Blindscan on linux with tbs cards

And if you would to try to change the demod to rf input assignment, you could look in tbsecp3-dvb.c
line 2036 and 2039.
Replace
tbs6909x_stid135_cfg, adapter->nr, adapter->nr/2)
with
tbs6909x_stid135_cfg, adapter->nr, adapter->nr%4)

Note that there are two lines: one fr version 1 of the card and 1 for version 2...

This will map as follows:
adapter 0 and an adapter 4 will be attached to RF0
adapter 1 and an adapter 2 will be attached to RF1
...
This would allow all 4 RF inputs to be used, but still not more than 4 simultaneous fft scans.
Obvously, in your user programs you need to use new adapter numbers. E.g., the the old adapter2 will
now be called adapter1 in tvheadend.

It is possible that this will lead to a driver crash, which then indicates a bug in the stid135-fe or
stid135_drv code (e.g., some programmer assuming that rf_input == adapter_no/2).

If you test this, please report back, because similar changes could be useful for tbs6903x and other
cards. This could be implemented as a driver option. In the long term I plan to make the RF to adapter mapping
fully configurable, but this requires quite a few changes. This will be more versatile.
For instance user space programs will be able to attach all 8 tuners dynamically to any of the inputs.
Or switching between tuner_mode will no longer require reloading the drivers.
 
I have verified that the above adapter renumbering allows spectrum acquisition on all rf inputs
The main downside is that the slave tuners are no2 4,5,6 and 7, but this can be "undone"
by adding
options tbsecp3 adapter_nr=0,2,4,6,1,3,5,7
in /etc/modprobe.d/tbsecp3.conf
and rebooting or reloading the drivers
 
after adding options tbsecp3 adapter_nr=0,2,4,6,1,3,5,7, fft is working for only 0, 1, 6 and 7
i dont know why but its not working in all 8 adapters!
 
actually its working randomly or having some pattern i guess
after few mins or using different adapters, it started working adapters which were not before
but only 4 adapters works at a time
 
stid135 have four FFT cells assigned to first four virtual demods.
 
Indeed: only four FFT modules on the card. There is a bug that I discovered recently: diseqc commmands
are not sent properly before a spectrum scan which means the spectrum can be from the wrong sat if
you use diseqc switches. It will be fixed in the next version, along with other bugs
 
can anyone please share formula to calculate bitrate of carrier/transponder?
 
Are this related with cmake or clang++ compile error
Code: 
[ 16%] Building CXX object src/CMakeFiles/neumo-blindscan.dir/neumo-blindscan.cc.o
/home/user/blindscan/src/neumo-blindscan.cc:1649:8: warning: decomposition declarations are a C++17 extension [-Wc++17-extensions]
                auto [found_freq, bw2] =
                     ^~~~~~~~~~~~~~~~~
/home/user/blindscan/src/neumo-blindscan.cc:1723:11: warning: decomposition declarations are a C++17 extension [-Wc++17-extensions]
                                        auto [found_freq, bw2] = getinfo(fpout_bs, fefd, pol_is_v, 0, band);
                                             ^~~~~~~~~~~~~~~~~
/home/user/blindscan/src/neumo-blindscan.cc:1948:10: error: no member named 'filesystem' in namespace 'std'
        if(std::filesystem::exists("/sys/module/dvb_core/info/version")) {
           ~~~~~^
2 warnings and 1 error generated.
make[2]: *** [src/CMakeFiles/neumo-blindscan.dir/build.make:76: src/CMakeFiles/neumo-blindscan.dir/neumo-blindscan.cc.o] Error 1
make[1]: *** [CMakeFiles/Makefile2:100: src/CMakeFiles/neumo-blindscan.dir/all] Error 2
make: *** [Makefile:136: all] Error 2

Debian clang version 13.0.1-3
cmake version 3.22.0
gcc version 11.2.0 (Debian 11.2.0-16)
SMP Debian 5.14.9-2 (2021-10-03) x86_64 GNU/Linux
 
Last edited:
Perhaps you use the wrong compiler?
 
How to check if i use wrong compiler, maybe need more old version of clang or gcc ?
 
I manage to compile, but are this work with tbs 6522 its si2183 based ?
Second are this blindscan will work on terminal only, without desktop ?
 
