_________________________________________________________ From: Markus Vester To: rsgb_lf_group@blacksheep.org Subject: XGJ, NA, XES and DRP (Feb 21/22) Date: Wed, 22 Feb 2006 5:42:11 PM Eastern Standard Time Dear John, Warren and Joe, last night, strong TA signals from XGJ, NA and XES appeared on the grabber screen. The phase masurement on John's carrier looked like this (same parameters as yesterday): The results of the CT1DRP phase display in the Eu-grabber panel during the last 24 hours are not quite clear to me. There seemed to be a systematic, slow colour cycling (about 360 deg/hour "clockwise") on both DFCW frequencies (22:00-22:00, 2 min/pix, 2 hour ticks): I first suspected that the 0.1 Hz software modulator used to split up the 136318 Hz GPS line might be inaccurate, however this would have caused equal but opposite phase roll on the two frequencies (one too high, the other too low). So it looks to me as if Brian's frequency was 2E-9 low on average, but then again, this doesn't really make sense for a GPS locked TX. Could it somehow have to do with the DFCW switching? 73 and best of luck Markus, DF6NM _________________________________________________________ From: Markus Vester To: rsgb_lf_group@blacksheep.org Subject: Re: LF: XES Carrier tonight Date: Tue, 21 Feb 2006 6:28:09 PM Eastern Standard Time Hi John and LF, here is the result of last night's phase measurement on the WD2XES carrier: This panel displays three amplitude and phase spectra, differing only in the a posteriori frequency shift applied to the right (H-field antenna) channel used for phase reference. Each segment shows the whole night's data from 22:34 to 8:30, at 5 min/pix horizontally and 0.657 mHz FFT resolution (10 mHz/tick) vertically. - left: no frequency shift applied. This is the normal colour-DF representation, referencing the phase of XES on the E-field antenna to that of the same signal on the H-field turnstyle. The westerly trace can be seen in blue against a background of purple-ish noise, originating partly from northerly Luxembourg-ICM on DCF39. Frequent grabber watchers will find it familiar ;-) - center: frequency shift +0.309506014 Hz. Same as yesterday, this brings the 137577.00 GPS line to XES's frequency and measures the phase between the two. - right: frequency shift -0.690493986 Hz, bringing the next GPS line on 137578.00 underneath XES. After applying an initial constant phase of 150° (reflecting the arbitrary start time of the recording vs. the 1 pps), the trace colours look very similar to those in the segment before. This proves that the phase measurement is not corrupted by soundcard samplerate errors, which would have caused phase drifts in opposite directions. Ok, so here we have a validated and relatively simple (?) method to measure propagational phase variations, which can be done without a fully GPS-locked receiver. But, what's it good for? Propagational phase and amplitude variations broaden the transmitted linewidth by a fraction of a milliHertz. Thus the stability of the channel sets an ultimate limit for extremely slow QRSS or PSK communication, and therefore the minimum transmit power requirement to overcome a given noise energy. Phase stability over an hour or longer suggests that we could still gain another 15 dB by going from 120s to 3600s symbol duration. Both yesterday and today, there were two relatively long stable periods, separated by an interval of fading and more rapid phase changes. In yesterday's high-resolution spectrogram (top right), one can even find an interesting Doppler shift between the openings, where the line dips about 3 pixels (0.4 mHz) below the center frequency, and then slowly creeps back up. This is most likely a real effect, caused by changing ionospheric reflection heights. Such Doppler spectroscopy on LF could provide interesting insights into the lower ionosphere, as it does on HF for the upper layers. Have started SndInput for another night, and also set up an experiment using SpecLab's 0.1 Hz amplitude modulator to shift a GPS line behind Brian's coherent transmission on 136317.9 / 318.1 DFCW - funny effects so far. Best 73 to all Markus, DF6NM In einer eMail vom 21.02.2006 21:26:00 Westeuropäische Normalzeit schreibt w1tag@w1tag.com: Thanks for the reports on the "dead carrier" last night. I've had a request to repeat it again tonight, so XES will be back on 137.57730950601398944854736328125 kHz again, starting at 2230 UTC. _________________________________________________________ Thema: LF: XES Carrier tonight Datum: 21.02.2006 21:26:00 Westeuropäische Normalzeit Von: w1tag@w1tag.com Beantworten: rsgb_lf_group@blacksheep.org An: rsgb_lf_group@blacksheep.org Internet-eMail: (Details) Thanks