Digital Modes: Applications
PSK-31 and EMCOMM
During the last SET it became apparent that a method of handling written (text) type traffic was needed. For routine inquiries, the "bike-a-thon" method works well. But in several instances that I heard, the traffic was pretty involved.
For example, the hospital is asking for a list of available supplies. The operator at the hospital contacts the EOC and begins to read the list. A couple of problems surfaced immediately: How do you pronounce some of the medical terms if you don't have a medical background? The second problem was at the receiving end: How do you spell those weird words?
Overall I thought the traffic was handled pretty well, considering. Yes considering...don't you hate that word? Most realized that things would have been much simpler if the originating station had the ability to simply pop the message into a computer and print it out at the far end. That way, any mistakes in the message would lie solely with the originator.
I understand that some attempt was made to use packet via the APRS system and it didn't work very well. There was also some resistance from operators outside the area that felt it was inappropriate to use the APRS in that manner. SAY WHAT?
A few of us have been experimenting with another digital mode to handle traffic that is readily available, easy to use and a piece of cake to set up. PSK31. This mode was developed for use on the HF bands and it works well there usually on USB although under noisy conditions the copy is usually less than perfect. Not the best choice for message traffic. But under the quiet conditions of FM the copy can be 100% even without an error correction system (fec). The throughput comes to roughly 60wpm and if error correction becomes necessary there are versions of PSK that DO use fec. These modes are only slightly slower. The ultimate in rough copy is the Olivia mode, available in MixW that survives under brutal noise and interference and maintains a throughput of 15 wpm. Not bad.
Here's a proposed architecture that might work really well if adopted:
First of all, this can be point-to-point or point-to-multipoint as required. No changes are required to select either condition. We confine the discussion to point-to-point.
Each end of the traffic link will be equipped with two computers, laptops preferred. One computer will be connected to an FM transceiver via a couple of patch cords and a VOXswitch. This computer will be networked to the second computer so that the traffic can flow back and forth without personnel tying up the radio computer. Traffic is originated by inputting it to the second computer. The operator then grabs it through the network and sends it as a file to the other end. At the far end the message is received and the operator cuts/pastes the copy to wordpad and routes it to the second computer or to a printer. If the site only has a single computer, the operator will need to do the inputting of the traffic as well as the transmitting, but even then it would probably be faster than doing it via voice and certainly less prone to errors. That's all there is to it.
The RF path is via an appropriate repeater or better yet, simplex if conditions permit. If the simplex path is too great for solid copy a crossband repeater could be set up. Of course to comply with FCC rules (actually to prevent FCC rules) proper ID needs to be used either via CW or voice and of course this needs to be done on 2 meters or above.
The interface
There are a number of commercial interfaces that could be used to connect an FM transceiver to the laptop but in the testing we did, nothing very elaborate was required. The path from the receiver to the laptop was nothing more than a short shielded cable with 1/8" phone plugs on either end. I used mono plugs because my W32 programs through the earphone jack (can you believe that) and a stereo plug doesn't work. By adjusting the levels on the laptop's sound card mic control, and then adjusting the output of the handheld, the computer copied perfectly.
You could do the same for the outgoing audio as well but then you would need to tie up one hand pressing the PTT and type with the other. Bad idea. Here's a nifty little gadget that I built to make a "fox" out of a mp3 player and a handheld radio. I call it a VOXswitch.
These parts are all available in either a well stocked junk box or at Radio Shack. (You have questions, we have batteries) Layout is not critical because it is pure audio. Shielding is used on the input and output cables to keep RF from getting into the gadget. I made mine on a piece of perf board using point-to-point wiring. There are only two adjustments. See the section below for the procedure.
Theory is easy to follow: The audio flows from the source, an mp3 player or in this application, the laptop's soundcard audio out. It passes through the isolation transformer to ground giving the audio amp in the source device a nice load. The audio is coupled through the transformer, through the 3.3K resistor to the mic jack of the radio. This design is primarily for JA handhelds such as Icom, Kenwood, and Yaesu. Most of these use the same type of circuit to trip the PTT. When a fairly high resistance is put across the mic line, (in this case 33K) the radio goes into transmit. The resistance is high enough that it does not swamp the audio from the mic. The same audio is also fed back via pin 2 of U1a to start a bias signal to the switch, Q1. When Q1,s base goes low, it conducts completing the circuit for the resistor across the mic line. For other types of radios this switch could be modified, letting Q1 pull in a small relay which would give you a hard, dry contact to key just about anything.
Adjustment
On October 8th we got a chance to test our theories as well as the equipment. We had a couple of stations set up in St Helena, configured much as described above and one station set up in Napa. We were able to communicate fairly well after a number of false starts. One thing we found in testing is while the PSK signals seem to pass through the repeaters fairly well, we never did get our levels set properly. While they sounded ok, and the traffic printed out ok, they were driving the repeater into clipping indicating that some level reduction was in order. The VOXswitch described above has been modified to provide the fine tuning required by the addition of a 10K pot on the mic line. In discussions with WA6BIS we determined that perhaps the best way to set the transmit level is to back the gain off to zero then slowly increase it until the receiving station gets a solid trace to print. That should produce something just below 2 kHz deviation which ought to be just about right. Experimentation will tell. The only other adjustment on the VOXswitch is the 10K pot wired to pin-6 of U1b. Set this pot so that your audio trips the PTT on the hand held radio.
Receiver gain: This looks pretty simple. I used a straight through cable with a 1/8" plug on each end. Setting the audio on the handheld too low produced an unprintable trace. Setting it too high produced multiple traces (harmonics) some of which printed and some of which did not. Backing it off to a point between the two produced a perfect trace and copy was 100%.
Another Interface. The interface described above was designed for use with a JA hand held radio. Transceivers with conventional PTT systems usually are equipped with some sort of mic jack on the front panel. The following interface, sent in by Mark, KE6O will key most of them and is very simple to build. Notice that it keys off the serial port of your laptop. If yours is not so equipped, you will need a serial/USB converter to make it work.
As you can see there is no adjustment but you will need to assign a port through the program you happen to be using for PSK31. If you don't have the manual for your radio, you can find the MIC PTT pin easily by grounding the pins on the mic jack, one at a time, until you hear the PTT relay click or see the TX indicator light up. The pin you want will be #5 on most Icoms. You should find about 5 volts on this pin as well.
Also notice that this schematic is lacking a means of controlling the audio levels. The receive line (from the radio to the laptop) will probably not require anything, but the transmit line (from the laptop's phone jack to the radio's mic jack) will most likely require a pad of some sort as described above. If you build this little beauty in a small project box you can include a couple of jacks and a pot to provide that padding.
Summary
This idea uses equipment that is already available and the software has a very shallow learning curve. I believe most of the freeware programs such as winpsk have file handling capabilities and for the ones that don't a cut/paste into a macro would work just fine.
The skeptics are sure to say "That ain't gonna work; in an emergency the repeaters will be down." To them I say, "Maybe, but what if they aren't?" As always, HF is still a backup option; using the same architecture as described above, same software, same learning curve. Skeptics who focus on things falling down, blowing up, blowing away don't have the big picture. The worst disaster you could imagine may not damage the infrastructure at all, but as we have seen in the past when conventional means of communication get overloaded they become unusable. Preparedness is having as many "tools in the toolbox" as possible; maintaining them, and knowing how to use them.
73, N6XN