Omni drive train
 

Hi,
I am looking for infomation about drivetrains based on omni wheels (main wheel), something like this https://www.youtube.com/watch?v=TtUeZnuL4bs
I know 33 had 4 omni wheels drive train but couldn't get any good picture of it.
Specificly I want to know, how different/hard it is for the driver and is it easy to push it with another robot.
http://www.vexforum.com/gallery/file...5/sexyness.jpg

Yarden

Re: Omni drive train
 

My team has never used an omni drive train, but I do know a little bit about them. Please correct me, anyone, if I am incorrect in any of these statements, and I hope this can answer some questions.

Holonomic Drive: More prevalent in FTC, it uses 4 omni wheels with 2 sets of 2 perpendicular to one another. In theory, this drive should be able to drive in any direction, but my understanding of why this isn't used in FRC is because of a terrible lack of traction and difficult programming.

Slide Drive: Also known as H-Drive, it provides around the same maneuverability as a mecanum drive, at the expense of the weight of an extra drive motor (for strafing).

Omni Drive: I don't really have a name for this, but it is what 33 and team JVN (VEX BuildBlitz) used this year. It was a 4WD, with omnis instead of traction wheels. I do not really have any info on this one, except I have been told that it allows for easy drifting and better handling.

Butterfly Drive: This drive is derived from octocanum, which was derived from nonadrive. The only difference between octocanum and butterfly is that omni wheels (butterfly) are used instead of mecanum wheels (octocanum). Butterfly articulates between traction and omniwheels, usually with the traction wheels geared low and the omni geared high. An example from this year is 3847.

Nonadrive: I may not be totally correct on this one, but it was first used in 2010 by teams 148 and 217. The difference between this and butterfly is that a slide wheel for strafing is deployed during the omni phase to allow for strafing.

In the end, to your question for driving, it all depends on how much time the driver has to practice, and how good the driver is.

Re: Omni drive train
 

You pretty much nailed them all. I just want to point out that "Holonomic drive" is sort of an incorrect name.

Holonomic just means that it is capable of moving in any direction, at any orientation. So there are many non-omni-wheel drivetrains that are holonomic (Swerve, octocanum, slide drive, nonadrive, and mechanum are all holonomic)

Though I do frequently see people use the term in the way you listed it. I've also heard it called "Kiwi drive", but some people only call it that when it's a three-wheel system (as opposed to four).

Back to the OP's original questions:

The system 624 is using in that video is known as Nonadrive, because it has nine wheels (Nona = prefix for nine). It switches back and forth between the five omniwheels and the four traction wheels. 148's 2010 robot also featured a Nonadrive (As far as I know they first invented it)

The advantage is that you can use the you can move omnidirectionally when in omni-mode, but, as you guessed, that leaves you open to being pushed around, because omniwheels offer very little traction. so when you get into a pushing match like that, you can switch over to the traction wheels for traction.

As 3175student17 mentioned, Nonadrive is not used very frequently because it has evolved into Octocanum (Four traction wheels, and four mechanum wheels).

Octocanum is used because it's simpler than Nonadrive; it requires one less set of wheels, and one less motor.

However the photo you posted, from the Vex forums, is a different kind of drvetrain, as 3175student17 also pointed out, that is usually called "Slide Drive". It gives you the same omnidirectional movement of Nonadrive, but you can't switch to traction mode if somebody is trying to push you. So yes, you will be very easy to "boss around" on the field.

Re: Omni drive train
 

Here's how it works in my head:

Re: Omni drive train
 

The little voice in my head (that sounds disturbingly like what I assume Ether would sound like) keeps yelling that what you've classified as "independent" isn't correctly classified as holonomic.

(what follows is a dim recollection of an explanation, I'll try to find the original post later)

While all degrees of freedom are controllable they are not all controllable ALL the time. For example, picture a "independent" drive system (think 16's drive this year) sitting with its wheels oriented to 0 degrees (along a Y axis). At this point it cannot control its position along the X axis.

Re: Omni drive train
 

While the system that we used this season is heavily based off of 148's Nonadrive, it's actually a simplified, lighter weight, and easier to machine variant that we call the GrassHopper drive. Only the front two traction wheels actuate down in traction mode, thus tilting the robot back at a slight angle and pushing the back two traction wheels into the ground as well. The chassis was assembled from mostly WCP VersaChassis components, limiting the machining complexity required.

