It was a trying day for me on Thursday, after completely assembling MDB (I can't be stuffed typing MrDotBoto anymore) only to find that there was no way to turn him! Why? Because I was suffering a problem known as maneuver drag! WTF is that?
If you have four wheels setup in a square, like MDB had, and try to maneuver the chassis by making one motor wheel go forward and the other go backward, to perform an about turn, then depending on the thickness of your tyres and the amount of tyre making contact with the surface below, you will create drag. If you have enough drag, and not enough power, you will not be able to turn!
So I toddled off to the local hobby store to see if they had any parts they could provide to setup some steering! Now, the only issue I knew I was going to have to face was not in setting up a steering system, but how to control it using the Serializer.NET board...?
Well, the hobby guys had more than enough bits and bobs to completely create a steering front end with suspension, but they all required a different chassis setup, and none included the "square" aluminium chassis I had created...I don't think they knew what to do when I assumed the foetal position and started sucking my thumb.
Then the boss man gave me that look, not the one that has you back against a wall (that was to come later), but the one where a distant image comes into the foreground, and your problem and his solution align perfectly! He runs out the back and comes out with an old petrol R/C chassis with a complete front end (no rear end) with working steering linkage...and what's more, it's got a HiTec HS-311 servo already hooked up. Now for those just getting into the robotics/automation scene with a Serializer.NET board, it supports two (well 4, but with a jumper change) servo's out of the box! So I spent the weekend stuffing around with the R/C chassis to remount all the bits from the old aluminium square to the new R/C chassis! What did it come out like...take a look for yourself:
Now, the only issue is that I had to mount the Mini-ITX motherboard in a really sketchy way, using some small PVC piping (just made custom stem mounts) and as you can see from the pics, there are a lot of wires and stuff on the top. Also, the Serializer.NET board is mounted under the Mini-ITX board, and the two batteries are mounted under the R/C chassis. Not ideal, but certainly enough to get started.
So what's next...?
Well, now that everything is running completely free of umbilicals, I can start to see what happens when MDB has to wander. I've started experimenting with a few things which I'm already realizing will be a challenge, particularly:
1. How to segment the application architecture
2. How to deploy new versions to something that only provides the ability to copy files to a share (yes, MDB has wireless connectivity for software updates and to keep itself patched). There isn't a "SDLC" for robots that I can find for .NET...so will start to put something together around doing Builds to a robot.
3. Start working on the chassis design for MDB Mk.2. By the way, if there are any Engineering/Design students who would like to help me with MDB, I need help with the following:
a. CAD design for the next version of the chassis (this will have provisions for boards, etc, which are now mounted with a variety of mechanisms including PVC piping and cable ties!!!) b. AI for the robot itself (although MDB is going to get a serious mind jolt when we head up to the Robot Olympiad on the Gold Coast in December and catch-up with Joel so he can DROP MDB!)
a. CAD design for the next version of the chassis (this will have provisions for boards, etc, which are now mounted with a variety of mechanisms including PVC piping and cable ties!!!)
b. AI for the robot itself (although MDB is going to get a serious mind jolt when we head up to the Robot Olympiad on the Gold Coast in December and catch-up with Joel so he can DROP MDB!)