for the reports on the "dead carrier" last night. I've had a request to repeat it again tonight, so XES will be back on 137.57730950601398944854736328125 kHz again, starting at 2230 UTC. The CW ID is at 30 minute intervals, and should be rather exactly at the top and bottom of the hour. The next step will be to send a little repeating PSK message, probably with 2-min/bit timing. Details tomorrow. Markus: Special thanks for your analysis thus far. One obvious issue is the phase stability of this sort of path, and the information will be useful. Am very interested to see any results from last night's unkeyed run. John Andrews, W1TAG/WD2XES _________________________________________________________ Thema: Re: LF: Slow PSK on Feb. 19/20 Datum: 21.02.2006 09:18:46 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org Doo-be-doo, two and two... In my last posting, there were at least two errors concerning missing factors of two, one obviously in the subject line. The other one regarding the narrowband FFT resolutions, that should have said 129.8 µHz (~ two hours of coherent integration) and 0.52 mHz instead of 64.9 and 0.26 - I had missed the doubled samplerate when resampling from IQ to real-valued wav format. Saw traces of John's silent carrier on the grabber between midnight and 06:30; narrowband evaluation will have to wait till the evening. As Jean-Pierre mentioned, static levels were quite high, nevertheless VO1NA "O" at 06:00. 73 de Markus (off to work) _________________________________________________________ Thema: Slow PSK on Feb. 19/10 Datum: 21.02.2006, 00:29 UT An: rsgb_lf_group@blacksheep.org Dear John and LF, WD2XES was received again here last night. On Argo 120 at 5.4 mHz resolution, the signal looked like this (22:15 to 8:15): <137577_Argo120_060219-20_2215-0815_small.jpg> A few hints of the PSK blips can be guessed, but nowhere as clear as in the UK recordings from Jim and Dave. No other signal was seen within a couple of Hz. Processing the narrowband recording produced the following panel. It shows XES on the E-field antenna at the top, and a GPS 1pps harmonic (coupled via 33 pF to the H-field antenna) on 137577.0 at the bottom. On the left side, there is a spectrogram having 64.5 microHertz FFT resolution, 10 minutes per pixel from 19:49 to 9:07 UT. On the right, the signal was despread by multiplying with a properly timed 2x15 minute PSK sequence, as done before: The lower trace shows that the OCXO, though still free running, provided much better stability than the simple homebrew heated crystal. Total frequency variation improved from 3.5 mHz last time to better than one pixel (0.065 mHz), including the effect of the soundcard on the 952 Hz audio. The stability of the traces encouraged me to try to make an absolute phase measurement and look for propagational variations. To this end, the GPS carrier on the right stereo channel was shifted by up by 0.309506014 Hz in software, and compared against the left channel containing the unspread signal from XES, using SpecLab's colour-DF setting to display phase difference as colour hue. This screenshot has 0.26 mHz FFT resolution and was flipped to have HF up: With this technique, any LO drift is eliminated, leaving only soundcard samplerate deviations, but these affect only the small frequency difference to the nearest Hz. For 0.31 Hz offset, the estimated calibration error of 0.05 Hz would accumulate a phase error of no more than +-25 degrees in 13.3 hours. The samplerate could also be accounted for if the recording had been wide enough to contain more than a single GPS line. This was not the case last night as I used a decimated rate of 11027.15Hz/10368. Other sources of phase errors could lie in the transmit and receive hardware (especially the narrowband antennas), or in the common GPS timing, but presumebly these are small compared to the skywave propagation effects. In case anyone is interested in the decimated audio data, I uploadeded it to http://members.aol.com/df6nm2/SlowPSK/ . The dat files generated by DL4YHF's SndInput tool are 4 bytes per sample (8 bits IQ stereo, LI LQ RI RQ), without a header. For tonight, I have switched back to 4096-fold decimation since 22:34. At the moment things look rather noisy - but we'll see. 73 and best wishes Markus, DF6NM _________________________________________________________ Thema: Re: LF: Slow BPSK Test Datum: 20.02.2006 23:03:13 Westeuropäische Normalzeit Von: w1tag@w1tag.com Beantworten: rsgb_lf_group@blacksheep.org An: rsgb_lf_group@blacksheep.org Tonight's WD2XES test will be a dead carrier on 137,577.309506014 Hz, interrupted only by CW ID's on the top and bottom of the hour. There will be no (intentional) phase