Zach M
This is just an FYI notice from our experience. This year at our regional competition, we learned a valuable lesson we will never forget. Use the friction tape provided by BEST. When we ran out, we bought a slightly different type. Unfortunately, it was so sticky that it caused us to barely be able to move. It cost us 1 nearly scoreless round and another scoreless round. It caused us so many problems and made us do a whole lot worse than our practice. Take my advice and use the right friction tape! [smile]
Quote 0 0
Always good advice to use real spec kit materials. 
If you ran out of tape because you used some failed other attempts, getting more is fine.
If you need more than one roll on the robot at one time, you shouldn't be using extra.

Friction tape has two sides, one is stickier.  I hope that was not your problem.
If you are using friction tape as a tire on the wheel, there are other things that work too.
  - string wrap as treads,
  - Bike inner tube provided, cut into strips.
  - rubber bands
  - rubber core from the bungee cord.
or even 
  - one Team Custom Part per wheel, using some kind of flat rubber strip from McMaster, and you have an extra that you can roll up into a 2x4x4 block to show that you pass compliance.

With the other materials, the challenge is to keep it on the wheel.
Quote 0 0
Zach M
We generally replace our friction tape every few rounds, but we didn't use more than one roll at one time. 

I don't think that was the issue this time - this stuff was pretty sticky on both sides. 

Yes, we're using it on wheels. Thank you for those ideas! I will remember them for next year. Because we got so "burned" from our experience, we are going to be meeting in the second semester to figure out means of motion and such so that that never happens again. We will try all those techniques. Thank you very much!
Quote 0 0
As long as you are testing different ideas, here are a few other things my team has tried or seen over the years.
  • "knobby" wheels - not a smooth circle, but many small cuts/ridges around the edge
                  We've seen ridges from .125 in wide to 2 in wide
                  We've seen knobby tires with and without friction tape
  • instead of string tread, take a long strip of friction tape and twist it into a strand (use a drill)
  • Hex-head sheet metal screws as treads
  • Omni-Wheel rollers

One year we considered printing a strip of tread from a flexible filament as our TCP (coiled for compliance as jgraber suggested), but we have not seen that used on an actual robot.

I am guessing that your robot moved forwards and backwards well, but could barely turn.  If you have 4 wheels, you may want them to have different traction properties to reduce turning scrub.  We've have ended up using onmi-wheels the past few years.  We have tested other options, but have ended up with 2 omni-wheels on the robot each of the past 3 years.

I've seen teams use friction tape on two wheels and electrical tape on the other two so that they slide side to side easier.

Good Luck.

Post some general findings if you get the chance.

Quote 0 0
Zach M
Actually, we have a bent PVC (4" pipe) skid on the front of our robot, and turning was quite fast, until we picked up the manikin which surprisingly moved the CG so far forward onto the skid that it slowed significantly. We had 2 10" wheels on the back connected to small motors for higher RPM. Again turning was still pretty fast. The problem was mainly so much friction on that skid that caused it to have difficulty going over bumps, and sometimes the tape of the starting box. I do not recommend electrical tape on PVC skids - for some reason that slowed us down worse than anything else!

Thanks for the ideas! We will definitely try all these and see what works and give a report. I'm sure our findings will be profitable for us and many teams to come.

From an ease of construction stand point - I like the screw into wheel idea. I imagine pilot holes so the wood doesn't split is recommended though? This idea could be made much faster than carving gear tooth wheels - especially if a CNC machine isn't available. (I've tried this with a dremel before.)

Another thing we've seen work in the past very well is the combo of gear tooth wheels and a castor wheel rather than skids. That is our first priority to make. Maybe gear teeth won't ware down as much as friction tape. We will see. 

Quote 0 0
Why didn't you have the drive wheels in the front, next to the manny/can lift, instead of at the back?

Wow, 10" wheels on small motors, that's fast! That is eqv of 20" wheels on large motors, which I have not seen lately, only up to 14" or so.
I'll have to check the compliance form survey to see what largest wheels were in Texas.

