change, and was it made to the comp bot too? How did you even expect the wheels to be the culprit of this, and not just assume is general inaccuracy?

1 Like

I think we have noticed stuff like this too. Carpets seem to have a ā€œgrainā€ (you can often see and feel this when vacuuming) and there can be a difference when driving with the grain and against the grain.

We were tuning our drive parameters by pushing the robot a known distance and counting encoder ticks (checking units and polarities should basically be ā€œstep 1ā€ of any feedback control bringup), and we noticed we got different numbers forward and reverse. We checked everything - making sure we were loading the wheels in the same way when pushing the robot, checking encoder mounting and wiring, etc. We verified the wheels rotated a different amount by putting a sharpie tick on the side and counting rotations visually. We also tried using passive follower omniwheels and had similar difficulties.

Once we put on roughtop, the problem went away, and we didn’t look back. It took us well over a week to diagnose and fix this problem, which really cost us programming time, and we had to scramble to have working auto modes for our first event.

FTFY.

I mean seriously, people wonder how a team can actually achieve a season like yours… it’s this level of attention to detail that does it. Something most teams would just shrug their shoulders about and say it’s ā€œnormal errorā€, you guys dig into and fix so you can have perfect results.

4 Likes

That is very interesting. If y’all ever follow up on this I’d love to read about it. I would have assumed that other than wear the wheels would have been quite similar.

1 Like

Colson wheels are relatively light for being ostensibly indestructible, quite inexpensive, get good traction on carpet and the plastics used commonly on playing fields, and are available in tons of different sizes. We love Colson wheels and use them frequently.

We have had plaction wheels fail on us before, although likely through our poor implementation. Still… Colson wheels have no fasteners to come loose or tread rims to break and are fraction of the cost of plaction wheels. Hi-Grip wheels, and other wheels with aggressive tread blocks, have been observed to peel up the gaffer’s tape used on many fields, sometimes resulting in a disabled drivetrain. We steer clear of these. Some teams ā€˜tune’ their center drop by machining off some material from colson wheels, and we’ve machined them narrower in some applications. This added flexibility can be key in optimizing your drivetrain, which I have not seen demonstrated by other wheel designs.

As for 4in vs 6in… 4in has less material and thus its diameter will change faster than a 6in and it can’t climb the same obstacles that a 6in can; but, a 6in is heavier and requires a larger gear reduction for the same free speed. A 4in wheel can be placed closer to the end of the robot fore and aft, increasing wheelbase and stability, but a 6in wheel is likely to generate ever slightly more traction than a 4in wheel because it has an effectively larger footprint and lower contact pressure, all else being equal. In 2016 we used 6in wheels to deal with the obstacles on the field, but 4in wheels in 2017 and 2018 where the fields were flat, or nearly so.

Fascinating. We had no such issues with our ~3 year old practice carpet and were seeing distance variations of <1%. We had good results on the competition fields as well (which is always new carpet).

Do you have a hunch as to why this occurred and why changing wheel styles helped?

Neat. We noticed the carpet pile problem last year when we had our first all Colson wheel robot. This year we used follower encoder wheels that seemed to read distance well, but I don’t recall any testing with/against/across the carpet pile to see how accurate they were. It’s hard to do without an actual field.

We did exchange our Colson wheels (just two this year, the middle wheel on each side of the drivetrain) for AM kit wheels. They have more traction out of the box, and are just as cheap and easy to replace when they become worn.

also interested in possible explanation of difference from Colsen to Roughtop

We have a white board that still has the differences we were seeing from going north vs. south - with Colsens - we did not resolve it. But how do rough tops change this? Is it that individual colsens slip (3 to a plane) while greater pile depth on roughtops still grab? we were not able to come up with a pattern but also did not try switching wheel types…

We had the exact same problem and drove us nuts. Spent a few weeks dealing with it (±2 feet spread on the swerve :frowning: ). Ended up switching to traction and life was great.

Until now I was in the colsons are the best camp… but we had the inconsistency issue but did not relate it to the colsons…, still don’t really understand what was going on - would spring suspension ala 2767 solve the issue also?

Nope, it doesn’t.

!! Thanks for feedback.

I guess we got ourselves a summer project.

I wonder if you could counteract the effect by cutting in horizontal tread into the colsons. I guess colsons kind of grip on top of the carpet grain, and the blue nitrile/traction material covered wheels grip **into **the grain?