changes at this end. The carrier went on at 2200 UTC. John Andrews, W1TAG/WD2XES _________________________________________________________ Thema: Re: LF: Slow BPSK Test Datum: 19.02.2006 23:46:55 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org Hi Scott, corona - boy there's a few stories I can tell. With my ever-too-small antenna made from a pair of 0.5 mm enameled copper wires, I have to run 15 kV rms to radiate 200 mW, and it always glows at one point or another: http://members.aol.com/df6nm2/antenna_ursa_major.jpg I have set up the decimated audio around John's 137577.31, so if you can get on, I guess it would be best if you joined in within +- 0.4 Hz from his QRG. I am now working on injecting a 1pps GPS reference to the second antenna input. This will (in theory) allow phase measurements if you transmitted spot on 137577.000 tonight, but bearing the risk of local QRM. 73 and good luck Markus In einer eMail vom 19.02.2006 23:02:36 Westeuropäische Normalzeit schreibt sthed475@telus.net: I'm headed out to repair the last remaining discharge point and will advise if I'll be running tonight. I could run in the wateringhole on 137778 as Steve is away from his station tonight. _________________________________________________________ Thema: Re: LF: Slow BPSK Test Datum: 19.02.2006 22:00:33 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org hmm sorry John, that was Saturday, right? Markus In einer eMail vom 19.02.2006 21:53:55 Westeuropäische Normalzeit schreibt MarkusVester@aol.com: I viewed XES on Friday morning using Argo 120 slow _________________________________________________________ Thema: Re: LF: Slow BPSK Test Datum: 19.02.2006 21:53:19 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org Hi John, Scott and LF Group, the gear is set up for receiving 137577.3 +-0.5 Hz tonight, recording 11025/10368 IQ-samples per second. I have connected a "proper" OCXO so stability should be better than last time, but still not GPS-locked. Surely I'd be pleased to find more than one signal - at 1.1 mHz per channel, there's plenty of space in a Hertz for anyone who wants to join in ;-) Re Argo speed - I viewed XES on Friday morning using Argo 120 slow, which produced a couple of blips around 4 am. I looked through the data from Scott's channel 137778 but nothing found - how long did VE7TIL stay on last night? 73, and best of luck Markus In einer eMail vom 19.02.2006 20:28:35 Westeuropäische Normalzeit schreibt w1tag@w1tag.com: WD2XES will try a run of slow BPSK on the following very odd frequency tonight: 137,577.309506014 Hz. This frequency appears to be clear of Loran lines for 1 Hz either side over a fairly large area, and is actually one of the rounder frequencies available from a 32 bit DDS clocked at 5 MHz. I will again use the 15 minute 0/180 degree keying, which would produce sidebands 0.55 mHz either side of the above frequency. There will be a CW ID at each 15 minute point. Start time will be 2230 UTC. Reports are welcome. _________________________________________________________ Thema: VE7TIL transmission Datum: 19.02.2006 02:42:41 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org Scott, >> TIL will go into key down mode on 137778.00Hz and will phase shift on the following sked starting at 0100UTC. H, H+15, H+30, H+45<< ok, all set up for monitoring, including NPG Dixon on the lowest grabber trace. Good luck! >> As a side note the combined stages are far more stable then a lone stage... The DC input current varies 1/2 as much as a lone stage with antenna variations in the wind... I expect this to improve even further with the 4 port combiner used for the 2KW version.<< - a very interesting point. Would that imply it is possible to partly absorb load reflections in the combiners? I had always thought that you'd need a nonreciprocal device like a circulator for that purpose - which might be sort of bulky at LF ;-) The reduced effect on current could possibly be due to a different phase of the reflection coefficient. 73, hope to see you before meltdown ;-) Markus, DF6NM _________________________________________________________ Thema: Re: LF: Slow PSK result Datum: 19.02.2006 01:04:44 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org Hi Scott, thank you very much. As I don't currently have a way to lock the receiver and soundcard to GPS, the best I can do is narrow spectrograms, and there are no magic numbers regarding frequency selection. I would probably select a frequency in the TA waterhole (137776 to 781) as it is monitored by several people, including Hartmut whose antennas are some 10 dB better than mine. 73 and best of luck, Markus, DF6NM In einer eMail vom 19.02.2006 00:15:31 Westeuropäische Normalzeit