QO100 beacon

[sv1bds]

Hi, I try to use it to receive QO100 beacon (Es Heil 2) without success until now. I can get a MER of about 10.2 using Polyphase Clock Sync & Costas Loop. But no constellation is archived after DVBS2 PL Sync. I suspect that I have to set ACM/VCM mode to False but I do not know what values must entered at PLS filter entries.

Any help is welcome !!!

[sv1bds]

DVBS2RX.PDF A simplified block is attached and the result screen. Thanks

[igorauad]

Hi @sv1bds ,

Thanks for your interest in the project. I'd recommend using the dvbs2-rx app instead of the hierarchical block, given the app has many options. I assume there is a way to pipe the PlutoSDR IQ output into dvbs2-rx, so you would run it like so:

pluto-sdr-source | dvbs2-rx --sink file --out-file /dev/null --gui --sym-rate 1.5M --samp-rate 3M --modcod qpsk4/5 --rolloff 0.25 --pilots auto

In this example, pluto-sdr-source is a hypothetical application feeding IQ samples into dvbs2-rx. If you don't have one, you can create one on GRC. You can feed the output of the PlutoSDR Source block directly into a File Descriptor Sink block configured for file descriptor 1 (stdout) and a complex input type. Then, generate the flowgraph and run it while piping the output IQ into dvbs2-rx.

I have never tried receiving this DATV beacon myself, so I'm not sure if the PLFRAMEs contain pilot symbols and if the FECFRAMEs are of short or normal size. I saw on your screenshot that you are configuring for short FECFRAMEs. If that's the case, you should add --frame-size short to the above command. Also, if you are sure there are no pilots, you can change from --pilots auto to --pilots off.

Alternatively, it is often helpful to run the low-level PHY block in ACM/VCM mode to double-check the unknown settings like FECFRAME size and the presence of pilots. To do so, run like so:

pluto-sdr-source | dvbs2-rx --sink file --out-file /dev/null --gui --sym-rate 1.5M --samp-rate 3M --modcod qpsk4/5 --rolloff 0.25 --pl-acm-vcm -d1

Option --pl-acm-vcm applies the ACM/VCM setting at low-level, while option -d1 sets the first debugging level (chose a higher number for even more logs). Then, if everything goes well, you might see lines like:

plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0

Based on that, you will know the correct settings. For example, the above would mean QPSK 4/5 with normal FECFRAME and no pilots. The MODCOD number you can see in Table 12 on the standard.

After that, you can go back to the regular command (without the --pl-acm-vcm flag) and just provide the exact configuration to it.

Let me know if that helps. And if you have an IQ recording to share, I'd be happy to try processing it.

Also, please note I have just pushed a couple of commits with some fixes and improvements. I recommend you pull them and try with the most up-to-date version. Besides, there was a fix there for the ACM/VCM mode.

[sv1bds]

Hi Igor, Thanks for your valuable information! To update code I just overide it? I prefer to write my own code as I write Python in my job. Your code is very useful for an end user. Thanks a lot for your effort!!! Any documention of using blocks will be appreciated. George SV1BDS

[igorauad]

Hi @sv1bds , apologies for the delay. I'm catching up with some pending issues now. Were you able to figure out how to use the blocks? If writing your own code is important, you can use the dvbs2-rx python code as a reference.

But no constellation is archived after DVBS2 PL Sync. I suspect that I have to set ACM/VCM mode to False but I do not know what values must entered at PLS filter entries.

Answering this question, note that the current version of gr-dvbs2rx does not support ACM/VCM fully. Only the PL Sync block supports ACM/VCM processing, and there the goal is to help with detecting the PLS settings (MODCOD, frame size and pilots) when these are not known a priori (e.g., trying to tune to an undocumented/unknown carrier). All blocks other than the PL Sync must be configured for a single MODCOD (CCM mode).

I understand you don't want ACM/VCM anyway, since you are trying to receive the QO100 beacon. For that, you can set the "ACM/VCM mode" option to False on the PL Sync block.

As for the PLS filter, even when operating in CCM mode, you may want to accept multiple PLS values. If you know the pilot configuration (on or off), then you can work with a single PLS value. Otherwise, you may want to accept two possible PLS values, one for pilots off, the other for pilots on. See how they are defined here: https://github.com/igorauad/gr-dvbs2rx/blob/master/apps/dvbs2-rx#L725. And see this routine for setting the PLS filters: https://github.com/igorauad/gr-dvbs2rx/blob/master/python/dvbs2rx/params.py#L231.

The low and high filters are two 64-bit (u64) variables with bit masks for the 128 possible PLS values. The PLS has 7 bits, so it has 128 possibilities. Each 7-bit value has the MODCOD (5 bits) and the TYPE field (2 bits), and the latter indicates whether the PLFRAME has normal or short size and whether it contains PL pilots.

