Date: Mon, 28 Mar 2005 22:53:20 -0600 (CST)
Subject: using ammo boxes for body
X-UID: 38
Content-Type: IMAGE/JPEG; NAME="tcsnap008.jpg"
Content-Type: IMAGE/JPEG; NAME="tcsnap009.jpg"

It looks pretty good in the CAD mockup with .30 cal ammo boxes for the body. 
This is practical. The US military tosses these things out in mass quantities 
so they are inexpensive at around $5 each (although it is difficult to find 
ones in good condition around here for some reason - may have to attend a gun 
show to find better ones). They are made of heavy gauge sheet steel with a 
baked on enamel or epoxy of some kind. They are also water tight and 
tremendously strong. Your US tax dollars at work here.

This simplifies some of the mounting of batteries and electronics. At worst, I 
can wrap them in foam and stuff them inside the boxes. The boxes are also a 
structural component and hold the frame tubes in relative position. 
Significantly larger .50 cal and 20 mm ammo boxes are also available. These 
would be advantageous as much larger items could fit. The .30 cal boxes won't 
even fit a mini-ITX board without some modification.

I've decided that I can't figure out completely the motor mount problem. If a 
team of dwarves can fumble for a few days trying to build a go-kart on Monster 
Garage, I think that I can do the same and defer the motor problem. I'm 
thinking that I will have to get two right angle drive attachments and then 
mount the motors along the axis of the frame.

I am concerned that I am growing weary. Where is the "I will stay awake for 
three days" energy? Now I'm thinking about going to bed with the pleasant 
memories of the very nice meal I had for lunch at Fadi's. Impressive robots are 
not built with such thoughts.

So it's a good thing that now I can start fabrication. I have to make an 
itemized list of the suspension parts or else I will become confused and make 
something twice or not at all.

I think that I should finally read that QNX Neutrino (real time OS) book I 
have. I finally read carefully an article that a friend sent me. One of the 
DARPA Grand Challenge teams used Windows and .NET for their systems software. 
It is a different perspective. They tried to create a full software abstraction 
for the hardware and move all of the real time functionality to as high a level 
as possible. It reminds me a little of telecom and the object models the 
systems engineers tried to apply to network elements. I think I need to read 
more about how other teams have approached this problem to gain some 
perspective.