schreibt sthed475@telus.net: If you like I can place TIL on this evening starting at local sunset ~0100UTC on a different frequency then John. I'd like to do some key down tests on the new 1KW amp module so I could kill two birds with one stone. Please advise on a frequency that would work. I can do anything a AD9850 will allow feed with a 20MHz disp. osc. will allow. _________________________________________________________ Thema: Re: SLowBPSK result Datum: 19.02.2006 00:55:21 Westeuropäische Normalzeit Von: Markus Vester An: peter.martinez@btinternet.com Hi Peter, thank you very much for your explanations, this is very interesting stuff, and a pleasure to read. Yes I had some clues about your work as Scott 'TIL forwarded your comments regarding his transmissions, along with data showing your (non)reception of his signal. Actually when I tried to run the spreadsheet I couldn't get the FFT function to work. Being not familiar enough with Excel, I exported the raw data to MathCad, and saw all the features you described. One thing that did amaze me then, and does again today, is how narrow the received Loran line really is - I had expected a couple of milliHertz skywave propagational spread. There was a mention of Clicklock but I didn't understand then what the name means. It looks like it's exactly what I need! I have fed GPS pulses to the antenna via a small coupling capacitor, and it worked very well. I have an old Motorola Oncore GT, and added a monoflop to get rid of the undefined falling edge. This unit has about 1 µs jitter but that doesn't hurt on LF, and averages down to about 50 ns or so in a minute. The good thing about coupling to the antenna is that it eliminates the phase variation caused by environmental antenna resonance shifts. I had fed the 1pps to the second (turnstyle H-field) channel of my colour-DF system, hoping to see propagational phase shifts on Brian's signal as colour changes. The difficulty was then that Brian transmitted on 0.2 Hz multiples. I should have made a /5 divider for the 1 pps, but instead I fiddled with artificial refernce lines from SpecLab's software signal generator, having to interpret the phase walk differences between two unstable lines in the end - not too comfortable. My understanding of John's frequency selection was his use of 5 MHz as a DDS reference, so apparently quarter-Hertz multiples have to be multiples of 19531.25 Hz as well. On the other hand, any other DDS frequency would be rounded but still perfectly predictable, and you could probably shift the frequency in software by any known offset. I have been reluctant to seriously offer contributions to your experiment for two reasons: My local receive situation is average at best, and Hartmut Wolff routinely gets perhaps 10 dB better SNR. The other one is the frequency stability issue. I can do a little better than last night by using a "proper" OCXO having a couple of ppb, and I could probably build some hardware to lock it to GPS or even Loran groundwave. But then there is still the soundcard clock issue which would have to be fixed by software. With the LO modifications I could also transmit synchroneously but I'm not sure if your project would really benefit from another Eu signal. Also, this TX antenna is very weather dependant (well, better call it fragile), and can be used only somewhat sporadically. One thing I am still wondering about is if you are really targeting quarter-hour PSK in the end, or if something like 1 bit per minute (which could still be synchronized manually) might be better suited to LF. BTW I have done a bit of work on Loran skywave monitoring in time domain, which included carrier phase measurements and samplerate referencing to the GRI. I still have to get around writing a proper description, but if you are interested there are some results on http://people.freenet.de/df6nm/LoranView/LoranView_2006.htm Best wishes, Markus, DF6NM In einer eMail vom 18.02.2006 23:44:09 Westeuropäische Normalzeit schreibt peter.martinez@btinternet.com: Markus: John Andrews showed me your waterfall of his signal on 136.71875 which he was transmitting for me. That was very interesting for me because it helped me to confirm how my own system is working. For info, attached is the spectrum that I saw at the same time. It's a 64-point FFT with I/Q samples integrated over 225 seconds, taken over the four hours 22:30-02:30. The centre line is the GRI 8000 line, probably mainly from the Slonim transmitter, and you can see the +/-0.555 mHz lines, and just detect the +/-1.666mHz lines 3rd-harmonics of the squarewave from John. The width of the whole spectrum is 4.44mHz. The centre peak appears to be shifted slightly to the right of