I have a robot with four 5" wheels. I am using tank style drivetrain to control it (individual joystick to control each side). It goes fine when going forward and backward. But when I try to turn it one side completely stops spinning and other side keeps spinning. I know nothing is wrong with program because I lifted the robot and tried it. Also it turns fine on a normal floor but not on field tiles. I am thinking 5" wheels don’t create enough traction? Is there any way to fix this? would filing wheels work?
Thanks:)
when you lifted your robot off the ground did the wheels spin at a normal speed or slower? and are you using clutches?
They spun at normal speed. And I am using old three wire motors so yes I am using those green clutches.
well if the holes do not look square shaped, i would consider testing the robot without the clutches to eliminate that as the issue
but they seem to work fine when I am going forward and backward. Also wouldn’t it damage motors if I don’t use clutches?
if you are careful and do not have much of a load, your motors should be fine without them. does the robot’s wheels turn in the correct directions for turning while off the ground or do they just move in forward and reverse?
They turn in correct directions (one side going forward and one side going backward).
the 5" wheels have too much traction on the foam tiles
try swapping out the two front wheels for 4" omni wheels if u have some
if not, it will help if you made the front and back wheels closer together (so you get a short and fat drive train)
We have omni wheels but I can’t bring down my robot lower any more.
Our competition is this Saturday. I don’t have enough time to change all that 
What about adding wheels in the centre without motors? Would that help at all?
Also right now I have 1:1 gear ratio for motors. 4 Old motors are connected directly to each wheel. Would changing the gear ration help?
I just want to make sure that it will work before changing anything
What motor/gear setup do you have for your 5" wheels?
What is the wheelbase length and width?
What is the weight distribution between sides (likely heavier on the stopped side, and lighter on the spinning side)?
There is pointer around here somewhere to a “will it turn” white paper on chief-delphi website that shows that long narrow wheelbases are much harder to turn than short wide wheelbase.
The usual Vex solutions are
- use omniwheels on the front or back wheels
- increase your wheel torque through gearing or smaller wheels
Unusual solution is crab/ackerman/swerve steer drive.
I used
http://www.vexrobotics.com/products/accessories/structure/angle-hole.html
and
other piece is 17.5" in length. my wheels are parallel to angle hole piece (short ones) so my wheels are very close to each other on both side. but i have big space between both sides.
All my motors are directly connected to each wheel. They are old three pin motors. And weight is even on both sides.
Adding omni wheel in front would make my robot slanted, and I can’t do that without changing almost everything. I will post picture tomorrow.
or you can add 4 omnis
it will still be better than what you have now
(unless the lowered robot will still need you to change everything)
Can you explain/show how the wheel axles are mounted? Any chance the wheels do not stay vertical when you try to turn?
Pictures would be good.
You can often move the axle to a higher or lower hole to accomodate different wheel sizes. If there is no hole higher or lower, you can use a pillow block bearing. Since the front and back wheels are not chained together, different size wheels don’t hurt much.
Adding unpowered wheels in the center…
- don’t help traction
- might shorten wheel base if they are lowered, which might help
Since one side doesn’t move when turning, you need more torque,
Smaller wheels, or using gears to torque up, will both help.
Field tiles at contests often have anti-static spray on them, which adds even more friction.
Your best bet at this point is to pull a tire off one or two of the wheels. That should lower your traction and make it easier to turn.
I have also seen this on a lot of robots with 4 in wheels and I believe it is a contributor to drive train early stalling. I recently posted a blog entry that touches on this. It still remains somewhat of a mystery to me:)
Would appreciate any comments.
Chris
Lots of good data in the blog post; length/width/motors/gears/wheels;
What was weight of robot?
What was the temperature of the room?
Looks like more and more people are noticing that the simple friction model isn’t fitting the real world.
Is this a typo? I parse it as turn0.5, forward0.66
A more complicated formula or lookup table would allow straight to be higher power, while scaling back power on the turns.
Brainstorm of ideas:
- Convert to non-cantilever; (I almost stopped reading when you said cantilever.)
- use a DVM in 10-amp series with battery to verify your current clamp hypothesis, and verify if the PTC is cutting out early.
- Swap Cortex, if PTC has some wear-down function
- swap front and back wheels, if that helps keep uncentered weight distribution over the high traction wheels.
- convert to all wheels the same diameter (all omni, or pad out the hi-tracs)
- convert from gear train coupling to front wheels to chain coupling
- remove the front/back coupling gear train, does that help any?
- convert to ackerman steering, or 4-wheel ackerman.
- try low-pass or linear filtered drive coding, to enforce smoother driving
- Smooth your axles from square to round where they pass through the bearings.
- use better spacer washers between wheels and frame, since this is one place that turning side loads can take effect.
- when spinning left, and the left wheel stalls from rotating backwards, pull the arcade joystick to the back.
- program up some buttons to send known drive values to the wheels when doing experiments, to get higher repeatability than from joystick/human calibration.
- precool your motors with cryogens
Thanks everyone 
I changed front wheels. Now I have 4" omni wheel and it turns perfectly.
Although I had to cut my sliders. But its working perfectly now so its okay!
I wanted to do a quick calculation on your robot that has 5 in wheels and is having one rail stalling during a turn. If you haven’t done this type of calculation you might find it useful. Hope I did the math right in my haste:)
Using static moment equations you can easily derive the minimum torque to just turn for a symmetric robot with four direct drive motors on wheels with radius r .
Define the half spacing between axles as “l” and the half width between wheels as “w” , the robot weight as W_lbs and the coefficient of friction for a tire side force as u. Also assume that the weight is evenly distributed on the four wheels and the center of turn is at the cg.
Then
Input turning moment per wheel = w* torque/r and this must be greater than the moment resisting the turn caused by the side friction forces on the wheel.
Resistance moment = lW_lbu/4 . The factor of 4 converts the W_lbs to normal force on the wheel.
So torque > r*(l/w)W_lbu/4
Lets plug in some typical numbers. r = 2.5 in, l=3.5, w = 7
This gives a required motor torque of 2.53.5/7/4W_lbs*u
or torque >( 2.5/8) * W_lbs* u
or W_lbs < torque*8/2.5/u
Without omni wheels, u is typically around 1 or higher and if the robot W_lb is 12 lbs then the minimum torque
torque_min > 3.75 in lbs
This represents about 60% of the old three wire motors 6.5 in lb spec torque . Clearly there is not much margin for any torque losses in the drive train or extra side friction on the wheels or extra weight from a heaver robot or from game objects. Also any contact forces resisting the turn will easily stall the turn.
Using a 4 in omni wheel in place of the 5 in traction wheels increases the input torque by 5/4 and reduces the u by about 1/3 so we get a combined improvement of 15/4 or 3.75 factor on the required torque.
min_torque_ 4 in omni = 3.75 / 3.75 = 1 in lb or about 15% of the available torque of an old motor. This is where you want to be in your design.