BucketBot Power Supply and Distribution

Loads

The DC power loads can be grouped as follows

Traction Motors

High peak currents (tens of amps), lots of switching noise on the bus from the PWM controllers

Computers/Controllers

Fairly low current (tens of mA) The current design uses a RabbitCore which draws very little power(<1W), but if, in a future incarnation, I retrofit some form of laptop to the robot, a different power distribution and supply scheme might be necessary.

Peripherals

Some high currents, but switchable. Examples are power for the servos (hobby type) controlled by the SSC and the video camera.

Radios

Small current, except when high power is needed (transmitting video, for instance).

Batteries

Primary power is provided by a set of 12V 7AH high rate "gel cells". These are the usual "bricks" that you see all over the place, and are a model designed for higher rate discharge (but not as high as the "UPS duty" ones that are designed to be discharged in 20 minutes).

As I add the processor, more batteries will be added. Multiple batteries, although adding weight, have proven to be more reliable for the following reasons:

• The power source for the traction motors (and, more importantly, the PWM controllers) tends to get a lot of switching noise, which wreaks havoc on the power for analog and low power digital circuitry.
• If the traction motor batteries go dead (or the fuse blows, etc.), you still have power for the processor and radios, so you can "call for help".

On Roger, I used some 24V DC/DC converters to generate isolated 5V power, but, the converters weren't very efficient (large unloaded static current draw) and they introduced noise of their own. I have some newer Vicor 24V to 5V DC/DC converters which are more efficient which I might try using.

The issue that hasn't been resolved yet is whether to use 4 batteries all told, 2 for traction power, 1 for computers, and 1 for radios.

Charging

For now, I charge the batteries separately using an off the shelf external 12V battery charger. It would be nice to have a single "charging port", but that would require switching the traction batteries from series (24V) to parallel. A single DPDT switch (or relay) can do this (rearrange series to parallel). An even better solution might be a 3P- or 4PDT that switches all the batteries from "use" to "charge", because the single DPDT scheme can potentially connect 24V to the 12V charger. Putting a diode in series to block reverse current will prevent the charger from working properly (because of the 0.6V drop across the diode).