Posted: 5/7/2014 7:21:12 AM EDT
| I've been doing some research on various digital modes and I keep seeing MT63 praised for it's performance in high noise/interference or jamming conditions. It would be interesting to see how JT9, JT65, MFSK16, Olivia, MT63 etc. perform at QRP levels in the presence of OTH radar / jamming (Firedrake, etc.). I was curious if anyone has actually put this to the test and tried to make a QSO during such conditions? |
|
Quoted:
I've been doing some research on various digital modes and I keep seeing MT63 praised for it's performance in high noise/interference or jamming conditions. It would be interesting to see how JT9, JT65, MFSK16, Olivia, MT63 etc. perform at QRP levels in the presence of OTH radar / jamming (Firedrake, etc.). I was curious if anyone has actually put this to the test and tried to make a QSO during such conditions? Could always try pre-planned QRP digital contacts during field day... That would be a good test. Aside from that or having some one with an amp intentionally jam you I am not sure how you would simulate? Kind of hard to make 5w work on a freq if some one in the area were putting out 500w+ not to many ways to swing that, waterfall will be flowing 
Of course the practice and thought process of Ham radio is to move on the spectrum and find a clear frequency that is playing well with propagation, avoid the jamming and interference don't fight it. Unless you have an amp  (I don't)
|
|
I've seen a number of sensitivity tests for copying (I think the yahoo Olivia group had
a pretty big list) but I don't think I've seen anything specific to robustness. Ultimately it depends on sensitivity and the depth of the forward error correction, the greater the FEC, usually the stronger the mode. The downside is the more error correction, the slower the mode. We use a pretty slow Olivia mode for the digital net and pretty much nothing short of a major solar event stops the net. I think we've had to call the net due to prop issues maybe 1-2 times a year, so that makes it 96-98% reliable. For the majority of us QRM never seems to be an issue. I would imagine any mode with equal or better FEC and slower data rates would be even more robust, so that would give the JTx modes a free pass (they've got more FEC and slow data rate than Olivia.) Many of the other modes are at a major SNR disadvantage compared to JTx and Olivia -- you get into higher data rate MT63 for example, and the SNR sensitivity really starts to drop off. This paper summaries the issues fairly well. A few times on net night we've tried other modes and I think most would agree we can find stuff that's faster than Olivia, but every choice sacrifices robustness/error correction. Note we didn't try JTx modes because it'd take a lifetime, but Contestia, MfSK, B/QPSK modes etc got tried out. |
|
VOA Radiogram is a Voice of America program experimenting with digital text and images via shortwave broadcasting.
It is produced and presented by Dr. Kim Andrew Elliott. http://voaradiogram.net/ Last weekend’s “battle of the 100-word-per-minute modes” produced useful results. Based on a quick analysis of the percentage of correct characters, the average of all responses is… MFSK32 98% MT63-1000L 90% Thor25x4 86% PSKR125 74% The MT63-1000L would have had an even better result if it had not been for four occasions of no decode at all. In situations where any MT63-1000L text was received, its performance was very close to that of MFSK32. MFSK32 does have the advantage of being about 20 words per minute faster than MT63-1000L. Furthermore, MFSK32, unlike MT63-1000L, can be used for images. This weekend, VOA Radiogram will return to an all-MFSK32 “set it and forget it” format. I do this at the risk of losing the interest of those of you who prefer a program replete with experiments. The program will be interesting in that it will include four images and an Flmsg CSV form with the updated VOA English