The first brainwaves of a rookie robot builder wannabee

Let's Make Robots
, , ,
1

So, let me start by saying that I really like LMR already.
For a couple of months now I've been wanting to start playing with robotics, micro controllers and pick up electronics again.
I did some electronics back in school and as a hobby but I didn't have time.

So, a while back I bought an Arduino Uno to start playing with uC's and electronics, drove my girlfriend mad with talks and showing designs I made, which where all not feasible or would be way too expensive to even think about.

Now you will probably think, stop rambling and start talking robots! And I will!

 

I decided that I will be building a 2DOF hexapod. From scratch.
I looked at kits from different suppliers and they all look totally awesome. More awesome than I could ever build myself, but still I want to do it all myself.
I haven't finalized any designs, but I did decide on what electronics I will be using.
I came up with the following:

  • Arduino Uno (Might upgrade to a Mega if I don't have enough room for the program)
  • Lynxmotion SSC-32 servo controller (Which is pretty easy to control with the Arduino)
  • 12 Servo's for the legs
  • 1 Servo for the head which will contain an SRF04 ultra sonic rangefinder
  • 5 cell battery pack to power the Arduino and SSC-32
  • 6 cell battery packs to power the servo's (might go with LiPo, not quite sure yet)
  • A stack of LED's for pretty and status indicators.

 

Next week I will be ordering the first electronics and start building some of the parts, to get familiar with the SSC-32 and creating the parts.

I still have to make my final choice on what material I will be using.
I have three options of which one has been ruled out already for the final assembly.

  • Wood: I ruled this out as a final material, since it is heavy and pretty expensive for the good stuff. I might use wood for prototyping some of the parts, like the legs, since it is easy to work with.
  • Aluminum: I really love aluminum but without really fancy tools it is going to be tough to get the show on the road. 
  • Acrylic or Lexan plates: Not too expensive, tough and easy to process without expensive machinery.

It's all gonna be handy work since I don't have access to fancy CNC equipment or anything. Dremel, hand/electric saw, power drill, sand paper, etc. is gonna be what's going to build this puppy.

I hope to finalize my designs this weekend, then get some good cardboard to build a first scale model (Actually a second one, but the first one is of a previous design which is not going to do what I want it to do) during the coming week, while waiting for my electronics to come in.

I haven't set myself a goal time wise for a couple of reasons:

  • I don't have all the money in the world, so I will be buying my parts piece by piece and build as stuff comes in.
  • I will be learning as I go since I don't have a lot of experience yet.
  • I will be manufacturing all the parts I need, besides the electronics, by hand, which can be a pretty daunting task. For example, six legs means 2x 6 equal parts, with very little margin for error.
  • I'm not in any rush to finish. It's a hobby, I have limited time and I want it all to work perfectly. Good enough is not good enough, it needs to work exactly like I have it in my mind. If that is not possible, I'm gonna need to make it possible.

 

So, that's my first blog entry, which is probably very uninteresting for most people.
I like writing stuff like this though, so you can expect more from me (if I don't get banned for writing bs like this :) )

Regards,

Benji 

2

Cool

Can’t wait to see it come to life.

3

Nice project! Can’t wait to

Nice project! Can’t wait to see it working!

I am working on a similar one, a 12 DOF quad. Same number of servos, but less legs. I will also be using an all in one controller board, based on the Arduino micro (ATmega328). You will be surprised how much code fits in there… No need to go for the Mega unless you want to drive 20 servos directly (a biped for example). The Servo library can control up to 12 servos, so for the sensor pan I plan on using a pico servo, for wich I will generate the pulses manualy. The idea is to set the servo pin High for the required amount of microseconds, then turn it Low and run the rest of the code (check the sensor, run the steps of the gait, etc) until the loop time of 20 miliseconds is met, then start again. The Servo library works in the background, so the timing for the pan servo might be at times off, but it is not so important to hold position. I am in the process of building the legs, hopefully I’ll have something working next week. If you want to take a look at my board, see it here: https://www.robotshop.com/letsmakerobots/node/24522

Cheers!

4

Thanks!

Thanks for your reply.

I know about the 12 servo limitation, that is why I am using the SSC32 board.
With this I will not have to use the servo library, but I will be able to send commands to the servo’s over a serial connection.
It will not be the fastest hexapod there is, but for a first robot it will be enough.

I’ll check your page out :slight_smile:

5

Great writeup

I enjoyed your post very much.  I admire a builder with a clear idea of what he wants to do and how he plans to do it.  Looking forward to seeing your progress.