Steering wheel

[Billfred]

Originally Posted by Hieb

As far as manuverability, I think that joysticks give more options while turning (tight turn radius/0 turn radius and so on). On the other hand, if the driver never makes use of all of the options available and can effectively drive using the steering wheel, then by all means go for it.

The robot has the same options for turning no matter what you go with--all you're changing is how you tell the robot to do it.

Assume you have one joystick for throttle forwards and backwards, then one steering wheel to determine rate of rotation. Calculate the difference in PWM values needed given the position of the wheel, then add in the throttle values. Since this might throw the PWM values out of the valid range, add in some code to scale back the numbers until neither is over 254 (or below zero).

IANAP (I Am Not A Programmer), but this would be how I'd attack the problem for a basic drive system without feedback or anything--those folks who actually know what they're talking about can elaborate there.

[Hieb]

Originally Posted by Dennis Jenks

As Cory stated, you can do all these things and more with a steering wheel. It's all about the implementation.

True. My initial impressions (while I was writting before) was that with the steering wheel it was more of an either/or proposition--I can either have a tight turn (ex. left wheels = 127, right wheels = 254) or 0 turn radius (left wheels = 0, right wheels = 254), but not both as would be possible with joysticks.

Further thought on the matter has led me to an implimentation that would allow both--I guess I just like joysticks because with a rookie team that only has 1 programmer (and she has no previous programming experience), I'll probably end up writting most of the code this year and joysticks are easy to implement.

[lukevanoort]

Originally Posted by Hieb

True. My initial impressions (while I was writing before) was that with the steering wheel it was more of an either/or proposition--I can either have a tight turn (ex. left wheels = 127, right wheels = 254) or 0 turn radius (left wheels = 0, right wheels = 254), but not both as would be possible with joysticks.

Further thought on the matter has led me to an implementation that would allow both--I guess I just like joysticks because with a rookie team that only has 1 programmer (and she has no previous programming experience), I'll probably end up writing most of the code this year and joysticks are easy to implement.

DISCLAIMER: I have never tried this
I think if you used the one joystick drive setup in the default code and just used the steering wheel's output instead of 'p1_x' and used the throttle joystick's y value in the place of 'p1_y' it would work.

[chris31]

Originally Posted by Hieb

True. My initial impressions (while I was writting before) was that with the steering wheel it was more of an either/or proposition--I can either have a tight turn (ex. left wheels = 127, right wheels = 254) or 0 turn radius (left wheels = 0, right wheels = 254), but not both as would be possible with joysticks.

Further thought on the matter has led me to an implimentation that would allow both--I guess I just like joysticks because with a rookie team that only has 1 programmer (and she has no previous programming experience), I'll probably end up writting most of the code this year and joysticks are easy to implement.

You could always have a button on the steering wheel or a switch to change drive modes.

[Billfred]

Originally Posted by lukevanoort

DISCLAIMER: I have never tried this
I think if you used the one joystick drive setup in the default code and just used the steering wheel's output instead of 'p1_x' and used the throttle joystick's y value in the place of 'p1_y' it would work.

Now that I think about it, this is exactly what the one-joystick code does. BRILLIANT!

[Hieb]

Originally Posted by lukevanoort

DISCLAIMER: I have never tried this
I think if you used the one joystick drive setup in the default code and just used the steering wheel's output instead of 'p1_x' and used the throttle joystick's y value in the place of 'p1_y' it would work.

K.I.S.S.

Now that I'm done beating my head against the desk for not recognizing the obvious I guess I can spend my time on more exciting things (like unpacking the huge order of supplies from McMaster-Carr and OnlineMetals--it really is Christmas time!).

Originally Posted by chris31

You could always have a button on the steering wheel or a switch to change drive modes.

That was the conclussion I came to after thinking about the problem a little more.

So now, as long as the steering wheel and the throttle either self-center or have a sufficient deadband built in to the code, I'm beginning to think this could be a very effective control that I had simply dismissed previously.

Who said you can't teach an old dog new tricks?

[Cuog]

For those who feel that the wheel will not give the same control options, I would have to say you are thinking too much inside the Joystick. The wheel can give you differences in control from only one or two pwm to 0 on one side and 254 on the other. Therefore any combination in between is possible with some rotation of the wheel.

[Cory]

Originally Posted by Hieb

True. My initial impressions (while I was writting before) was that with the steering wheel it was more of an either/or proposition--I can either have a tight turn (ex. left wheels = 127, right wheels = 254) or 0 turn radius (left wheels = 0, right wheels = 254), but not both as would be possible with joysticks.

Further thought on the matter has led me to an implimentation that would allow both--I guess I just like joysticks because with a rookie team that only has 1 programmer (and she has no previous programming experience), I'll probably end up writting most of the code this year and joysticks are easy to implement.

I am not a programmer. I understand the basics behind what we did, and the principle, but not much else.