After championships we made a few modifications to the original drive and saw even more maneuverability in omni mode, as well as increased pushing ability in traction mode.

Personally, I prefer this drive to Octocanum. I think that it allows more maneuverability in non-traction mode without the added weight of four shifting modules. With our off-season iteration working quite well and brains churning on possible future improvements, I would not be surprised if this drive made a return in upcoming seasons (game permitting of course).

You can find some info on our original iteration here: http://www.chiefdelphi.com/media/papers/2968?
If you have any questions on our drive, feel free to ask.

Re: Omni drive train
 

Huh, very interesting, thanks for the correction!

This is a really great drive idea, much simpler than the others. I'm not super familiar with the Versaframe setup though, so I'm having trouble understanding the shifting front wheel.

Is it physically connected to that tensioning cam, welded or something? If not, how is it constrained, why doesn't it just slide around all over the place?

Re: Omni drive train
 

There were a few things holding the VersaBlock in place on our original design. The tension of the chain held the block in the horizontal axis by pulling it tight against the cam (which was not physically connected, just rotated to achieve the correct chain tensioning). The "elastic tension" you see in the pdf consisted of surgical tubing pulling the block upward until the bottom lip of the block was tight against the bottom of the frame. The piston would overcome the force of the surgical tubing, sliding the wheel into the downward position.

As you've probably realized, this original design was a bit of a makeshift solution, and one of the modifications we made going into the off-season was to attach plates to the blocks that pivoted close to the omni wheel. That meant that the traction wheels, when actuated, would follow an arc rather than sliding up and down, keeping the chain tension more uniform throughout the range of motion. We saw a much more smooth and consistent actuation. I attached an image of a quick mock up to show you basically what it looks like now.

In future iterations we could possibly move the plate's pivoting point back to the omni wheel's shaft, so that the traction wheel would follow an arc such that the chain would always be the same tension.

https://www.dropbox.com/s/1ffm7908vk...nActuation.bmp

Re: Omni drive train
 

Hello! Just last week we hosted our 5th annual off-season event, with about 36 teams. Our drive system was very modular, and we had omnis in the front and blue nitrile treaded plaction wheels in the back. One of our traction wheels cracked, and we only had extra omni wheels. We swapped out the wheels so that we would be driving all omni and hoped for the best.

Our first match with the new drive was one of the most fun match me (coach) and our drivers have had all season. The robot was gliding across the field like a piece of butter on a hot skillet. Handling and turning improved greatly. Overall it is a whole new world when you start driving it. It makes your robot movements more fluid, and encourages you to keep your robot moving. You are immune from t-bones putting robot at the complete mercy of your opponents, though you do have to maneuver away as you don't have any staying power.

In regards to pushing power and defense: With the wheels you won't be able to push worth beans, and you won't be stopping anyone. However you can still very much get in the way of robots scoring, and ram (at a non-damaging speed) to disrupt game piece manipulation. You can still play very solid defense with a good driver, just be more strategic in doing so.

Our robot performed its best all season with this drive system. I recommend that teams give it a try.

Specs:
Vexpro 6" omni wheels in a square config.
Each wheel powered by its own CIM and gearbox (4 CIMs total)
Gearboxes geared for 11 fps
Mounted on Wild Swerve modules Zip tied into a fixed position. (worlds most expensive tank drive :yikes: )

See attached thumbnail for picture of module (not completely assembled)

Re: Omni drive train
 

We've used the 4 omni wheel holonomic for a few years now. We love it and it's a great system. It's super simple to build,weighs very little, and not that hard to program(provided you know basic trig and a little trick with the system). here is our last year's bot during practice and a game match:

practice:
https://www.youtube.com/watch?v=lE8Q...7xo1Oi08f0dphg

actual match:
https://www.youtube.com/watch?v=Kr3jjJhHSsM

i also have a few vids uploaded of a swerve we are prototyping for next year. While the holonomic system works great for us we aren't very fast and get pushed around by pretty much anyone.

Re: Omni drive train
 

haha Yep! I was the driver, and there was a noticeable performance and handling improvement. I remember doing donuts down the field like swerve :D Aside from cracking the wheel, we actually threw the tread too!