We are a new VRC team, coming up from VexIQ. While looking at VRC drivetrain designs we notice that many teams use 4 omniwheels even when designing a tank drivetrain. Why not two traction wheels and 2 omniwheels?
In, VEX IQ, using two traction wheels would give us much more consistent results, specially for programming.
Why are most VRC teams using 4 omniwheels when design a tank drivetrain?
Thanks in advance for any insights! Much to learn.
Cris
For me the change of drive trains from iq to vex is for 2 main reasons:
because the field is a different material (foam not plastic) which already creates more then enough traction to do things accurately.
Driver preference, people like driving robots that have a turn radius in the center and omni wheels also help with turning scrub.
An alternative that was widely used in Nothing But Net is having 6 wheels, 4 omnis in the corners and 2 traction in the middle, this still gives the center turn and clean turn and also stops your robot from being pushed in a match. The only issue is that this only works well with 3.25" wheels because the other traction wheels are slightly smaller than the omnis of the same size. This makes it so you have to slightly increase the diameter of the wheel by rubber bands or other means.
We like the center turning point as well. But it is not always in the center of the robot - it depends on the weight distribution of the robot.
We were planing to use the 6 wheel configuration you described above, with 4" wheels. Sad to hear that this might not be a feasible option due to the wheel size discrepancy.
We might start with 4 omni-wheels then. Let’s hope the robots don’t get pushed around too much in the “In the Zone” game.
Usually when you have two traction and two omnis it can make the system slower at turning or have friction when turning with such massive robots. I also go with 4 omnis because turning is very much faster since the center of turning is centered.
Being apart of FTC, VEX, and FRC (FRC is not completely transferable since matches take place on carpet instead of foam tiles), I have noticed a couple reasons why a team might choose to use 4 omni’s when designing a tank drive system because by using omni’s you are able to create a smaller turn radius like others have said, but it also allows you to “drift” into your turns and higher mobility. Some teams with practice have found that the increased mobility is a big asset when trying to out maneuver their opponents to game pieces. One thing to note, like others have stated is that when you combine traction and omni wheels your center of turn is moved from your center of mass to the axle of your traction wheels. Sometimes this is very helpful and teams want to incorporate this into their robot and other times they do not. As for only using 4 wheels, that is because there is hardly a need for more than 4 wheels on a drive train in vex unless you are trying to get over an obstacle or are using a drop center. By using more that 4 wheels you are adding more weight, parts, and complexity to the design. However if your team feels like they need more than 4 wheels that by all means feel free to do so.
The turning is much better with four omni wheels. I think teams have been able to turn just find autonomously with all omni wheels.
The 6-wheel configuration you mentioned will indeed not work. I was interested in that sort of drive to start the year but upon building the first prototype, found that the traction wheels are roughly .2 inches smaller in diameter from the omni wheels. Thus if you put the traction wheel in the middle it will either have no effect at all on the drive or will do very little.
Once more weight is added + foam ground, it shouldn’t be too much of an issue. (I had that type of drive during nbn) Buuut even though that’s said, the circumference the 4 wheels on the outside is bigger than the two traction wheels in the center, which will mean the 4 Omni wheels have to turn more to make a full 360 degree turn. This will cause some issues when turning, but it’s so little it shouldn’t really matter.
(998th post)
The reason that teams use 4 omni-wheels is if they are powering all 4 of them. If they are only powering the back 2 wheels, they can use traction wheels in the back and omni-wheels in the front to remove skidding. If all 4 wheels are powered and the back wheels are traction wheels, it will skid, but if powering 4 wheels and all wheels are omni-wheels, then there won’t be skidding.
That is not true. The issue with all omni-wheels in a tank-style drive is that it can slide sideways fairly freely by the very design of the omni-wheels, regardless of how many of them are driven. So if another robot pushes from the side, such an all-omni-wheel robot will slide or drift. The same thing happens with similar pushes. Let’s say you have a ramp and you want to drive horizontally along it, then the omni-wheels won’t stop the robot from sliding down the ramp. Similarly, drifting or skidding will still occur if you cut a corner and still have some radial momentum. You can deal with this by driving more slowly or cutting turns more gently, but as long as you go for full mobility, you’ll get that skidding.
No, this isn’t just theory. I’ve been playing around with an all-driven all-omni-wheel robot in the last week and have watched just such skidding happening. The skidding happens both theoretically and experimentally.
What I meant by skidding was that the wheels won’t stop the robot from turning by itself, like for example if you have 4 traction wheels and you want to turn then it will be much harder than turning with 4 omni-wheels. I wasn’t referring to another robot pushing it, but rather the resistance to turning by itself. But yes, you are right that no amount of onmi-wheels will stop it from moving horizontally (assuming that the robot has a simple tank-drive base) when another force is acting upon it.
This thread has given lot of good reasons. One last reason is that tests performed by community shows that on foam the omni wheels have more traction than the high traction wheels.
Also the drift sideways on all omni wheels is pretty negligable. I can promise you that. (I localize off of only omni wheels with milimeter accuracy over several minutes.)
The standard 4" wheels can be shimmed out to be slightly larger in diameter than the 4" omnis. The last time I did this I used small (VEX-legal!) pieces of 4" wire ties inserted into the slots on the wheels, under the tires. I used needle-nosed pliers to insert them. Worked great.