Lynfield Drive

It seems unusual to name a wheel configuration after a team, but it’s the name that has been used in NZ by other teams and frequently at worlds, so I’m using it here.

The front two omni wheels are tank driven via chain from the motors at the back of the robot.
The motors for the back two wheels (strafe wheels) are sitting below the main drive motors.
The pneumatics are entirely unrelated.
This configuration is exceedingly useful for packaging the robot into a small base as it doesn’t require 6 wheels for holonomic drive. It also provides a large wheelbase to prevent tipping. Any problems with driving perfectly sideways can be fixed with code, and most drivers find the shallow arc to be useful when positioning the robot in front of goals etc.
Most importantly (at least for round up), the chasis doesn’t have anything in its center, meaning a robot with Lynfield drive can easily drive over round up rings.
Finally, the chasis has plenty of room in the center of the U for the arm and claw mechanisms.

2 Likes

but in the picture front right wheel on that bot has a high traction tire there. I am clearly missing something.

No, you’re not missing anything at all. It was I who didn’t see that. There I was thinking this was the best photo we had… Evidently this was the week we got the new high traction tyres and had put one on the robot just to test it out

Edit: I don’t believe it was very successful with the high traction tyre…

My team did a drive fairly similar to this this year, and it didnt have horizontal strafe, even with two omnis. However, our back wheel base was closer together. What may be happening is that while turning the back wheels they offset the movement with moving the front wheels.
the major problem with the two back wheels is only having front wheel drive in all directions, which caused us to have problems after getting a wheel stuck somewhere or in a tube.

A good demonstration of this chassis is the tv coverage of the finals of the New Zealand nationals, unfortunately lynfield loses as they are no match to goal lifting as we had not implemented this strategy yet but it is a good video of the strafing.

the only downside we see to this drive is that at max, there will only be two motors with the front/back movement and 2 motors for the left/right movement
there isnt really any way to “combine” the power of all 4 motors

We don’t see how this is a disadvantage over say a more traditional H drive as there is the same power going forward and sideways. Extra motors can be geared to the main motors powering the forward and back wheels by attaching them to the outside of the tower (this was tested and proved to be effective). Furthermore, the power on the wheels can be added together to some extent by driving the robot on a diagonal. Which is actually more common than one would expect.

To get over the only front wheel drive thing, you could have an extra set of wheels in the back also chained together.

The problem with having the strafe wheels in the back of the robot is that it tends to drift when you strafe. In a standard H holo, the strafe wheel is near the CoG, so it pushes directly on the robot. Here, the strafe tends to pivot a little around the center, depending on how much friction you have on the front wheels. It is really hard to code for without a gyro because the amount of pivot you will have is not always the same.

While it doesn’t strafe in a perfect sideways line, in some ways that’s not that necessary. They’re agile enough to slip around the field quickly in any direction. I very much like this drivetrain.

A) during the game, the slight turning while strafing wasn’t a problem, I wrote a drive program that corrected for it, but jacko preferred the ‘real driving’ feel of doing the correction himself when he needed it
B) during auto it was more of an issue, I had to drive the front wheels slightly to keep it straight, but once I had figured out the speed ratios I needed between sideways and forwards drives for different arm positions, it wasn’t too bad
C) strafe wheels not moving us straight sideways meant we could use them to speed up our turning. It was FAST :smiley:
D) in regards to being only front wheel drive, I don’t see how this is a problem. I did think that it would make line following tricky, but managed to get around that with the strafe :slight_smile:
E) getting stuck on tubes - because all of our wheels were powered, if they got stuck on a ring, we could just drive them off again

1 Like

AURA used a similar drive system on our college robots for Round Up, with the notable difference of putting two high strengths on each front wheel, meaning that we lost no pushing battles. The extra maneuverability was a big advantage in situations where we were fighting for position with opponent robots.

As for line following, it turned out that in round up you didn’t want to do much line following anyway. In our autonomous codes, we used line detection rather than line following to make sure the robot was parallel or perpendicular to objects depending on what we needed. Driving slower during autonomous also helped with the accuracy. We found that the robot did drive mostly straight, and the back strafe wheels turned the robot a lot faster than normal tank drive or H drive.

1 Like

One disadvantage of this drivetrain is that a substantial portion of your normal force is over two wheels that provide no thrust when moving forward. All else being equal, a robot with four wheels in the same orientation powered by four motors will have more ability to “push” than one with only two powered wheels.

In the real world, though, other factors may limit what you see. You may see robots suffering current overload and thermal shut down before anyone gets pushed, and wheels can slip on the tiles rather than “dig in.”

The most common result I’ve seen in pushing matches are shutdowns and robots not moving. In VEX it’s not usually a good idea to develop a strategy which involves stalling your drive motors.

Experienced teams can try whatever they want, but I urge new teams to focus on mobility and strategy and not brute-force solutions.

1 Like

This is true - a 6-wheeled H drive will have ~2/3 of its normal force over its foward driven wheels while this drive has less than 1/2, so in theory this drive should be more likely to skid when it gets into pushing matches.

However, this is complicated by the fact that the wheels are at the front and the things you are pushing are always at some distance off the floor. This means that these robots push much better backwards than fowards, because when pushing backwards the back of the robot will tend to try to tip fowards and the front wheels will dig themselves in. As long as your drivers know that when they need pushing power they need to drive backwards, traction shouldn’t be much of a problem.

This isn’t really a drive designed for pushing force, though - it’s an alternative to H holo where you trade ease of driving and programming for turning speed and space inside the robot.

1 Like

I just wanted to emphasize this point here, because it’s very important when talking about traction. Only half of the robot’s weight is on the wheels that power it forward, so you’ll only have about half the traction of a robot with two 393s powering all wheels forward. Slipping is not an issue for most teams, but you should realize that this setup effectively halves the weight of your robot in pushing battles. It also looks like your CG is pretty far back (unless the arm/intake is heavy), so don’t use this drive if you want to play defensive by pushing. A more traditional (and bulkier) H-drive is better for that.
With just two 393s on the forward drive though, this should not be an issue, unless you decide to gear it for torque (I’m assuming the drive was 1:1?).

I had a nicely typed out reply to exactly this about a week ago, but it went for mod approval and never came back.

Here is our (college) robot, which used the same drive but with additional forward motors:

With 2 393s on each front wheel (1:1 torque), it was able to lose traction when it tried to push forwards. The trick was to push backwards instead, which meant the front wheels would dig themselves in. Since we had to drive backwards to dump goals anyway, this wasn’t really a problem.

It obviously isn’t a drive designed for pushing power - it’s designed as a modification of H drive which allows you much more room inside the robot. If driven correctly, though, it doesn’t have to be bad at pushing.

1 Like

This might be a slightly clearer photo for what Oliver is trying to explain above:

1 Like

i see!
you have 6 motors total on the drive (4x HS 2x 2-wire)
well if you did it that way, then i guess it uses the same number of motors as the “H” drive
the only advantage this has over the H is more space for the intake (a U shaped chassis)
however, strafing may be an issue (you fixed this with code)
so i’d say they are pretty much alike
so does that mean you have a 2 motors for the arm and 2 motors for the intake??

we used the same type of design on our demo tank. except ours have 4 forward facing drive wheels and one perpendicular wheel that turns it.

Actually we were a college team so we got 12 motors :wink:
4 for the lift, one for intake, and one for descorer.

Although we had 1 HS motor and 1 269 on each side of the drive, and one HS on strafe for us, leaving 1HS and 2 269 for lift and 2 269 for the intake (we are high school, and had a very… very… weird motor configuration…)