H Drive questions

[Tottanka]

Hey all,
Im looking for a few tips from anyone experienced with designing and using an H Drive.
First of all, are there any problems with driving in a direct straight line with that kind of drive? I'm thinking the Omni's sliding ability will make driving less "expected" than the usual traction wheel tank drive. If so - how do you solve that?

Second, how much torque do I need to use for the non conventional direction of movement? Is the same gear ratio as other wheel acceptable, or would i need more?

Last but not least, do i have to install the 2 non conventional wheels on the center of geometry/mass of the robot in order to slide well, or is would it be OK to install them at say, 1/3 of the length on the robot towards the back side?

This is all assuming i'm using 4 Omnis in a tank drive style (X movement) with 2 more added at the back of the Drivetrain (Y movement).

Thanks!

[Ginger Power]

The GreenHorns built a slide drive for our Robot in 3 days. We had no problems at all driving straight, turning left, right etc. It seemed to operate exactly like you would expect a tank drive to operate. We drove the center wheel with a toughbox mini and 1 cim motor. This amount of power seemed to be more than sufficient. We had no trouble strafing. We installed our single center omni wheel towards the middle of our robot and like I said it strafed as you would expect. We tried shifting around our center of gravity around with heavy ballasts and they didn't affect how it strafed. I can't speak to moving moving the wheel that far off center. Intuition would tell me that it may arc more than strafe perfectly straight.

[Ether]

Originally Posted by Ginger Power

We drove the center wheel with a toughbox mini and 1 cim motor.

For the record:

- omni wheel directly driven from gearbox output shaft?

- standard 12.75:1 gear ratio?

- what wheel diameter?

Also, did you try driving over a scoring platform?

My apologies if these questions have answers already posted somewhere. I haven't had time to watch all the videos yet.

[Ginger Power]

Sorry should have thought about it more:
We had a 16 tooth pulley mounted on the output shaft of the toughbox and a 48 tooth pulley on the omni wheel driven by a belt. Yes it was a standard gear ratio. We used a 6" aluminum dualie omni wheel from Andymark. Most of the decisions we made were based on material availability because we had very limited resources. All of our components were generously provided by 4607, just a third year team. We did some testing with the scoring platform, but it didn't take much testing to figure out our configuration wouldn't work. We used 4" omni wheels on the AM14U2 and only had about a half inch of ground clearence. The frame would get caught up on the scoring platform even at full speed.

I hope this clarifies things a bit. If I forgot anything please let me know!

[Ether]

Originally Posted by Ginger Power

Sorry should have thought about it more:
We had a 16 tooth pulley mounted on the output shaft of the toughbox and a 48 tooth pulley on the omni wheel driven by a belt. Yes it was a standard gear ratio. We used a 6" aluminum dualie omni wheel from Andymark. Most of the decisions we made were based on material availability because we had very limited resources. All of our components were generously provided by 4607, just a third year team. We did some testing with the scoring platform, but it didn't take much testing to figure out our configuration wouldn't work. We used 4" omni wheels on the AM14U2 and only had about a half inch of ground clearence. The frame would get caught up on the scoring platform even at full speed.

I hope this clarifies things a bit. If I forgot anything please let me know!

Here's what I think you are saying:

- you used a 6" omni for the center strafing wheel, and 4" omnis for the corners?

- the 6" center omni had a total speed reduction of 12.75*(48/16)=38.25:1?

- so the rotary free speed of the center 6" omni would be 5310/38.25=138.8 RPM (ignoring losses)?

- so the linear free speed of the center 6" omni would be (138.8/60)*(2*pi)*(6/2/12) = 3.63 fps (ignoring losses)?

Am I reading that right?

[Ginger Power]

Originally Posted by Ether

Here's what I think you are saying:

- you used a 6" omni for the center strafing wheel, and 4" omnis for the corners?

- the 6" center omni had a total speed reduction of 12.75*(48/16)=38.25:1?

- so the rotary free speed of the center 6" omni would be 5310/38.25=138.8 RPM (ignoring losses)?

- so the linear free speed of the center 6" omni would be (138.8/60)*(2*pi)*(6/2/12) = 3.63 fps (ignoring losses)?

Am I reading that right?

Yes this is correct.

[Ether]

Originally Posted by Ginger Power

Yes this is correct.

OK.

To teams designing H drives: please note the 3.6 fps free speed (3.6*80%=2.9fps estimated top speed). Fine control and high torque.

[Ginger Power]

Originally Posted by Ether

OK.

To teams designing H drives: please note the 3.6 fps free speed (3.6*80%=2.9fps estimated top speed). Fine control and high torque.

Thanks for helping figure this out. I'm sure many teams appreciate it.

[MrRiedemanJACC]

Thanks for this! Our team hasn't gotten this far yet with how we would set it up. We are seriously considering H-Drive, but believe a suspension would be needed for it to be effective tavelling over the scoring platform, but there is a lot of work that will go into that (spring and shock? spring only? what spring force needed, exactly how much travel, etc, etc - do you have any other questions we should be asking ourselves?)
So since are looking at the suspension end of it, we haven't looked at gear ratios. 3.6 seems really slow, but with only one wheel pushing the robot sideways, it makes sense you need to torque to accomplish the task. Even more sense is that you don't need to move quickly sideways, just enough to position where the robot is needed.

[GeeTwo]

Originally Posted by Tottanka

This is all assuming i'm using 4 Omnis in a tank drive style (X movement) with 2 more added at the back of the Drivetrain (Y movement).

What's the rationale for 2 wheels for strafing? Are they located in the same line? It sounds like they are.

The whole idea of H-drive is that you would like most of the advantages of a skid-steer, but are willing to give up some traction (including almost all of your side-to-side traction) for the ability to "strafe" just a bit, that is, move right-to-left with none or at least minimal rotation, but just enough to "line up" on some sort of target, not something that you're planning to use as you cruise down the field hoping to reach free speed.

The amount of traction you can get from each wheel is limited by its coefficient of friction and the weight on that wheel. Assuming for simplicity's sake that you're putting the same weight on each of your six wheels, you're putting 1/3 of your weight on the strafing axis, 1/3 on the left tank line, and 1/3 on your right tank line. With a classic 5-wheel H-drive, the numbers are 1/5, 2/5, and 2/5. Do you feel that increasing your strafe traction from 25% to 33% is worth reducing your forward/reverse acceleration from 75% to 67%? If so, by all means continue! The next classic step, by the way, is Killough or mecanum, each of which supports 50% forward and 50% strafing.

And, of course - why at the back? I'm assuming you pick up loads at the "front", which may be in error. If you aren't going to put the strafe force at the center of gravity, it is probably better to err at putting it on the side that you actually want to move left-to-right.

[GeeTwo]

Originally Posted by MrRiedemanJACC

We are seriously considering H-Drive, but believe a suspension would be needed for it to be effective tavelling over the scoring platform, but there is a lot of work that will go into that (spring and shock? spring only? what spring force needed, exactly how much travel, etc, etc - do you have any other questions we should be asking ourselves?)

The ideal suspension for a center-wheel of an H-drive would be one that always pushed it against the floor with the same force, presumably 1/5 of the robot weight, though you might come up with a different number. In order to keep an essentially constant force against the floor, you will want to use a setup in which the spring(s) is/are pre-compressed by the desired force, but in which the springs don't actually increase greatly in force as they make the necessary deflection for you to get over the corner of the platform. Disclosure: I've never done and H-drive before, and our H-drive this year is not going to attempt to climb onto the platform.