help with navtex/medium wave receiver sensitivity and decoding

Actually accurately placed high loaded Q (over 100, can use up to 1000) in the RF section will actually help this mode (50 baud fsk 170 Hz shift).

Perhaps an old analog PLL would clean up the signal for further processing in teh digital world.

Thats an interesting idea...I would need a slow moving error loop to avoid pulling it all over the place. Unfortunately at an IF of 6kHz and a 100baud Ive only for 60 cycles in which to lock and recognise a particular frequency, but perhaps more like 20 cycles to allow for reasonable edge detection between a logic 1 and zero. If i were using a to part conversion it would be a neat solution and I'd have a higher

I'm trying to resist putting a very tight filter on the front end as I would like to be able to choose between 518kHz and 490kHz, although practically this may be a reason why a certain commercial low end receiver I tried (which is also dual channel) was so hit and miss.

For US use, only 518 kHz is used. If you use really high q filters (200 or more) dual filters is most appropriate. Cheap units cannot afford to do this. The distance between the two "carrier" frequencies is way too far for single high Q rf tuned circuits. Multiple high Q tuned circuits works a lot like combined AM and FM IF strips.

I'm in Great Britain...and theres some useful local weather forcast information on 490kHz. I have to do some research on filters. Many thanks for your time with this Joseph