Opportunities and limitations
Next
Previous

The advantage of the slower JT9 modes is that they still decode at levels where the "traditional" JT9(-1) mode fails. The theoretical improvement compared to JT9-1 is:

Computer simulations (dry tests) confirm these values pretty well:

With on-air tests it is much more difficult to determine the exact values, but tests on 136 kHz and 475 kHz confirm a significant improvement (more decodes) of JT9-2 over JT9(-1) and of JT9-5 over JT9-2. For JT9-10 the results are a bit more ambiguous.

 

But there ain't no such thing as a free lunch, the slower JT9 modes also have their limitations.

It is obvious that a QSO will take longer for slower modes (but at least there will be a QSO). If no transmissions get lost a basic QSO takes 5 minutes for JT9(-1), 10 minutes for JT9-2, 25 minutes for JT9-5 and 50 minutes for JT9-10.

But as the SNR drops, a QSO in a slower JT9 mode will proceed faster due to the fact that less transmisions get lost:

The graph above shows that for an SNR better than -26.7 dB JT9(-1) is the fastest mode. Between -26.7 db and 30.4 dB JT9-2 takes the lead. From 30.4 dB to 34.4 dB JT9-5 will be the fastest and for an SNR worse than 34.4 dB it is JT9-10.

 

The above graphs assume perfect frequency stability. However, the slower the JT9 mode the narrower the bandwidth is and the more vulnerable it becomes to frequency instabilities.
The acceptable frequency drift seems to be about 1 Hz/min for JT9(-1), 0.5 Hz/min for JT9.2, 0.2 Hz/min for JT9-5 and 0.1 Hz/min for JT9-10. For strong signals the acceptable frequency instability is more relaxed, but for these sigals you don't need the slower JT9 modes.
Frequency instability can be caused by oscillator drift at the TX and RX, and also by ionospheric fluctuations. On 136 kHz and 475 kHz most state-of-the-art rigs have a sufficient frequency stability for modes up to JT9-10, but on higher bands this might become a problem. On 475 kHz, and to a lesser extent on 136 kHz, ionospheric fluctuations can cause a Doppler shift that exceeds the permissible drift for JT9-10 and sometimes even JT9-5 mode.

On-air test have shown that over a 5000 km transatlantic path JT9-5 and JT9-10 signals could be copied over longer periods on 136 kHz, so most of the time ionospheric fluctuations will not be an issue on this band. Over the same path JT9-5 signals were copied regulary on 475 kHz. There were also occasional transatlantic copies of JT9-10 signals on 475 kHz.