Gary, the 4WD experiment

Let's Make Robots Robots
1

After some years following different DIY sites and forums, I decided that it was time to jump in.

I choose to start with a premade platform to ease first steps (which was wrong, but I discovered it later ;) )

The 4WD DFRobot platform is well made, instructions are available and quite complete:

unboxing

Step after step, the work started well:

motorsplatform assembled

Add the wheels, put a controller board:

cables in placebase montée et cablée

 

I decided to post on LMR at this stage but in fact I worked further since the last picture.

I'll complete with my Sharp IR installed and of course... video or it didn't happend ;)

My problem is now on the choice of a 4 wheels platform. The motors are controlled side by side and the result is that it doesn't turn very well.

I wonder if it will be best for me to find a way to replace rear motor with simple axis or if I try to add a second pair od motor driver to learn how to implement full 4WD...

Later I will either add my work on understanding how to build a pan tilt, or you can look at my Flickr to see the wood prototype I've done.

Update 01/12/2014:

Just a quick picture of Gary with its Sharp IR sensor

 

Sharp IR

no more for the moment, I choosed to work on another project this week end and discovered the miracle of hot glue ;)

Update January, 28th 2014:

I added the video I have, even if it's with the previous Romeo that was bugged.

So now, I can claimed it: it happend! ;)

Update February 23th 2014:

That's been a long time since last update. Maybe I should experiment less and finish some :)

I always wanted to add wireless communication to Gary and my Raspberry Pi wants to travel: Deal!

I bought a nice RaspiRobot kit, but it comes like this:

Here we go!

First time for me to solder so nice but tiny things :)

I think I got it pretty well:

Everything solderedI think it's acceptable

I adapted the case I have with some little drill game:

Gary got a new brain

I forgot to take a picture with the Pi and the Wifi dongle.

The python code works and I will translate it to Golo-lang or Java.

I also have some camera streaming code working, but the Pi camera is really not a good piece of hardware. I'll search another solution.

 

  • Control method: autonomous, (remote interaction planned)
  • CPU: Arduino Romeo
  • Power source: 5AA
  • Programming language: C
  • Sensors / input devices: Sharp IR
  • Target environment: indoor flat surfaces
This is a companion discussion topic for the original entry at https://community.robotshop.com/robots/show/gary-the-4wd-experiment
2

It seems that I didn’t

It seems that I didn’t explain my problem very well :wink:

I know that I can drive the motors as you describe. It is even documented by DFRobot this way.

My problem is the fact that this configuration implies that there is some skid on the tires that prevent precise turns.

After some research, I saw 3 solutions:

- no rear motors, free wheels and a steering mechanism on the front wheels/motors

- remove rear wheels and put a ball caster, as a 2 wheel platform

- replace wheels with tracks

3

Need more Power Mr. Scotty!

You will need to give the motors more voltage to get good turns.  Look into Nmh AA rechargables or a 7.4v Lipo pack.

4

I have 5 AA so I’m near the

I have 5 AA so I’m near the 7.4 lipo. In fact, I have a 7.4 lipo :slight_smile:

But the mechanical problem will still be there, it will just skid harder and quicker :wink:

5

omni wheels?

I have the same issue,  bumping the power did wonders, but it still doesnt turn great.  I changed the code so it doesn’t try to “spin” in place but makes a more gradual turn with the one side running half speed… however, I keep wondering if there are omni wheels that will fit the motor wheel mount.  seems like it might be a good solution.

6

I wonder if tracks will

I wonder if tracks will solve the issue, or if turn with tracks implies some skidding too.

Anyway, I also ordered a ball caster to give it a try and… I started a somethingLikeChopStick.

hexapods are in my mind now :slight_smile:

7

by the way, Omni wheels for

by the way, Omni wheels for DFRobots servos exists: 

https://www.robotshop.com/en/60mm-omniwheel-w-hub-servo.html

But the frame needs some modification first in that case

8

tracks may be good

my first project had tracks and turns really well.  I think the big difference is the material… rubber wheels vs plastic tracks that slide but have low traction.

9

I think I can still learn

I think I can still learn other things and keep it as it is.

I will see things like put my 7,4V lipo battery (keep connector or not), replace the IR with 2 Ultrasonic but using pan servo…

In the meantime, I believe I found the next project during my tries with a kind of Chopstick quadruped. All I want to experiment now is hexapod :slight_smile:

 

So expect some improvment here, but not especially on the wheels at least for now.

24h days are so shorts :wink:

10

Nice!

We built a similar project.  However, we installed an Android device onto a RC car and used an IOIO board instead to control it autonomously. Here is the preliminary result:

http://youtu.be/vkvkfcqEUkk

11

I’ll post an updated front

I’ll post an updated front picture, but if you look at the last update, you will see that the brain is now a Raspberry.

I used a board called RaspiRobot to drive the motors and the sensor.