transmission schedule. http://voaradiogram.net/post/81671836096/voa-radiogram-5-6-april-2014-includes-the-voa-english VOA Radiogram, 29-30 March: Battle of the 100-wpm modes Videos of VOA Radiogram, 29-30 March 2014, “battle of the 100-wpm modes.” VOA Radiogram during the weekend of March 29 and 30 included a “battle of the 100-word-per-minute modes.” The modes were PSKR125, MT63-1000L, Thor25x4, and MFSK32. Videos kindly provided by three VOA Radiogram listeners demonstrate the comparative performance of the modes. TW in Japan provided this video of the beginning of the broadcast Saturday at 0930 UTC on 5745 kHz. Because of the challenging transmission path from North Carolina to Japan, even the MFSK32 text was not 100%. The PSKR125 that followed was mostly unusable: youtu.be/jl1SKcfZXAY Tim in Colorado provides this YouTube video of the entire broadcast Saturday at 1600-1630 UTC, along with some of his commentary typed in the Fldigi transmit pane. This is a good example of the decoding of the four modes under fair — meaning not great but not too bad — shortwave reception conditions. At 22 minutes into the video, Time declares “DA WINNER” of the battle of the 100-wpm modes. youtu.be/-VqIXc9SEXw Nicolas in Colombia produced this video of his reception Sunday at 0230 UTC (Saturday evening local tme) on 5745 kHz. It demonstrates the decoding of VOA Radiogram modes under even more challenging conditions, and using a relatively inexpensive shortwave portable with an analog dial. It appears that Nicolas edited out his mostly unsuccessful decode of MT63-1000L. Thor25x4 was more successful with only a few errors. Nicolas’s decode of MFSK32 had some errors, as shown in the video, but my decode of MFSK32 using audio from his video was error-free: youtu.be/H2-o4y_fkhQ Here is a screenshot from Nicolas’s video… 
1 month ago The QSL for VOA Radiogram, weekend of 22-23 March 2014. Even in cases when the image was fuzzy, usually due to the distance from the transmitter in North Carolina, the text portion of the program was received 100% because of the forward error correction of the MFSK32 mode. 
The QSL for VOA Radiogram, weekend of 22-23 March 2014. Even in cases when the image was fuzzy, usually due to the distance from the transmitter in North Carolina, the text portion of the program was received 100% because of the forward error correction of the MFSK32 mode. VOA Radiogram transmission schedule (all days and times UTC): Sat 0930-1000 5745 kHz Sat 1600-1630 17860 kHz Sun 0230-0300 5745 kHz Sun 1930-2000 15670 kHz All via the Edward R. Murrow transmitting station in North Carolina. To decode the digital text and images transmitted on VOA Radiogram, download Fldigi, Flmsg and Flamp from w1hkj.com. See also how to decode the modes . |
|
Quoted:
Of course the practice and thought process of Ham radio is to move on the spectrum and find a clear frequency that is playing well with propagation, avoid the jamming and interference don't fight it. Unless you have an amp (I don't)My interest comes from a currently theoretical (historically realistic) situation where Russia, China, etc. decide to switch on their radar or jamming transmitters (Russian Woodpecker, Chinese Firedrake, etc.) that cause major interference on the amateur portion of the bands, and what modes would give you the greatest reliability under such interference. Thanks everyone for the replies! Good stuff! I really appreciate the radio community here on AR15.com - questions like this on other forums (qrz.com etc.) usually just end up dying or turning into degenerate arguments between over-inflated egos. |
|
I have a large number of both Olivia 16/500 and MT63-1K QSOs in my log. Of the two modes Olivia is by far the best performer in weak signal conditions. However if you can actually see the signal on the waterfall MT63 is a great deal more robust. I've had successful MT63-1K QSOs with Olivia stations as QRM right in the middle of the MT63 signal.
