[Part 9] Electrical Design - The Main Board

I draw a big blank circle over this part because this is the most complicated part of building a UPC and I don’t think I can afford to go about explaining every single details I did here. So I’m going to write just a general description of what I’ve done.

First I’m using a Parallax Propeller USB Board. This board has a very small area to work with so I had to cram in as much as I can to get everything in place. I’m designing my Segway in such a way that it can be disassembled easily. So every single data channel cables and power jumper wires had to be detachable. The easiest way to accomplish this is to buy multiple pin wires and use latch-hook connectors to connect them. I’m personally using five 10-channel ribbon wires (two ribbon wires dedicated for the OSMCs, one for the display module, one for the headlight power supply and finally the last one for external sensors and control hub). The next picture shows five latch-hook connectors for these. It is basically showing the other side of the same board.

On the bottom right, there are two green PVC boards. The one sitting vertically up is the LISY 300 Gyroscope. It has to be oriented in that way because the direction of its angular speed measurement is specific in that orientation. The other PVC piece houses Hitachi 3-Axis Accelerometer and since it can measure acceleration in all direction, its orientation does not matter.

I heard from somewhere that ribbon wires can handle up to 1W per each channel. So I had to a little math to make sure that they don’t melt especially when I wanted to use them to transmit power instead of data (I actually ended up melting couple of wires by ignoring this simple procedure).

To be able to control OSMCs, I needed five data channels. Since I was using a 10-pin wire for each, this was really no problem. The picture above shows logic convention for OSMCs. No matter what function does the OSMC carry out, AHI and BHI pins always take 1. So I just tied them permanently to high 5V in my propeller board. ALI and BLI pins were connected to the Propeller pins directly through current limiting resistors. The disable pin is pretty much useless in designing self-balancing machine because if you stop to thing about it, even if you are not driving the machine, it has to be on at all times just to stand up and maintain its position. So I tied the disable pin to low (ground).

By the way, if you check out the OSMC design, there are current limiting resistors built-in in form of a voltage divider. I didn’t follow the schematic. So the use of current limiting resistors is mandated. Another nice thing about the OSMC is that the active data pin voltage can be anything from 2.5V to 12V (if my memory serves right). This is certainly a very nice feature since almost all the microcontrollers fall within this range!

Now that I think, using 10-pin connectors on OSMC is a bit waste of space since it actually only requires 4 pins (or 3 pins) to be able to function properly. 

The display module in my control hub runs at 1.5W. Therefore using only one channel for power transmission in a ribbon wire is not a good idea. I devoted two channels for power supply (so total four channels, two for high 5V and the other two for its ground). This will be just enough to supply all the instruments in my control hub (two potentiometer, one light sensor, one display module). There are several channels left unused which is good for any possible future modifications.

The whole Propeller board is not mounted directly to the main chassis. Instead there are long four nylon screws connecting the board to the mounting plate. This is done in a way that the board kind of hovers a couple of inches away from the mounting plate with a vibration damping material placed in between (black high viscous foam in the picture). I did this later to isolate the board from the vibration of main chassis. Between the plate and the main chassis, there is another damping material inserted for more layers of damping.

The main chassis houses everything, from motors to the navigation pole which then houses the control hub. So it is a single most critical element in my design. Although everything structural element is conductive metal (aluminum), it is not connected to the ground as usual. This will make it safer to drive under high static condition, I think.

No comments: