Daisy Build Report 2

April 29, 2001

Despite some setbacks, building is proceeding at an acceptable pace.

The biggest setback was our giving up on the homemade speed controllers; they are malfunctioning and there isn't time to debug them right now. Shows what you get using a schematic off of the internet! Still, it's been educational and I bet we'll get them working eventually. Just not for SF.

So instead we are going with Innovation First Victor 883 speed controllers, which seem to be working just fine.

As a side effect, the overpowered and complicated microcontroller setup has been discarded as well for now, which costs a little flexibility but should pay off in reliability.
 

 
First step: attach the wheels to the motors. For wheel hubs, we used some QD bushings from McMaster Carr, which we were able to bolt directly to the wheels. After that, all we really had to do was cut a keyway into the axles and add some shaft collars to stop the wheels from moving around.
 
The wheels and hubs, on the motors, with the motor mounts.
Once the motors were attached to the wheels, we attached them to the base. There are bearings (not ball bearings, just oiled bronze bushings really) supporting the wheel on either side so the motor shafts don't take any stress. The bearings and the motor mount are supported by a 1/4 inch piece of rubber sheet, which allows everything to move a bit if the axles get slightly bent or otherwise damaged.

Vacuum cleaner belts are stretched around the wheels and get pretty good traction. They still need to be glued and screwed into place but we will surely destroy one set during drive testing so we'll wait for the glue anyway.

The little protrusion on the bottom of this picture was for a contoured front, which has been discarded now in the interest of saving weight.

 
After cutting aluminum sheet into armor, 3/4 inch aluminum tube into frame parts, and using about 70 hardened steel machine screws, Daisy really started taking shape, so we wired it up messily and took it out for a spin.

It is still only 34 pounds so we can't say definitively but it seems to be pretty quick and durable -- smashing into things doesn't seem to hurt it. Didn't measure the speed exactly but it seems to be about the expected 10 miles per hour or so.

 
We are using an Ohmark mixer in place of our old PowerPC control processor. It seems kind of touchy, which I had pretty much expected -- with the old processor we were free to program an extended dead zone and an exponential response curve for more precise turning, but it looks like we'll have to replace that with driving practice. It wouldn't be too hard to make our own mixer out of a tiny AVR chip, that could do most of what we want, but time is getting short.

 
So really we could stop and go with it pretty much like it is, but what fun is a boxy wedge? The box is really supposed to be a carrier for the weapon.

We have this nifty standard-issue EV Warrior motor and a bunch of half-inch shafts, so we got a 5/16 to 1/2 spacer/bushing thing, reamed it out a tiny bit to fit the metric motor shaft and cut out some slots so it will slide into the pin sticking out of the motor shaft (that rolled spring steel pin seems pretty strong to us so we won't bother replacing it with something else).

 
The motor will attach to the weapon with a half inch (V4) belt via 3 inch pulleys (1:1 ratio). Hopefully the pulley/belt mechanism will allow a little slippage and reduce the shock on the motor when something crunchy gets in the way. Ideally it will even prevent the motor from stalling to limit the current a bit, but that might be asking a bit much.
 
Here's what we're talkin' about!
This assembly weighs about nine pounds. The sharpened stainless steel "petals" sit in grooves we cut into the barbell weight so they can't go anywhere and then they are securely bolted on tight with six military spec steel bolts.

Our only worry is that it will be too heavy for the EV warrior to spin up energetically. Time will tell!