Balancing weight over wheels is always an important issue, and even more so when picking up relatively heavy loads, to avoid your kind of traction issues.
Good balance without low CG and slew control also leads to jerky robots, and often requires wheelie bars in both sides of drive wheels,  which can then lead to high-centering the robot on game objects. 
  Wheel notches are fast to make with a band-saw.  I'm against hex heads as wheel treads, because they tear up the field, and leave sawdust on plywood ramps.
  Castors are surprisingly easy to make from 1/4" threaded rod with 2-4 bends, 7 nuts, and 0-2 base plate bearings.  I thought for hours about how to make a jig to get the bend angles and places exactly correct; I finally just bent the rod by hand to see how the jig should be made, and it was so easy, I never made the jig. 
  If you want two castors, a second wheel can be made with a hole saw, and 0-2 wheel bearings.

Quote 0 0
Zach M
Well, at the time we had no particular reason for doing that, but in the end I'm quite glad we did, because our greatest weight came from the chemical drums. We picked all three up at one time, so the wheel position helped tremendously there. Our robot went significantly faster than normal with the cans. Since the cans were heavier than the manikin, our wheel position worked well.

Yes, well, it was supposed to be. A note to anyone thinking about using small motors: like jgraber said, CG is super important. Our CG didn't maximize the mathematical potential of our wheels and motor choice, so it didn't turn out to be extremely fast by the time we got all our equipment on the base of the robot. I think that for our weight and the way our robot was setup, 20" wheels with large motors may have been the better choice, as they would have had more power but equal speed. I believe I saw a robot at South's BEST that likely had that arrangement. 

For that reason, I would like to look into where the CG was, and where it should have been to achieve max speed per the wheels we had. I would also like to see what the stall weight for the small motors and 10" wheels would have been, to estimate where the line should be drawn as to whether or not small motors should be used. 

Ok, thanks for the ideas. I see your point about hex heads - likely not a good idea. I'm pretty sure we're going to try gear tooth wheels first in the second semester. 

Ok, we'll give that version of a castor wheel a shot! Thank you for the idea. 

Also, here's a tip for jerky robots - this year we programmed a slow button which allowed us to half our speed when performing high precision tasks. This allows you to have max speed when you need it, but super slow precise movements when necessary. 
Quote 0 0
So cans over wheels in back, and manikin over skids in front?
The other issue with 20" wheels is that they narrow the mounting location to the center.
If I could come up with a decent belt material, like metal pipe hanging tape, I'd like to try larger motors with 4" pulley belt drive to 10" wheels with 2" pulley. 
Another interesting idea that allows very narrow (<6"?) robots is large motor with large wheel on top of small motor with small wheel.  Summer BEST training generated a protobot with that motor/wheel combination, and it drove ok.  That also allows a small and large motor for mechanisms, if needed.

More drive control:  Yes, half speed feature and non-linear joystick conversion can make robot easier to drive.  In 1987, ElectroGlass brand semiconductor wafer probers had an X,Y,theta joystick that twisted to switch between 3 modes:  scan (fast), Jog (repeated small steps), and single step.  That would be interesting to try as drive mode.   When aiming a pea shooter, you get more resolution if the exit point is centered between the drive wheels.
Quote 0 0
Zach M
Yes, but never at the same time. Our strategy required that we finished rescuing him before we got the cans. (this lead to 3rd place in the Founder's Award at South's BEST.)

Also, we had just one skid in the middle of the front, which turned out to be a bad idea. All the weight in the front was on that one skid, so friction was not spread out between another skid. This lead to us mysteriously getting stuck on the tape of the starting box, but for whatever reason could easily go over the slat of wood in the entrance to the hot-zone. Still not sure I understand how that happened, but I'm fairly confident that what we learned is that castor wheels are better than skids and 2 skids are better than one if you have to use skids. 

Ok, I don't really understand the part about the large wheel on top of a small motor with a small wheel, but what is Summer BEST Training? Is that available to students, or just volunteers?

Also, do you have any resources for getting smarter with all this technical stuff? There's a lot that could be learned, I know, but I don't know what would be the best way to learn it. 

Very true. We should have taken this into consideration while we were designing where to put our water launcher this year. We will keep that in mind if they allow projectiles again next year (which I really really really hope they do, that was lots of fun.) I think the theta joystick idea (3 modes) is definitely applicable for all years, though. 