The PLSC (Reed-Muller encoded) of each incoming PLFRAME is decoded and the PL Sync block checks whether the incoming PLS value is one of the accepted ones, according to the PLS filters. See https://github.com/igorauad/gr-dvbs2rx/blob/master/lib/plsync_cc_impl.cc#L917. If you end up using the wrong filters, the PL Sync block will achieve frame lock but won't output the decoded XFECFRAMEs downstream for the subsequent blocks (LDPC/BCH decoders, etc). You can tell whether this is happening by checking the count of rejected frames. With dvbs2-rx, if you run with option --log-stats, you will see this count at the frame_count.rejected field in the periodic JSON-formatted logs.

Let me know if that is helpful at all and whether I can close the issue for now.

Thanks!

[sv1bds]

I try this : dvbs2-rx --source rtl --gui --sym-rate 1.5M --samp-rate 3M --modcod QPSK4/5 --rolloff 0.25 --pilots auto -f 741500000 --rtl-gain 20 -d 3 /usr/local/bin/dvbs2-rx:16: DeprecationWarning: The distutils package is deprecated and slated for removal in Python 3.12. Use setuptools or check PEP 632 for potential alternatives from distutils.version import StrictVersion log :info: Starting DVB-S2 Rx log :warning: Warning: failed to XInitThreads() Warning: Ignoring XDG_SESSION_TYPE=wayland on Gnome. Use QT_QPA_PLATFORM=wayland to run on Wayland anyway. gr-osmosdr 0.2.0.0 (0.2.0) gnuradio 3.10.1.1 built-in source types: file fcd rtl rtl_tcp uhd hackrf bladerf rfspace airspy airspyhf soapy redpitaya freesrp Using device #0 Realtek RTL2838UHIDIR SN: 00000001 Found Rafael Micro R820T tuner [R82XX] PLL not locked! Exact sample rate is: 3000000,178814 Hz [R82XX] PLL not locked! Allocating 15 zero-copy buffers frame_sync :debug: -------------------------------------------------- frame_sync :debug: PLFRAME found frame_sync :debug: Peak after: 1233396; Timing Metric: 32.010540; Locked: 0 frame_sync :debug: Sym: 1233396; SOF: +14.6 +0.3j; PLSC: +17.4 %+.1fj plsync_cc :debug: SOF count: 1; Index: 1233306 freq_sync :debug: PLHEADER phase: 0.129828 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490 frame_sync :debug: -------------------------------------------------- frame_sync :debug: Peak after: 26004; Timing Metric: 31.689281; Locked: 0 frame_sync :debug: Sym: 26004; SOF: +14.7 +0.0j; PLSC: +16.9 %+.1fj plsync_cc :debug: SOF count: 2; Index: 1259310 freq_sync :debug: PLHEADER phase: 0.567732 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490 frame_sync :debug: -------------------------------------------------- frame_sync :debug: Peak after: 51961; Timing Metric: 30.537136; Locked: 0 frame_sync :debug: Sym: 51961; SOF: +13.1 +0.2j; PLSC: +17.4 %+.1fj plsync_cc :debug: SOF count: 3; Index: 1311271 freq_sync :debug: PLHEADER phase: 0.548318 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490 frame_sync :debug: -------------------------------------------------- frame_sync :debug: Peak after: 103828; Timing Metric: 30.980776; Locked: 0 frame_sync :debug: Sym: 103828; SOF: +14.0 +0.4j; PLSC: +17.0 %+.1fj plsync_cc :debug: SOF count: 4; Index: 1415099 freq_sync :debug: PLHEADER phase: 0.42755 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490 frame_sync :debug: -------------------------------------------------- frame_sync :debug: Peak after: 51961; Timing Metric: 34.359901; Locked: 0 frame_sync :debug: Sym: 51961; SOF: +15.1 -0.2j; PLSC: +19.2 %+.1fj plsync_cc :debug: SOF count: 5; Index: 1467060 freq_sync :debug: PLHEADER phase: 0.564757 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490 frame_sync :debug: --------------------------------------------------

It seems that I can not receive. Any help please ?

[igorauad]

Hi @sv1bds ,

That's good progress. I can see the timing metric is not that much strong yet. Ideally, the timing metric goes up to 57 when the signal has high SNR and the AGC performs its job properly. The frame sync algorithm will lock after only two consecutive timing metrics above 30 observed at the right (expected timing). That has not happened on your logs. I believe you would see that if the Peak After log were 32490. If this beacon signal is QPSK 4/5 with normal FECFRAME size and pilots off, then you have 361 slots of 90 symbols, hence an interval of 32490 symbols.

