Introduction: High-Torque Steering Mechanism for Really Large Remote Controlled Toys

This 'ible leans heavily on instructions given in my previous 'ible on building a pannable vision system. As such, it's a little less step-by-step and more a photographic tutorial on the concepts involved.

The position sensor feedback circuit used in this steering mechanism is the same as that used in the previous 'ible, a simple potentiometer.

This is the steering rig I build for my ATRT (All Terrain Robotic Trike) vehicle. You can see that in action here, if you like. The panning vision system that I covered recently was built later, to be an add-on module for the ATRT, so covering this subsytem now is kind of back-tracking for me.

I think this system could be adapted to other large 3-or-4 wheeled remote control vehicles or robots. Adding a suspension system or all-wheel-drive would probably complicate the system of tie rods needed, but it could still be done.

Step 1: Build Steering Knuckles

Ok, this step is not an easy one. You will need a fairly large sheet of scrap sheet metal, and several pieces of additional metal to reinforce the mounting points for the tires.

If you have invested in a welder and a cutting torch, then you will probably enjoy this step a lot more than I did. I did my cutting with a metal blade in a jigsaw, and by hand with a hacksaw. I used a power drill for all of the holes, and lots of nuts and bolts in lieu of welding.

I did get a good deal on nuts and bolts: 80 metric M6 machine screws and matching nuts in a convenient organizing case at the local dollar store. I guess it's because they were metric, that the normal retail stores didn't buy them.

Feel free to use or adapt the imperfect hand-drawn pattern that is shown.

Step 2: Create the Front End of Your Vehicle

I am using a single 2x6 piece of lumber. I chose to make my wheels about 24" apart. I used both wheels from a discarded BMX bicycle.

I would drill the holes for the pivot-axles of your steering knuckles next, then, using the "arms" of your steering knuckles as a guide, measure out what length of tie rod you will need, and create one from a 1/4" solid steel rod.

I had to round the corners on my lumber, for clearance. It also has cut-outs in the middle for other parts to fit through.

Step 3: Hack in a Gearmotor to a Worm-drive System.

This assembly took me all day to come up with, build, test, and get working sufficiently. It went through a few more revisions, after it broke a few times. What you see here is the end result.

The assembly will have to be able to pivot, too, to stay aligned.

Step 4: Add a Position Sensor.

I would again like to point my readers to this 'ible for details.

What this position sensor does is provide an analog voltage that an MCU can read.

In a more traditional R/C system, if you are not using an MCU, you might be able to wire this onto a servo-control IC like the one in this schematic that I found at the excellent site www.seattlerobotics.org

Step 5: Attach the Front End to the Vehicle, and Do Some Wiring.

I attached the 2x6 to the base of my ATRT with decks screws.

It's best to do this before mounting other parts to the base, because apparently more than one object cannot easily occupy the same region of space. Where's Geordi LaForge with a particle beam when you need him, right?

You'll need to wire power in to your H-Bridge board.
Wire the motor power to the H-Bridge board's output.
Wire signals lines from your MCU / Servo IC to the H-Bridge.
Wire the position sensor's potentiometer to your MCU / Servo IC.

The wiring that I built for my ATRT is shown in the photo, but simpler, "bird's-nest", point-to-point direct-soldered wiring is fine, too. I made my cables more robust because of the generally modular nature of the robot.

Thanks, to anyone who reads this 'ible.

Comments

author
vareah (author)2010-09-07

I am currently doing a project similar to this project.Programming of PID Controller on Vehicle Platform via Bluetooth. I am required to upgrade the steering with feedback. any recommendation??

author
psymansays (author)vareah2010-09-07

Well, the potentiometer on this acts as an analog sensor, for the MCU. You could use that same solution, or an automotive throttle position sensor, or a hall effect sensor (they use those for electric scooter throttles), or even use an optical encoder strip, and count the clear/opaque bands on the strip as they're fed through an optical sensor.

There's a lot of solutions to the problem, but they're all similar in that they rely on a simple position sensor system. Most likely, the best solution for simplicity and accuracy is an automotive throttle position sensor.

author
snowpenguin (author)2009-05-27

I saw this and thought, RC Office Chair, w00t!!! Oh well.

author
psymansays (author)snowpenguin2009-05-28

Oh, I see...that sounds like a fun project all on its own :) But wouldn't a motorized office chair be the same thing as an electric wheelchair?

author
snowpenguin (author)psymansays2009-05-28

Ummmmmmmmmm..... No... Maybe.... Yes? I don't know. It would be pretty cool though. Just imagine ,you're sitting in your cubicle and you look back only to see an office chair drive by.... With no one in it!

author
lemonie (author)2009-02-14

If I understand this correctly the motor unit uses the thread on the steering rod to crank it along. You've got a hex-nut on the rod with the blue gear teeth on the outer which driven by the motor and held in place by metal plates. How fast does this move, e.g from central to full-lock? L

author
psymansays (author)lemonie2009-02-14

Well, mine is pretty slow, because the motor is fairly low-powered. It can go from center to full-lock in about 2 seconds, if the vehicle is moving forward. If the thing is stopped, the motor turns more slowly because it needs to make more torque. While slow, it is drivable. A motor with more RPM's and more gearing-down would move the whole thing much more responsively, though.

author
lemonie (author)psymansays2009-02-14

It's a fairly tight pitch thread, and none too smooth I guess. But seems to work OK nevertheless. Had you considered a cam mechanism instead? I think that whatever is used, like you say, a more powerful motor would give better response. L

author
psymansays (author)lemonie2009-02-14

I had not considered a cam system. A rack-and-pinion setup was my first idea, but I had a hard time finding those parts. I did use a tap and a die to smooth the threads on the rod and inside the hex nut, and lubed it up, so it moves rather smoothly. A motor from a cordless drill or an R/C vehicle might fit in this mechanism, to give it more speed, though. The all-thread rod that I used, has angled, standard threads. Most worm-drive rods use "flat" threads that are not designed to lock a bolt in place like machine bolts are, so that does decrease the efficiency of the worm-drive system.

author
lemonie (author)psymansays2009-02-14

This is a solid worm-type mechanism, it's not going to shift unless it breaks, so that's a good feature. If you had one you'd want to worm the can (pun there?) but you'd probably find the parts hard. I like the use of bits here - it's simple and it works. L

About This Instructable

12,274views

29favorites

License:

Bio: Hi, I'm psymansays. I'm an engineer from California. I enjoy sunsets, and long robot test drives on the beach. More from me: http ...
More by psymansays:Patio Garden Wicking Bed From 55 Gallon Steel DrumThe Bare Minimum DIY LatheBrowser based teleoperation for the LaptopWheels project
Add instructable to: