Stu2 - W7IY

Master

Messages are exchanged with the remote as described above. The master selects an antenna by sending an 6 bit number to the "C" port of the pic, which is connected to 4 Allegro Current Sink Latches. Then, the corresponding latch is 'strobed,' which turns on the appropriate sink - and the corresponding relay. The relay connects the radio to the antenna.

When power is first applied, no relays are selected and there is no 'memory' from the previous power on state.

Logic

Currently, any station can select any antenna. The same antenna can be selected by all stations at once. No attempts were made to split the signal. This would require a much more complicated relay arrangement. If this turns out to be a problem, we can update the firmware to limit antenna selection. (e.g. an antenna can only be used by one radio at a time.)

I added a button to envoke a test program. Just after reset, the program senses the logic of the 'test' button. If it's low (0), the PIC releases all the relays and switches them on - one by one. If this works, you know the PIC can control all the relays and the relay coils work.

Construction

A prototype board from Basic Micro was used to build the master controller. This saved time and was fairly inexpensive. The board costs $9 and the parts cost about $15. The parts actually come as a 'kit,' which includes two sets of parts and costs about $30

At first, I wired the Allegro latches to the prototype board, but then discoverd (after three hours!) the protoype board had some pins linked together. This would have shorted out the latches. So I removed the rats nest and decided to make my own PC boards using one of the services advertised in Nuts and Volts.

The boards were expensive. Plus, I ended up spending another 6 hours learning the software and laying out the boad. They cost about $65 for two - 3"x2" boards. On the plus side, instead of spending three hours doing point to point wiring, I soldered the four chips and connectors on the new board in about 15 minutes. I also had a lot of fun learning the method. Next time, I won't have to go through this learning curve.

The tricky parts were 1) using the right part from the PCB parts library. I actually had to make a part from scratch. 2) Picking the right trace sizes. I used 25mil traces and are they small! 3) Mapping the PCB drill sizes to the manufacturer's standard sizes. The E-PCB web site had tips and a script to automate the process.

A LCD module displays the messages from the remotes, but we elected to keep it disconnected. We didn't have a place to mount it. Again, nothing tricky because mbasic takes care of the details... unlike using C.

All the boards were mounted in a nice plastic cabinet purchased from e-Bay. We used #6 screws and nuts for stand-offs. The cables connect to barrel F connectors mounted on the bottom of the box to a strip of aluminum. This grounds all the connectors together. A wire is inserted in the 'box side' of the barrel, which is connected to the approriate terminal on the relay board.

The relay board was "yet another amazing hamfest find" by my Dad, WA0DYJ. I told him we were planning this project and by chance, he stumbled across this board at Dayton for $15! I reverse engineered it and added the appropriate jumpers to make it fit my needs. This saved a lot of money and time.

Power

A 24VDC power supply provides voltage for the relays. When all the relays are switched on (which I did accidently during prototyping,) they draw about 400mA. A 9VDC wall wart provides power for the master and remotes. They draw about 350mA total. Standard CAT5 computer network cable and DB-25 connectors are used to link the remotes to the master. The maxium cable run is 20 feet.

Each remote and the master have 7805 voltage regulators for power. The 'chips' require 5VDC. To keep the remote regulators cooler, I used 9VDC wall wart instead of the common 12V supplies.