Our wheel acts exactly as a joystick would. If we turn the wheel a tiny bit, the robot turns a tiny bit. It's exactly the same as if you pushed one joystick forward a tiny bit and one backwards a tiny bit. It's scaled this way throughout the travel of the wheel. Our driver vastly prefers driving the robot this way, and after 3 years he's gotten pretty darn good at it.

The wheel we use is the NASCAR Charger 2 by Thrustmaster.

[Alan Anderson]

Originally Posted by slickguy2007

I will try to find the brand of steering wheel that we use. It turns out that the steering wheel we have already has a serial port. After a little bit of tweaking, we were able to make the wheel function exactly how we wanted it to.

What are you connecting the serial port to? There isn't anything on the OI that can deal with a serial output from a controller device.

[slickguy2007]

Originally Posted by Alan Anderson

What are you connecting the serial port to? There isn't anything on the OI that can deal with a serial output from a controller device.

Yep. You are right. I meant it already had the 15 pin analog interface already. It was something we specifically looked for in a steering wheel before we bought one. It plugged right in with no changes necessary. We bought a discontinued wheel off of Ebay. I still haven't gotten the chance to check the brand yet though.

[Mike Hendricks]

Originally Posted by Cuog

For those who feel that the wheel will not give the same control options, I would have to say you are thinking too much inside the Joystick. The wheel can give you differences in control from only one or two pwm to 0 on one side and 254 on the other. Therefore any combination in between is possible with some rotation of the wheel.

Further supporting the wheel ..

Think of the range of motion you have with a joy stick. 90 degrees, at best. With a steering wheel, you have over 200 degrees of control (if not more, depending on your wheel). This was the argument supporting the wheel presented to us by a mentor from Team 60 (after 60/254's extremely successful 2004 season).

It is MUCH easier for students to learn how to drive the wheel than with (a) stick(s). And that's not limited to students, if you have a sponser, little kid, future team member, etc .. driving, it's much easier for them to get a feel of it without crashing into something or someone.

With that said, 1388 has used a wheel in 2005 and 2006. In fact, we already have our wheel for 2007. It's a Saitek R80, and we've got all 3 of ours new off ebay for less than $15 each. There are plenty of USB wheels out there, and if you have a rich team that wants a pretty wheel, then consider the new IFI USB dongle for the OI.

There is some fine tuning on the driving code needed for the wheel. There is also a lot of room for improvement. We have added features into our wheel code that allows for much easier turning (pretty much turning in place). There is also a switch on our 2006 control box, that lets you jump from a wheel/joystick combination to a single stick for driving.

[Jon Jack]

Originally Posted by PhatMike

Further supporting the wheel ..

Think of the range of motion you have with a joy stick. 90 degrees, at best. With a steering wheel, you have over 200 degrees of control (if not more, depending on your wheel). This was the argument supporting the wheel presented to us by a mentor from Team 60 (after 60/254's extremely successful 2004 season).

It is MUCH easier for students to learn how to drive the wheel than with (a) stick(s). And that's not limited to students, if you have a sponser, little kid, future team member, etc .. driving, it's much easier for them to get a feel of it without crashing into something or someone.

After trying a steering wheel for the first time this year I'd have to agree...

Having the steering wheel and one joystick for throttle made it much easier to turn, turn inside our footprint and drive straight (encoders of course helped out too).

We had a first year driver and it only took him a couple minutes to get the hang of driving around with the steering wheel. After about a day on our practice robot he started looking like natural on the wheel.

[John Gutmann]

I don't know if anyone has realized it but using a steering wheel and throttle setup is the same thing as using one joystick......

steering wheel = left/right movement of joystick
throttle = front/back movement of joystick

-John

[MrForbes]

maybe this is too obvious...but how about an r/c car controller? it has a steering wheel and throttle control, and driving an r/c car with one is very easy.

http://www2.towerhobbies.com/cgi-bin...I=LXPNM9**&P=7

for example

[Mike Hendricks]

Originally Posted by John Gutmann

I don't know if anyone has realized it but using a steering wheel and throttle setup is the same thing as using one joystick......

steering wheel = left/right movement of joystick
throttle = front/back movement of joystick

-John

For the most part it is, but from the teams I've talked to that run wheels (including ours), there are some "features" added to make driving easier (turning harder at certain positions of the wheel, jog L/R, shifting ease, etc). Main advantage of the wheel, you have a much larger range of motion to adjust your control.

Originally Posted by squirrel

maybe this is too obvious...but how about an r/c car controller? it has a steering wheel and throttle control, and driving an r/c car with one is very easy.

http://www2.towerhobbies.com/cgi-bin...I=LXPNM9**&P=7

for example

This is a summer project I've wanted to do for awhile, but have never got around to doing it - using a mini EDU controller and the r/c car controller to try and drive the robot. Sadly, I don't have the programming background to set the EDU up to do it. It would be really cool to be able to eliminate the full sized RC and OI for doing demonstrations and stuff where the full control system isn't needed.