Sunday, March 13, 2011

Internet Repeater Linking and the PC-based Repeater Controller

In the past, amateur radio repeaters have been controlled by dedicated controllers, whether commercial or one-offed. Linking has been carried out by radio or wireline/telephone. This works well and is proven technology, but today there are alternatives in the PC-based controller and VOIP linking.

I recently became involved with a project to use PC-based controllers with voice over IP linking to connect three to four repeaters in three separate cities that are roughly in a line and each separated from the next by about 60-90 miles. There wouldn't normally be any issue in linking these repeaters by RF, if they were at hilltop tower sites, but these repeaters are what is commonly referred to as "backyard repeaters." Since these repeaters don't have a clean line of sight to each other, the conventional alternative would be to use a phone line, but as you know, phone lines are expensive and inflexible. So in this case, since the repeaters are at homes and businesses, we decided to leverage the ubiquitous broadband Internet connection for linking.

The software that we are using both as a repeater controller and VOIP end point is thelinkbox. There is also a Yahoo Group (http://groups.yahoo.com/group/thelinkbox/) dedicated to thelinkbox, and this is the best place to find both the latest version and information on how to setup and configure thelinkbox. Once I complete my node of the system, I will be documenting how to build and interface circuit and how to configure thelinkbox. Until then, here are some pictures of the PC that we are using, the interface circuit and my repeater node.

From top to bottom, there is a Micor 75W power amplifier, two GE MP-A handhelds in vehicular chargers, a Micor power supply, a Decibel Products band pass / band reject duplexer.
This is my homebrewed repeater interface board, there are transistor inverters, audio isolation transformers, a pull up resistor and some current limiters. There is also a DC blocking and bypass capcitor to keep the preamp voltage from the transmitter out of the PC soundcard. This is very important because applying voltage to the PC soundcard inputs and outputs could damage the ports.
Backside view of the interface board. Not shown in this picture is that both sides of the PC audio out must be connected, because repeat audio is on one channel and tones / ID are on the other channel.
This is the carrier operated relay that I added to switch the receiver audio on and off with the COS signal that is also being sent to the PC. I had to do this because thelinkbox, unlike dedicated controllers, doesn't seem to handle muting this input audio. If we were using speaker audio or audio from before the speaker amplifier this wouldn't be an issue, but we are using discriminator audio so that the audio profile is flat. This keeps us from having to mess with the de-emphasis algorithms. It is important to note the diode in reverse across the coil terminals on the relay. This is so that when the COS goes low and the magnetic field in the coil collapses, the generated voltage spike is dissapated and cannot damage sensitive transistors and integrated circuits in the receiver or PC.
There is quite a mess with all of the connections to the prototype interface board, but that will be resolved soon because we are going to create a circuit board layout and have some of these made up with standard connectors so that we can deploy new nodes easily.
The PC that we are using is the EES-3610, also known in the tlb community as the "Little Blue", a fanless VIA based single board computer with two serial ports, two network ports, two USB ports, a parallel port and a built in sound card. It is a good PC for this application but it does require some tweaking to get everything working.
Testing requires the usual repeater test equipment, plus a way to watch logic signals and some headphones are useful for tracing the audio path. Also a PC with the ability to access the web interface and SSH are necessary as well.
Sealed up and working, this repeater is in the process of being programmed to id by CW, voice and Text To Speech. Timed IDs are programmed using the tlbcmd utility and the Unix style cron program on the little blue. Once I receive coordination this repeater should be on the air.