Re: LF: EbNaut LF test from iz7slz

[Markus Vester <markusvester@aol.com>]

> At LF/MF if messages are short, say less than 30 minutes, the tx frequency only needs to drift by less than about 0.5 Hz/hour. That's within range of a reasonably stable undisciplined oscillator such as an OCXO.

 

Yes a 0.5 mHz (not Hz) hourly drift won't affect decoding much. I think the reason for chirp tolerance is simply that a quadratic phase associated with a linear frequency drift has smaller excursions at the end, compared to a linear phase slope from a constant frequency offset. The difficulty is only that the actual center frequency is not easily predictable, and the search can be cumbersome when the signal is weak.

 

Domenico initially used an undiscipled OCXO for EbNaut on LF with good success. During our recent MF QSO, I was using my FiFi-SDR which is equipped with a free-running Si570 oscillator and a glued-on NTC-and-transistor heater (which was fabricated quick'n'dirty style during the night before our 2012 trip to India). This RX drifted up to 50 mHz over the hours, especially after transmit rounds which also affected receiver temperature. The attached curly phase plot is what the receiver made of Domenico's 6.6-minute transmission at 20:40 UT - this was strong enough to let me easily tweak the freq offset, ending up at a magnificent 15.2 dB Eb/N0.

Best 73,

Markus (DF6NM)

-----Ursprüngliche Mitteilung----- 
Von: Paul Nicholson <vlf0403@abelian.org>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: So, 22 Nov 2015 10:06 pm
Betreff: Re: LF: EbNaut LF test from iz7slz

Markus wrote: 

 

 > your direct-sampling LF receiver actually seemed to be 

 > doing rather well! 

 

I think so, the rx itself hasn't had any problems. 

The timing system has got better too with the incentive 

to reduce the jitter without sacrificing speed of response 

to clock drift. 

 

It seems that whenever I get involved in amateur radio 

activities I end up with some improvements which get fed 

back into natural radio reception. 

 

I was thinking about oscillators.  At LF/MF if messages are 

short, say less than 30 minutes, the tx frequency only needs 

to drift by less than about 0.5 Hz/hour.  That's within range 

of a reasonably stable undisciplined oscillator such as an 

OCXO.   I'm thinking that the phase search in the decoder 

would accommodate that amount of drift.   Is it worth trying 

some transmissions which are not GPS locked? 

 

Rx battery is holding up here although I think it must give 

out overnight at some point.  Tomorrow I will recharge it. 

-- 

Paul Nicholson 

-- 

 

sym_11152043.png
Description: PNG image