Tony, Static crash resistance is not the only parameter to consider. The problem is that you can have static and weak signals at the same time. MT63-1000 has a -5 dB minimum S/N, but MFSK16 has a -13.5 dB minimum S/N, so the static tests you made must be at signal levels high enough that MT63-1000 decodes, which may not be a realistic level.
Last summer, during the lightning season in Florida, MFSK16 turned out (after three months of testing) to be the most static-resistant mode of all, even surpassing Thor, which we had worked on so hard to harden against static crashes. However, THOR is tolerant of mistuning, whereas MFSK16 is not, and MFSK16 needs AFC, which Thor does not, but overall, we concluded that MFSK16 was the best for NBEMS messaging on HF unless conditions (QSB and QRN) were such that a faster mode would work. Of course our tests were to find the best mode for messaging, which has to be a combination of reasonable speed and minimum S/N, and MT63-2000 is the only MT63 variant that is fast enough to overcome the extreme latency of MT63 and allow successful ARQ transfers without unreasonable wait times. MT63-1000 is not fast enough. The problem is that MT63-2000 is 3 dB worse on minimum S/N than MT63-1000, so the spread in minimum S/N between MT63-2000 and MFSK16 grows to about 11 dB, which is a LOT! As you point out, the list of variables is very long, and a mode for one situation may not work for another. As you observed during the MT63-1000 tests we made together, MFSK16 would print 80% when MT63-1000 would not print at all, and Olivia was printing 100% under roughly the same conditions. There is a resonably acceptable speed for message transfers, with and without ARQ (ARQ cuts the speed in about half), and a different reasonably acceptable speed for QSO's, just as JT65A is acceptable for short exchanges, but not so much for QSO's. So, for NBEMS, since the primary objective is messaging, on HF we found MFSK16 to be most suitable overall, but on VHF, where there is no static, for instance on 30m there is little static (where PSKmail operates), PSK250 can be used instead, when it is impossible to control the static crashes, or even noise, on the lower HF bands from capturing the AFC and shifting the tuning off frequency on HF, simply because you need to have AFC for PSK250, and between ARQ exchanges, there is no signal to lock on, so the AFC locks on a noise burst. Olivia would be great to use, but takes forever to get a message through, so the better minimum S/N of Olivia has to be sacrificed for greater speed in messaging and use MFSK16 instead, and let the ARQ just resend blocks when necessary. Of course, at some point, enough blocks may be damaged that the link simply fails or times out. Once you add ARQ to MFSK16, you have a speed of only about 20 wpm, which is very slow for anything than a very short message, but the ARQ guarantees error-free reception in return for the slow speed. Minimum S/N, QSB, QRN, doppler distortion, inter-symbol interference, tolerance to operator tuning, transceiver frequency stability, minimum necessary bandwidth, etc. etc., all figure into the decision as to which mode is "best". "No one shoe fits all", and we can only choose the "best" mode for our particular mission out of all the many available choices. 73, Skip KH6TY NBEMS Development Team ----- Original Message ----- From: Tony To: digitalradio@yahoogroups.com Sent: Saturday, March 21, 2009 5:05 PM Subject: Re: [digitalradio] THOR is static-proof (Re: KV9U - MT63) Jaak, > What about THOR? Thor stated to be more static-proof. It depends which THOR mode is used. It seems THOR-22 is the best of the bunch for static crash resistance. I've done a few static crash tests by generating noise at regular intervals; the noise obliterates the signal in short bursts. I would imagine this method would give some indication of on-air performance. I'm sure there are simulators out there that can produce more accurate results. The list of variables that would add to the mix are endless; ionospheric distortion, weak / strong signal performance, QRM etc. As the disclaimers say, your mileage may vary! See below... Tony -K2MO _______________________________________________ Text Message: Quick Brown Fox Pangram Static Crash: Duration: 1 second Interval: Every 5 seconds THOR-11 µ9i$:neíICK olrsplnOX JUAnopco vsR THE l¶unknOG TËq ©E QUICK BRetqksˆX JUMPS«aa±n THE )txeTaTic DOG X erEÒtCK BROsbßnn”X JU 5¶R THE ¡t,a0ssY DOG TŒi R ta BROWN THOR-22 THE QUICK BRwnoacebnOX JUMPS OVER THE Lti ) tla ey tktzlQ HE QUICK BROWtzoh JUMPS OVER THE Lpc·¢fG THE QUICK BROWN L xth Ítl t1 JUMPS OVER THE LAZYk rNyp+THE QUICK $ MT63 1K Long Interleave THE QUICK BREWQUICK BROWN FOX JUMPS OVER THE LAZY DOG THERQUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG MFSK16 THE QUICKl||½ OWN FOX JUMPS hqPeavHE LAZY DOG THEvaŽÊICK BROWNza«cpFOX JUMPS OVER Taetf ‡E LAZY DOG THE Qh tCK BROWN FOX JU3 ]S OVER THE LA¬cc tsa ÕOG _______________________________________________ Text - Quick Brown Fox Pangram Static Crash: Duration: 2 seconds Interval: Every 5 seconds THOR-11 Tseor'Ka °ANROWN F7ueNpg r epitUX s 3àn MDBxhvuntF^yš THE õ ¾bSyK BROWN tq?yõP×7 eZ ²opHE L 8p!t es OGCK Ä A/pttªOX JUMPS OfdròSe THE LAZY Do trtn THOR-22 THE QUICK BuA qklt ¬ JUMPS OVER ta97tncx2td/R>ZY DOG THE QUIceË Ái daÖWN FOae t pQ R m ©t OVER THE elNtîi oMcsiG THE QUICK rLbu otiSoWN FOX MT63 1K Long Interleave THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG THE QUICK BROWN FOX JUMPS OVER THE AOY JOMPS OVEU THE LAZY DOG THE QUICKEBRAWN FOX THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG MFSK16 CK BROWN FOX JUMPS THE LAZY DOGqnæwbih THE QUbs up,‡CK BROWN FOE&l„UMPS OVER THtY DOG G¨¨aId-E QUICK BROW)o tÌieEX JUMPS OVER gt ----- Original Message ----- From: "Jaak Hohensee" <jaak.hohen...@eesti.ee> To: <digitalradio@yahoogroups.com> Sent: Saturday, March 21, 2009 3:55 AM Subject: Re: [digitalradio] THOR is static-proof (Re: KV9U - MT63) > > > Tony wrote: >> >> The most impressive thing about MT63 is how it seems to resist heavy >> static crashes. I made a few recordings with short segments of >> the signal removed to simulate this type of QRN and there was little >> effect on copy. >> > What about THOR? Thor stated to be more static-proof. > > Jaak > es1hj/qrp >> >> > > -- > Kirjutas ja tervitab > Jaak Hohensee > >
