Date: Fri, 27 Jan 2006 02:25:21 -0600 (CST)
Subject: SLA battery and new steering motor
X-UID: 129
Content-Type: IMAGE/JPEG; name="img2573.jpg"
Content-Type: IMAGE/JPEG; name="img2575.jpg"

For $20, I found a 12 volt 4.2 amp-hour SLA (sealed lead acid) battery as 
a Christmas toy leftover. That includes a slow charger. This will run the 
digital side for at least two hours assuming the specifications are 
correct. Weight is around three pounds which is acceptable. It's a little 
tall but otherwise fits under the electronics box with the existing mount.

The community that closely parallels what I'm doing are amateur radio 
operators. They typically have a basic knowledge of electronics and are 
often concerned with adapting existing COTS (common off the shelf) 
technology to mobile and portable use in the field. This involves problems 
that the electronics hobbyist is unlikely to encounter.

Specifically, I found a clear explanation of why my homemade battery packs 
didn't work. Others have tried this with typically poor results. The 
battery contacts and thin gauge steel wires in the typical plastic battery 
holder introduce too much resistance. In my case, there was so much 
resistance that the wiring inside the pack was becoming very hot. This 
explains why the voltage would drop very quickly. The hot wires had 
greater resistance, which increased power dissipation, generating even 
more heat, etc. Eventually, the resistance was so high that voltage dipped 
below the cutoff for the DC-DC power supplies on the computer boards.

I have some industrial style solder tab battery holders that probably will 
work. They're still not as efficient as spot welded commercial packs. But 
I have lots of NiMH batteries and these battery holders on hand so will 
probably make a bipolar pack (-6 volt, gnd, +6 volt) for the steering 
motor and cooling fan. However, right now the pack I have does seem to 
work so I'm going with it for Sunday.

This time, I'm trying to transfer lateral loads to the screwdriver body. I 
think that hanging the other screwdriver off the shaft causes the gearbox 
to seize up. So I found a 7/16" ball bearing (the OD of the shaft is 
.431). By happenstance, it fits perfectly inside a 1 1/4" compression 
coupling. It's like it was made for it.

Mating the coupling to the screwdriver body was more problematic. I had to 
shorten it by sawing off one end and then facing it off in the lathe. It's 
almost too big to fit in the chuck. Then it was degreased and epoxied to 
the end of the screwdriver with JB Weld. This steel epoxy has one noteable 
property that makes it very useful. It doesn't run like most two part 
epoxies do.

The bearing end of the compression coupling screws off. So tomorrow I'll 
remove it and pour some more epoxy between the coupling and the 
screwdriver. I don't know if this will hold up as in my experience epoxies 
have a difficult time bonding to many plastics.

So I'm on schedule. The task list for this evening.

1. charge the NiMH drill battery packs for Sat testing and Sun field trial
2. reinforce and reposition the battery mount for the SLA battery
    (shift it down with rubber washers and reinforce with aluminum bar)
3. mount SLA battery with industrial velcro
    (have to degrease surfaces first)
4. solder wires into the new cordless screwdriver
5. mount new screwdriver in place of the old one
    (may defer until Sat depending on additional epoxy cure)
6. check/modify electronics box wiring - no shorts or loose connections
7. close electronics box and mount on robot
8. basic function test of electronics, as much as fits in available time

Saturday will be testing and adjustment.

For Sunday, I'll use WiFi, ssh into the onboard computers, and run the 
robot with shell scripts. It'll perform a fixed sequence using what I've 
got now, nothing really new.

A loose end I've almost forgotten - the webcams. These must still be 
mounted. I'll have to fit in Friday or Saturday sometime.