R/C car pistol grip style controller

[yash101]

I'd like to ask, but why do you want a pistol-grip controller? Is it easier to use? What is the advantage of using such a controller?Are you going to have Ackerman Steering? It seems cool!

What if you take apart a regular joystick, and build another enclosure for the pots? You can re-engineer the joystick to look and feel like piston-grip!

Think like an engineer! Think about how you can modify what you HAVE to get what you WANT!

[MrForbes]

Think about how much time it takes to make a controller housing with all the moving parts the way you want them, vs. buying a pre made housing?

We initially developed the RC car controller concept to try a prototype Ackerman steering robot for the 2008 game, during the first week of build...then decided not to use that chassis design, but on a whim we tried the controller with a conventional skid steer chassis, and it was amazingly easy for even a klutz like me to drive.

My sons passed the concept on to the teams they mentor....

[yash101]

Yeah. That is true. However, putting together small pieces of aluminum with potentiometers attached and tons of bearings doesn't seem overkill. It would look weird, but would be nice to play with. The controls would move more easily! You can make sure to make the device best for the driver. One things FIRST is about is: "The robot helps the driver drive. The driver doesn't help the robot drive". I believe that there was one team at the phoenix regional who won an award for that.

[EricH]

Originally Posted by yash101

Think like an engineer! Think about how you can modify what you HAVE to get what you WANT!

Incorrect. Engineers do not think about how to modify what they have to get what they want, nor do they particularly think about how to get what they want.

An engineer will determine what is NEEDED first; any extra features are a nice-to-have bonus (what they want). The engineer will then look at COTS items to see if there is one that will fit the bill. He or she will get as close as possible using that before doing any modification. OR the engineer will determine that no COTS item will get "close enough" and create a custom design, incorporating COTS items like bolts and nuts where possible. Only if the engineer cannot get all the way with COTS or custom builds will he or she modify what's on hand to get the result.

Doing it any other way than evaluating all options is not going to be productive in the long run.

[kuraikou]

Pistol grips are probably the controller that I would recommend to any and every FRC team, they are extremely easy to use, offer great control, and are relatively easy to set up.

For our 2013 season we used a pistol grip controller that we bought from a local hobby shop and took all its guts out. We then took a game pad circuit board and wired all the pistol grip controls to the axis on the game pad so that the right axis was controlled by the turny bit on the pistol grip and the left stick was controlled by the trigger part of the pistol grip, so essentially it was a arcade controller with a different shell. We then used the rest of the buttons on the circuit board to control our shooter.

If you need more info our programmer who worked a lot on this would probably be happy to help you.

[Abhishek R]

Originally Posted by yash101

I'd like to ask, but why do you want a pistol-grip controller? Is it easier to use? What is the advantage of using such a controller?Are you going to have Ackerman Steering? It seems cool!

What if you take apart a regular joystick, and build another enclosure for the pots? You can re-engineer the joystick to look and feel like piston-grip!

Think like an engineer! Think about how you can modify what you HAVE to get what you WANT!

The control system should match what the driver feels most comfortable with. Having played around with joysticks, gamepads, and steering wheels, I find the gamepad most intuitive to pick up - using one stick forward/backward, the other left/right - because I played around with [not very sophisticated] R/C cars when I was younger, then enough Forza kinda locked in that decision. People used to the grip-style controller from R/C cars will find that most natural to them.

Some people just prefer to use the joysticks, it makes more sense to them, or another form and control. If you use what feels most natural to you, when a time comes to make a split-second decision you don't have to think because your "instinct" will make a move - no need ot worry about if you're gonna go left or right, it will just happen.

On another note...why would you take a complicated route for no real gain? Rather than modifying a radically different setup into a makeshift R/C controller, I think it would be extremely more efficient to take an existing controller and just change that around to be coded to the functions of the robot. Yes you will have to take it apart, but it will be infinitely easier and you'll end up with a better final product anyway.

[yash101]

I agree that overcomplicating things would be a bad idea. Even better would be to make a glove full of pots so you can control your robot by moving your hand around! That's what I'd call magic!