Thank you so much for all your ideas. It has been really helpful. 
Quote 0 0
My name "Summer BEST" has some official name about nationally branded BEST training intended for teachers.  It usually gives Continuing Education credits, lasts 4 ish days.  There are some grant funds available for public school teachers, and I hear next year for all teachers. It has been going a couple years at UT Dallas, but I hear they are trying to expand to three locations this summer.

re Large. small:  The usual minimum width of a BEST robot is constrained by having motors back to back.  How can you get narrower?
- use belts or some other non-direct transfer of motors to wheels, so the motors can be mounted differently.
- mount motors non co-linear side by side in horizontal plane, so that the wheel centers are not directly across from each other; how does that affect steering in each direction?
- mount motors non co-linear top/bottom in the vertical plane, so that the wheels centers are parallel in the vertical plane, but not in horizontal plane
-- When top/bottom mounted, the wheels have to be different sizes to avoid tilting.  Might as well make the larger wheel to be 2x larger, and use the large motor. 

Learning Technical stuff:  I dont have a good answer.   There is a lot of stuff on BEST website, team resources. 
Quote 0 0
Zach M
Thanks for the ideas and help!

I created a caster wheel with a single ball bearing, and the result was phenomenal. I highly recommend casters to teams trying to get more maneuverable and fast. The robot literally spins on a dime! A single ring of friction tape around the drive wheels is working currently. It appears that the caster vs. the skid is putting far less stress on the motors since they have hardly any resistance with the caster, which will probably increase battery life.

About the carpet at our regional competition: Our coach found out that other teams had the same problem as we did, and as of now we are thinking that it could be the fact that we had the small motors driving the wheels. We think that they were simply not powerful enough to overcome the slightly softer carpet. The super sticky friction tape certainly didn't help if that's the case. We will be brainstorming different ways to be more adaptable between our local competition and regional competition. 
Quote 1 0
Can you post a picture of your new caster?   that would be worth a 1000 words of description.
An old team used a 3/4" bearing ONLY as the caster wheel, and failed on thick carpet.

Since you are new to casters,  you should practice dealing with its notable problem: 
"Caster snap"

- Made worse by long arms, larger caster offsets, longer distance between motor axles and the caster pivot mount.
- Most noticeable when changing from straight forward to straight back.
- As the robot starts to back up, 
the caster first rotates around the pivot point while caster tread spin-grinds into floor,
then the pivot joint wants to circle around the caster tread at the caster offset radius,
which exerts side forces on the rear of the robot, making it jerk to the side.

For warehouse robots that always turn at the same place on the floor,  the caster snap can cause the floor to wear unevenly.

I usually start teams with 8-9" wheels on small motors, not 10".
This reminds me to do a demonstration on balance on skids vs castors in teacher training this summer.

Quote 0 0
Zach M
Yes sir! I will also provide some CAD because there's a lot you can't see from the picture... (it will probably take me a while, I'm busy these days.) 
Haha I bet... I'm  using the razorblade wheel and the ball bearing is for the rotation of the wheel. I will post a picture shortly. 

Ah yes, I have noticed that some. Thankfully it is not too noticeable, but I can see that if the base wasn't level after the caster and drive wheels were put on, the caster snap would be worse. Assuming the base of the robot is a flat plane, do you think that adjusting the heights of the caster and drive wheels such that the base is level could decrease the effect of that issue?

Ok, I definitely want to try a variety of different wheel sizes. My plan is to bring a bunch of sizes to gameday to prepare for the worst after our experience at regionals. 
Quote 0 0
If you plan on trying different wheel sizes, you may need an adjustable height wheel motor mount to keep ground clearance the same.  

The use the smallest caster offset that trails correctly to minimize caster snap effect.  
If you make a two part caster, pivot + wheel carriage, with a slide mount between them, you can use the slide mount to easily adjust the caster offset.   Caster offset I define as the horizontal distance between the vertical pivot and the caster axle.
Quote 0 0
Zach M
Good point. It may be good planning ahead if we make something we can adjust like that... it would also take the guesswork out of making the base of the robot level if  you could just adjust it! Thanks for the idea! 

Another good idea! I just drilled a lot of holes in the carriage for adjusting, and I think I found the sweet spot.  IMG_2855.jpg  IMG_2880.jpg  IMG_2881.jpg 

These are the pictures of the caster I made. The block of wood is the ball bearing receptacle, and I think the rest is fairly self explanatory. 
Quote 0 0