the centre graticule line, but this is not true. It's a bug in the Excel chart plotting. The raw data shows that the centre line is indeed in the centre of the spectrum. I had asked John to transmit 15-minute BPSK so that his signal (which had to be on 136.71875 because it was the only frequency we could both use) could be seen each side of the LORAN line! My Clicklock system removes completely the receiver drift by injecting a harmonic-rich pulse from the GPS into the antenna and locking in time and phase to it, so the software is as stable as the GPS even if the receiver is drifting. I can now receive on any frequency and not just round-number fractions of 1Hz as I was doing last night. This will make it possible to try coding some data with the BPSK modulation, because we can now QSY away from the GRI 8000 line frequency. I plan to try some other ideas soon. I can give you more details if you are interested. I really would like other people to experiment with this technique. I am not active on the LF bands myself. 73 Peter G3PLX _________________________________________________________ Thema: Re: LF: Slow PSK result Datum: 18.02.2006 23:15:52 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org Klicken Sie mit der rechten Maustaste auf das Objekt, um die Optionen anzuzeigen. Dear John and LF, based on an idea by Peter Martinez, I then tried to despread the PSK signal by multiplying with a 15+15 minute (+1, -1) sequence before doing the Fourier Transforms. This works as expected, splitting up the Loran carrier and focusing John's signal to the center frequency: The recording was done using the E-field channel of my normal colour-DF grabber hardware, ie. SSB receiver and standard soundcard. It might be worthwhile to add in 1 Hz reference lines by coupling GPS 1pps pulses into the RX input. One could then possibly use SpecLab's tracking function to compensate for the LO frequency drift, but I haven't tried this in such a configuration. John, are you intending to transmit again tonight? 73 de Markus, DF6NM _________________________________________________________ Thema: Slow PSK result Datum: 18.02.2006 20:53:46 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org Klicken Sie mit der rechten Maustaste auf das Objekt, um die Optionen anzuzeigen. Hi John and LF Group, here is the result from last night's recording of John's slow PSK transmission: This ultra-narrow spectrogram has a resolution of 0.329 milliHertz (16 times finer than Argo 120), and the duration is 13.2 hours. The wandering and fading central line is Loran GRI 8000 from Russia, with my receiver drifting up and down by about 3.5 mHz. Between 3:45 and 5:45, the line is accompanied by +- 0.56 mHz sidebands; these are John's PSK transmission with 30 minute modulation period. 73, and thank you for the interesting experiment Markus, DF6NM _________________________________________________________ Thema: TA Feb 18 Datum: 18.02.2006 11:26:01 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org Dear Group, excellent TA signals from XNS (partly readable here for the first time), XGJ and NA - see http://members.aol.com/df6nm2/TA_Feb06/TA0602180600.gif (291kB). On 136718.75, John's XES enganged in a struggle with the Russian line, with some interesting results on Argo-120 between 03:00 and 06:00: 73, and have a nice weekend everybody Markus, DF6NM _________________________________________________________ Thema: Re: LF: Slow BPSK Test Datum: 18.02.2006 00:49:02 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org Hartmut, I believe such magnitude jumps are to be expected, due to the superposition of a phase-reversed signal from John and a constant-phase carrier from the Russian Loran. Markus In einer eMail vom 18.02.2006 00:40:42 Westeuropäische Normalzeit schreibt w1tag@w1tag.com: Hartmut, No, the power has been steady, as far as I know. We were out for about an hour after the start-up, though. The AC power was out here when I came home from work, so I'm hoping that the GPS-sync got re-established quickly. John A. Hartmut Wolff wrote: >John, >have you reduced power since 23:15? > >Hartmut _________________________________________________________ Thema: Re: LF: Slow BPSK Test Datum: 18.02.2006 00:12:05 Westeuropäische Normalzeit Von: Markus Vester An: rsgb_lf_group@blacksheep.org John and LF Group, I have started to record decimated IQ audio in a 2.7 Hz channel around 136718.75 for offline analysis. Unfortunately my receiver is not GPS-locked, just a simple temperature controlled crystal. I do see the Chayka line on slow Argo and will look out for systematic amplitude variations, but I have little hope to view the phase-switching blips directly from here. Best wishes Markus