Re: LF: 74550 - opds-64

[Markus Vester <markusvester@aol.com>]

Hi Bob, LF,

 

I am very pleased to find that your Opera-64 signal produced five correlation detections here last night. In the attached spectrogram, the symmetrical Opera modulation sidebands are clearly visible between the many DCF-77 interference lines.

 

Unfortunately the timestamps from SpecLab temporarily suffered from an unexplained offset, affecting both the spectrogram and the exported spectrum data. Such transient errors have also happened to other slow SpecLab instances, eg. 8.97 kHz - sometime after startup, time seems to spontaneously jump back, and suddenly comes back to normal after a few hours. In this case, timestamps were found to be exactly 5h 46m 4s early, allowing to derive corrected spot times (appended to the end of the lines here):

 

   date      time   call   distance  frequency bandwidth  snr    correlation
2013-12-09 07:53:55 WG2XRS  6448km  74549.500Hz   1mHz -49.5dBOp  95% 15.6dB
2013-12-09 05:42:27 WG2XRS  6448km  74549.500Hz   1mHz -47.3dBOp  82% 16.7dB
2013-12-09 04:36:41 WG2XRS  6448km  74549.500Hz   1mHz -44.4dBOp 100% 19.6dB
2013-12-08 22:50:39 WG2XRS  6448km  74549.500Hz   1mHz -45.2dBOp  91% 19.1dB  04:36:43 (same transmission as above)
2013-12-08 21:44:53 WG2XRS  6448km  74549.500Hz   1mHz -41.8dBOp  92% 19.3dB  03:30:57
2013-12-08 20:39:07 WG2XRS  6448km  74549.500Hz   1mHz -49.2dBOp  98% 16.1dB  02:25:11

 

The grass is always greener on the other side... Regarding the choice of frequency, the high resolution spectrogram seems to confirm the gap in the DCF77 interference around 74552.0 Hz. So Bob, Dex and others may consider moving up there, further reducing the QRM for other observers (eg. Hartmut) as well.

 

My understanding is that Opera's 28 bit payload can encode six character callsigns, but no extra appendix. This would result in ambiguities if several WG2XGR/n stations were transmitting simultaneously. As a simple workaround, I would suggest to replace the number "2" by the number from the appendix, eg. just use WG4XGR instead of WG2XGR/4 etc. One could argue that "sending the wrong callsign" might be considered illegal. But it would be unabiguous, and we could publicly announce it as an appropriate extension of the Opera encoding scheme.

 

Best 73,
Markus (DF6NM)

-----Ursprüngliche Mitteilung-----
Von: Markus Vester <markusvester@aol.com>
An: rsgb_lf_group <rsgb_lf_group@blacksheep.org>
Verschickt: So, 8 Dez 2013 11:05 pm
Betreff: Re: LF: 74550 - opds-64

Regarding opds-64 (or 65?) setup, I had forgotten to mention two more lines to be edited in opds.ini:

 opspeed=64    ' speed definition
 spotdelay=58  ' same timestamp offset as from Opera

 

In Graham's words, there's really a whole "cactus" of spikes in the 238 µHz spectrum. Most are on a 16.67 mHz raster, because large parts of the BCD time-message from DCF77 are repeating every minute. I have now recalibrated the frequency scale, and restricted the opds search range from 74548.1 to 74550.0 Hz.

 

Anyway given the QRSS levels from Bob last night I am quite optimistic.

 

Best 73,

Markus (DF6NM)

 

From: Markus Vester

Sent: Sunday, December 08, 2013 8:40 PM

To: rsgb_lf_group@blacksheep.org

Subject: LF: 74550

...

For the night to come, I'm set up for monitoring opds-64 in a 4 Hz slot around 74.55 kHz, and will be uploading screenshots and possible detections to the usual opds grabber http://dl.dropboxusercontent.com/u/26404526/opds.htm

In case others want to try: SpecLab FFT settings can be similar to opds-32 but more decimation is needed for half bandwidth (0.238 mHz). In opds.ini, center frequency and decimation should be edited: fc=74550, dec=50331648 .

 

Best 73,

Markus (DF6NM)

 

 

opds64_131209_0800.jpg
Description: image/pjpeg