There are some trade-off's with MT63. The 1K version requires quite a bit of bandwidth. It requires precise tuning of the signal, usually +/-50Hz for most of the software demodulators available (by software demodulator I mean, for instance, the code in Fldigi that does the actual demodulation and decoding of the signal). It also has a rather long latency, with approx. 12 seconds required before print will start, due to all of the FEC used. This latter issue often frustrates MT63 newbies because with decode not starting right away they will fiddle with the tuning during this time and each time they fiddle with it the decode process has to start over. This results in no decode and they can't figure out why. So with MT63 just trust you've got it right for 12 seconds before fiddling with it! I would be happy to have some MT63 QSOs with people when I get back home (sometime this weekend). It is actually one of my favorite modes. JT9 and JT65 I suspect are even more robust, but of course they are only appropriate for "emergency action message" type traffic. I had suggested once that we designate a few freq's that were ARFCOM JT9 guard channels, perhaps even make up some ARFCOM EAMs. These would be transmitted at a relatively high power for the mode in order to assure reception. While the idea got a lot of arfcom hams interested in JT9 and JT65 it never really went anywhere. Maybe we can revive it? |
|
Quoted:
JT9 and JT65 I suspect are even more robust, but of course they are only appropriate for "emergency action message" type traffic. I had suggested once that we designate a few freq's that were ARFCOM JT9 guard channels, perhaps even make up some ARFCOM EAMs. These would be transmitted at a relatively high power for the mode in order to assure reception. While the idea got a lot of arfcom hams interested in JT9 and JT65 it never really went anywhere. Maybe we can revive it? If I could get JT9 or JT65 running on my radio, I'd be using it. I can't seem to get the app to transmit. |
|
using a debain sid linux machine, two serial interfaces (one for control, one for PTT), an Icom IC718, and an auto-tuner.
The problem is getting WSJT-X and WSJT to run correctly. For some reason the soundcard works correctly in EVERY OTHER APPLICATION except WSJT and WSJT-X. Once I get home I'll post a thread about it. |
|
Quoted:
using a debain sid linux machine, two serial interfaces (one for control, one for PTT), an Icom IC718, and an auto-tuner. The problem is getting WSJT-X and WSJT to run correctly. For some reason the soundcard works correctly in EVERY OTHER APPLICATION except WSJT and WSJT-X. Once I get home I'll post a thread about it. Try using pulse audio and pavucontrol. In WSJT-X set the soundcard to pulse, then in pavucontrol find the application and select what soundcard you want to use. |
|
Got the KX3 / PC audio cables last week, and was playing with WSPR this weekend. I think I may be running into some of the frequency stability problems that others have reported with the KX3, which would be disappointing to say the least.
On RX I get nice clean horizontal WSPR signals on 30m, but when I switch to 20m every signal trace has a funky wave shape to it. Other times I get a clean horizontal line with a frequency shift step halfway through the cycle. This shows up on all the signals present so I know it's me, and not the TX station. I live in an RF noisy townhome with the entire building's AC feed on our outside wall so maybe I'm getting some funky RF interference that the radio tries to correct with AGC or something? Doubtful, but maybe? I'd like to go to a RF quiet area and try it again, and see if it makes any difference. I did a little transmitting on WSPR with 500mw on 30m with good results, so I don't think my problems are heat shift related. When I check my transmitted signal on the WSPR database I have 0hz/min drift on 40m, 0-1hz/min on 30m, and 3-4hz/min drift on 20m. I will do some more testing this week and see if I get even more drift on higher frequencies (15m+). I will also try a side by side with my FT-817 to see if I get similar results, so I can be 100% sure it's the radio, and not my laptop causing problems. Stay tuned... |
|
Propagation wasn't great on the higher bands, so I ran 20m WSPR again this evening. After watching the waterfall it does appear I'm getting frequency instability 2-3 cycles after my transmissions, and then it stabilizes for the most part. I'm surprised because I didn't figure I'd be creating enough heat at 500mw to do anything. The exterior of the case is room temperature....Bummer...
http://i239.photobucket.com/albums/ff54/aurich85/12MAY14WSPR20m.jpg http://i239.photobucket.com/albums/ff54/aurich85/12MAY14WSPR20m500mw-2.jpg http://i239.photobucket.com/albums/ff54/aurich85/12MAY14WSPR20m500mw-3.jpg http://i239.photobucket.com/albums/ff54/aurich85/12MAY14WSPR20m500mw-4.jpg |