Date: Tue, 24 Jan 2006 22:49:09 -0600 (CST)
Subject: replace battery packs and steering motor
X-UID: 128
Content-Type: IMAGE/JPEG; name="img2570.jpg"
Content-Type: IMAGE/JPEG; name="img2572.jpg"

My original plan is to use individually charged NiMH AA cells to power the 
electronics. AA batteries have the highest energy density as most research 
and development is in this size due to demand for consumer electronics. An 
intelligent Maha charger then charges each AA cell which are then packed 
together in battery holders.

This can't work.

The huge voltage drops attributed to internal battery resistance yesterday 
are really contact resistances. Each spring loaded metal to metal contact 
point adds resistance. The .5 ohms per cell measured is roughly accurate. 
But the resistance is mostly in the contacts and wiring.

A 19.2 volt battery pack with 8 ohms of internal resistance is almost 
useless. That's what I've made. It is just enough to power one of the 
computer boards alone. But a computer board with a PCI card overloads it. 
I even heard popping from the pack which means something became very hot.

This makes sense when you think it is drawing on the order of 1 to 2 amps. 
That works out to a voltage across the terminals of 19.2 - 1 * 8 = 11.2 
volts to 19.2 - 2 * 8 = 3.2 volts. Power dissipated is from 1 * 1 * 8 = 8 
watts to 2 * 2 * 8 = 32 watts!

So how does anything run off batteries at all? The answer is simple. The 
number of contacts and lengths of wires are kept to a minimum. A battery 
pack has all cells soldered or welded together. The battery pack becomes a 
monolithic unit. This is not done only for convenience. It is a necessity 
to avoid large internal resistances.

The right thing to do at this point is use standard R/C battery packs to 
power the onboard computers and WiFi router. This is a prepackaged battery 
power solution.

I've decided that I'm going to replace the steering motor. It does work. 
But it also becomes stuck too often. I think the quality is just too low. 
I need a low voltage cordless screwdriver with a better gearbox. What 
attracted me to the one I have now is the square body. At the time, that 
seemed to me easier to mount on the robot than a cylindrical shaped tool. 
And it was cheap. Now I think it was too cheap.

My goal is to get this robot in the parking lot and rolling under radio 
control this weekend. I'm still swapping out a lot of parts which worries 
me. But I think I can make it if there are no major electrical accidents.

I go off-call tomorrow morning. So I have Wednesday evening to pick up the 
battery packs and new steering gearmotor (have to find these - what to 
buy, where?). Then I should get these on the robot, wired, and powering 
the electronics Thursday evening. Friday evening is final assembly. 
Saturday is function testing in my apartment. And Sunday is initial 
testing in the parking lot.

On the good side - it's funny how much this is like some of those robot 
competitions I've read about. People are always working on their machines 
at the last minute, sometimes just before their competition run. Very 
basic things will fail.

Many robots I've read about use somewhat specialized parts whereas mine 
has gravitated towards major components available from Radio Shack, 
Northern Tool, Home Depot and Walmart. So for last minute changes, 
sourcing parts is easier. Being able to hold things in the hands is very 
helpful.