[Andrew Schreiber]

Originally Posted by yash101

I agree that overcomplicating things would be a bad idea. Even better would be to make a glove full of pots so you can control your robot by moving your hand around! That's what I'd call magic!

It's been done... a few times. I'd assert it's not the most intuitive control system. Something about how mapping X input into Y output isn't obvious to a casual observer.

[Rangel]

From my experience as a driver for 2 years on 842, the pistol grip controller was extremely intuitive. I had been practicing for a month with joysticks but was nowhere near as good as I was with the pistol grip. With gyro assisted drive and even just a moderate amount of practice, the pistol grip can turn anyone into a top tier driver. Last year we decided we didn't want the cypress board anymore so we took apart the guts of a regular gamepad controller and mapped the potentiometers from the pistol grip controller to to the joystick inputs. After doing this, the computer will just see the pistol grip as a regular gamepad.

[Tyler2517]

We have messed around with implementation a pistol grip controller in the past, when we started using holonomic drives we could not use the full functionality of them and wanted to keep the controllers constant between all drive types. If you are using a Ackerman steering system it would be very user friendly.

[yash101]

Originally Posted by Tyler2517

We have messed around with implementation a pistol grip controller in the past, when we started using holonomic drives we could not use the full functionality of them and wanted to keep the controllers constant between all drive types. If you are using a Ackerman steering system it would be very user friendly.

Yeah. It seems like the best controller for Ackerman steering because you need a control for the speed and a control for the turning wheels. I guess this could work with swerve, but wouldn't be very intuitive. Otherwise, as I have been reading, for regular Ackerman and tank drive, it seems like the best control. How would you code tank drive to use this? Would you have the accelerator controlling the speed of both sides, and the turn wheel controlling the delta in motor speeds?

[Oak1477]

Our team created a rc remote for last season to drive with. We used a Spektrum dx2e remote. It was gutted and the potentiometers were wired in to an arduino uno. We found some code that would make the arduino be recognized as a logitech joystick so it could be easily changed in case of emergency. We did it because it was driver preference but we have found it is easy to learn.

[Alan Anderson]

Originally Posted by yash101

How would you code tank drive to use this?

You wouldn't. You'd code it as arcade drive.

[TheOtherGuy]

Originally Posted by yash101

I guess this could work with swerve, but wouldn't be very intuitive. Otherwise, as I have been reading, for regular Ackerman and tank drive, it seems like the best control. How would you code tank drive to use this? Would you have the accelerator controlling the speed of both sides, and the turn wheel controlling the delta in motor speeds?

Swerve is the drivetrain, the control type (holonomic/omnidirecitonal/etc) requires 3 inputs: translation in the x/y plane and rotation. You have only two inputs with the pistol grip, so you cannot control all three DoFs of an omnidirectional setup with a pistol grip controller.

Arcade and tank are two control types as well - the drivetrain is a skid-steer. You are mapping two inputs to two outputs: arcade requires a linear input and a rotational input, and tank requires two linear inputs. The output is the same for both (left and right sides of drivetrain). Since you are using a wheel (rotational) and a trigger (linear), you are using "arcade" style control and should code it appropriately.

I'll add myself to the list of pistol grip supporters.

[Andrew Schreiber]

Originally Posted by TheOtherGuy

Swerve is the drivetrain, the control type (holonomic/omnidirecitonal/etc) requires 3 inputs: translation in the x/y plane and rotation. You have only two inputs with the pistol grip, so you cannot control all three DoFs of an omnidirectional setup with a pistol grip controller.

Arcade and tank are two control types as well - the drivetrain is a skid-steer. You are mapping two inputs to two outputs: arcade requires a linear input and a rotational input, and tank requires two linear inputs. The output is the same for both (left and right sides of drivetrain). Since you are using a wheel (rotational) and a trigger (linear), you are using "arcade" style control and should code it appropriately.

I'll add myself to the list of pistol grip supporters.

You're making a big assumption - maybe the swerve isn't controllable in one of those outputs… Like say a certain 148 robot that could not turn…