There are many things you can try. Perhaps you can message me on Telegram (username igorauad) for discussing in real-time if you wish.

I would try the following:

• Using a higher gain on the RTL-SDR. From 20 to 40 dB gain is the range I typically use.
• Try with --pilots off if the pilots are really off in this carrier.
• Try with --pl-acm-vcm -d3 if you are not sure about pilots and frame configuration.
• Try with the in-tree symbol sync algorithm using option --sym-sync-impl in-tree. That's a little more CPU-intensive but a bit more flexible in configurations. You can also play with the --sym-sync-loop-bw parameter, but let's leave that for later.

In any case, the signal looks pretty clear. I think you will get it working soon.

[sv1bds]

Hi, These are d 3 debuga data from a personal flowchart using Pluto for Rx. I am unsure for short or normal FEC setting. Can you determine the short/normal from these data? I get a MER of about 10 but still can not get 4 groups of dots at constellation. Any idea ? It seems that I get hi metrics. Why I do not lock? rame_sync :debug: -------------------------------------------------- frame_sync :debug: Peak after: 19069; Timing Metric: 59.164867; Locked: 0 frame_sync :debug: Sym: 19069; SOF: +25.7 +0.4j; PLSC: +33.5 %+.1fj plsync_cc :debug: SOF count: 545; Index: 16862263 freq_sync :debug: PLHEADER phase: 1.8342 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490 frame_sync :debug: -------------------------------------------------- frame_sync :debug: Peak after: 20197; Timing Metric: 59.724411; Locked: 0 frame_sync :debug: Sym: 20197; SOF: +25.9 -1.0j; PLSC: +33.8 %+.1fj plsync_cc :debug: SOF count: 546; Index: 16882460 freq_sync :debug: PLHEADER phase: 1.37593 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490 frame_sync :debug: -------------------------------------------------- frame_sync :debug: Peak after: 20559; Timing Metric: 62.262856; Locked: 0 frame_sync :debug: Sym: 20559; SOF: +27.9 -0.4j; PLSC: +34.3 %+.1fj plsync_cc :debug: SOF count: 547; Index: 16903019 freq_sync :debug: PLHEADER phase: 1.49037 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490 frame_sync :debug: -------------------------------------------------- frame_sync :debug: Peak after: 11734; Timing Metric: 56.077709; Locked: 0 frame_sync :debug: Sym: 11734; SOF: +24.9 +1.1j; PLSC: +31.1 %+.1fj plsync_cc :debug: SOF count: 548; Index: 16914753 freq_sync :debug: PLHEADER phase: 2.09109 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490 frame_sync :debug: -------------------------------------------------- frame_sync :debug: Peak after: 20203; Timing Metric: 60.772141; Locked: 0 frame_sync :debug: Sym: 20203; SOF: +27.1 +0.1j; PLSC: +33.6 %+.1fj plsync_cc :debug: SOF count: 549; Index: 16934956 freq_sync :debug: PLHEADER phase: 1.23626 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490 frame_sync :debug: -------------------------------------------------- frame_sync :debug: Peak after: 19985; Timing Metric: 60.077423; Locked: 0 frame_sync :debug: Sym: 19985; SOF: +25.9 -0.2j; PLSC: +34.2 %+.1fj plsync_cc :debug: SOF count: 550; Index: 16954941 freq_sync :debug: PLHEADER phase: 1.62958 plsc_decoder :debug: MODCOD: 8; Short FECFRAME: 0; Pilots: 0 plsc_decoder :debug: n_mod: 2; S: 360; PLFRAME length: 32490

[sv1bds]

Hi again, There is a bug in QO100 beacon and output is broken. I have to verify first proper beacon operation before testing your software... Thanks

[igorauad]

Hi @sv1bds ,

You will only see the four constellation points after the frame sync stage, whose success is indicated by the Locked: 1 frame_sync log. You haven't reached this state yet on the logs.

After frame synchronization, the other forms of synchronization kick in. The coarse/fine freq offset corrections and the phase correction. After that, the four constellation points become visible.

The timing metric is stronger on the latest logs but the interval between frame sync timing metric peaks is not consistent yet. Do you have any IQ data loss by any chance? If you are using an RTL-SDR, I'd recommend testing the RTL with rtl_test -s 3000000. Not all RTL-SDRs will be able to sample at 3 Msps without IQ loss. Not sure if that's your case.

[sv1bds]

Hi, Yes you are right! It loose data. It can receive to about 2400 max.

[igorauad]

Hi, @sv1bds . It seems like I can close this issue for now. Let me know if you need any further help, and feel free to reopen the issue if necessary.