Hi Alfred, Thank you for your fine report. Some comments:
Rx audio filtering isn't important for FreeDV as the demodulator has it's own filters - one for each modem carrier in fact. 1/ Based on reports so far (and my tests) - I don't think 700 has any advantage over 700B. We'll probably remove 700 from later releases, it's included for compatibility for now. 2/ The initial frequency offset estimator for the cohpsk modem used in 700(B) has a small initial lock range of +/-60Hz (FreeDV 1600 is +/- 200Hz). Doesn't seem to be a problem in practice, once you are used to it. Good to see it's locking up fast - I put a lot of work into it! 3/ 700(B) has been designed for a low crest factor (PAPR). You might be able to drive the PA harder (if your PA doesn't get hot - pls be careful). 700B works down to around 0dB SNR (measured in 3000Hz), so perhaps there is an issue in your SNR estimation code for strange waveforms like FreeDV. Can you pls tell me how well 700B worked in low SNR channels compared to SSB? That is an important goal of FreeDV for 2015. Thanks, David On 31/08/15 21:57, wully wrote: > Hi David and all interested > > Today, the conds on 40m where usable after a long period of complete > loss of signal because the MUF was only around 6.5 MHz for 300 km. > (dc6ny is about 275 km away from my location). > > But now dc6ny and i have made our first freeDV700B-qso today at about > 1100 UTC. > My audio frequency filter for tx and rx was from 700 Hz to 2400 Hz for > this mode. > > I noticed some very positive things: > > 1) the sound of the 700B mode is - at least what I remember from our > experiments a year ago - > comparable in sound quality to the 1600 mode. I was very pleased > about that. We did not test the 700 mode, but in my local experiments, I > found the 700 mode not as comfortable as the 700B. Since the modulation > scheme is the same, I assume, that 700 as no advantage over 700B? > > 2) As soon, as the rx-frequency is adjusted correctely, the modem locks > very quickly. In the 1600 mode it seemed to me not as quick. > > 3) I used about 25 to 28 Watts peak measured just before the antenna > coupler. dc6ny gave me a report of 59+ on his hpsdr system. Switching to > SSB, my peak power was between 250 and 300 W. > > Question: I show the estimated SNR from the libarary on my sdr. In > 1600mode I always saw about 12-20 (positive) with good signals (whatever > these figures mean). In the 700B I see a figure around -6.2 (negative) > with good signal. Probably I have made an error in the presentation? > > We are planning experiments with different SNR conditions to find out > the limits of this new mode. > > The first results are very encuraging! > Thank you very much for your constant effort on freeDV! > > 73, Alfred, hb9epu > > > > > > > ------------------------------------------------------------------------------ > _______________________________________________ > Freetel-codec2 mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/freetel-codec2 > ------------------------------------------------------------------------------ _______________________________________________ Freetel-codec2 mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/freetel-codec2
