Andy, K3UK and I had a very interesting experience this evening testing several sound card modes on the lower end of the 40 meter voice/image portion of the band. I have been wanting to do some comparison of the single tone high speed MIL-STD/FED-STD/STANAG modem in the RFSM2400 program with a multi tone OFDM modem.
Signals were not that good between our locations, with deep QSB at times. Which was good for testing purposes. Andy running 100 watts and I running barefoot at times and up to about 250 watts when running digital (more on SSB). Once a minor glitch was taken care of, we discovered that the program is extremely easy to use. Just enter the station you wish to connect with and the negotiation is done by the software. There are some additional windows that can be opened that provide feedback on the actions being done. One window can be opened that includes an oscilloscope like screen that displays what appears to be a real time constellation of the received signal. Very impressive. Since we can only use this technology for sending image/fax here in the U.S., we elected to send some moderate sized jpeg images. I sent Andy a file that was 28521 bytes and it took 731 seconds to make the transfer which figures out to 312 bps. The software gives you this information upon completion of the file transfer. Andy then sent me a file of 7323 bytes which took 197 seconds or 295 bps. The software will display the received S/N ratio on screen and between each packet will update the transmitting station with how well the signal is (or isn't) being received with a S/N number. Even when I was only running 25 watts and turned off the amplifier, there was some throughput but it was reduced. The general report that I saw was 5 db S/N without and 10 dB with the amplifier. I would have expected something closer to 10 dB difference. When receiving the S/N appears to be close to real time on the receiving stations display and with QSB was all over the place with peaks at around 15 dB, but dropping down to 4 dB and I believe that I did see some negative numbers but would like to test this further. The point of all this is that it is possible to send data, even if fairly slowly. The ~ 300 bps speed roughly equates to 37 bytes per second or 370 wpm if you were sending uncompressed text. I don't know if any compression is added for the program, but perhaps not. That means that with text data we might be able to nearly double the throughput. Next we tried to use EasyPal, currently a popular digital OFDM program used for SSTV. I sent a file that would normally be a bit too large but it compresses it down to a much smaller file although you can adjust the quality on that as well. The file size at the default value became 20K and was to be sent in 94 seconds. There were something over 180 blocks sent. Andy received perhaps one sixth of them, but we suspected that it would take many tries to get the data through fully intact using the manually requested BSR system. This confirms that these OFDM programs tend to require a signal better than +5 dB and perhaps +7 to +10 is more reasonable for throughput to work well. It does seem that the RFSM2400 single tone modem worked better than OFDM, but I wish we had a comparable ARQ OFDM mode that did the corrections on the fly. It would be very helpful to have more hams do further comparisons, under varying conditions, and share their experiences. One thing that we need to test is how well that ~300 wpm throughput compares with FAE 400 and FAE 2000 under the same conditions. We would have had to move back down to the text data portions of the band to perform such a test. I am available for further testing with these modes and other modes too, such as NBEMS and ALE/FAE 400, or other programs you would like to compare. Just think ... a year ago almost none of this existed for ham use. Many thanks to the programmers who are helping us with the new sound card technologies. 73, Rick, KV9U