Well, why don;t you try it and see if it works? si2183 is supported but it was never tested on 6522.
There is command line blindscan and gui blindscan software available.
 
I have been working on improving the blind scan drivers for the montage m88rs6060 chip as used in the tbs6904se card.
Blind scan is partially working and the most recent success was to extract a high-quality spectrum from the card. The resolution is
187kHz. Not as good as with stid135 and there is a risk on some sporadic artefacts. The scan below is part of the high-horizontal band
on 28.2E. Capturing a complete horizontal spectrum takes less than 30 seconds. So, it is pretty fast as well.

The code will appear on github soon.

spectrum.webp
 
Can't wait for the driver for 6904se (also available for 6902se? same chip). But until then, I'd like to point out (I hope I'm wrong) the missing letter "s" in the command: ./install.h:
...
make dir DIR=../media
make distclean
./install.h
This is the only way I could install the drivers, but unfortunately, with an error 2 (on screenshot).
 

Attachments

  • bitrate & error 2 .webp
    bitrate & error 2 .webp
    82.3 KB · Views: 25
Last edited:
satesco said:
Can't wait for the driver for 6904se (also available for 6902se? same chip). But until then, I'd like to point out (I hope I'm wrong) the missing letter "s" in the command: ./install.h:
...
make dir DIR=../media
make distclean
./install.h
This is the only way I could install the drivers, but unfortunately, with an error 2 (on screenshot).
Click to expand...
The ./install.sh file extension denotes a shell script and assuming it has been made executable I can’t see how it could possibly run by dropping the s in the file extension name. As far as I can see that install file hasn’t run at all. No doubt deepthou will shed more light on this. Err 2 is an indicator.
 
satesco said:
Can't wait for the driver for 6904se (also available for 6902se? same chip). But until then, I'd like to point out (I hope I'm wrong) the missing letter "s" in the command: ./install.h:
...
make dir DIR=../media
make distclean
./install.h
This is the only way I could install the drivers, but unfortunately, with an error 2 (on screenshot).
Click to expand...

Satesco,
indeed. it should be install.sh. I will correct this part of the documentation in the next version.

6902se should also work, but the proof is in the eating of the pudding as they say.
 
Satspot said:
The ./install.sh file extension denotes a shell script and assuming it has been made executable I can’t see how it could possibly run by dropping the s in the file extension name. As far as I can see that install file hasn’t run at all. No doubt deepthou will shed more light on this. Err 2 is an indicator.
Click to expand...
Just a typo....
 
New blindscan drivers released.,

The main change is support for m88rs6060 chips. This make tbs6902SE and tbs6904SE work
with the blindscan code and with neumoDVB. Everything works, except constellation display.

These cards have the unique feature that they support automatic detection oif the Physical Scrambling Code.
This is supported as well.

The cards also feature an FFT processor, and the drivers exploit this to capture a spectrum with a resolution of
187kHZ.

The drivers also support low bit rate muxes (e.g., Alouette, 667 kSs on 5.0W), but beware that small LNB local oscillator
deviations can cause problem. Also, the spectrum is not fine enough to detect these muxes. So you have to enter an approximate
frequency and symbol rate for the initial scan.

All changes in release-0.9.0

  • Blindscan and spectrum upport for montage m88rs60606 based cards: tbs6902SE and tbs2904SE.
  • Support m883s6060 module unloading
  • stid135: BER is now reported properly othrough the PRE instead of POST error counters
  • Improved correction of discontinuities in spectrum due to small errors in RF level
  • stid135: detected multi-stream IDs are now accumalated internally
  • Added DTV_BIT_RATE readout via dvbapi
  • stid135: improper estimation of required llr rate, causing rai multistream on 5.0W to not tune sometimes (green blocks in picture)
  • stid135: changed default spectral resolution: slightly slower scan, but more narrow-band muxes found
  • Add support for timing lock flag
  • stv091x: return more correct data about currently tuned mux (instead of returning what user requested)
 
Well, Frankviana,

unless I am mistake, ebspro does not run on linux.
So you have a a number of options
1) install linux and use my drivers and then post something good or bad in this thread
2) keep using windows, test the latest drivers and open an another thread to complain about them.
3) complain to the manufacturer
4) keep waiting
5) combination of one of the above

Remember that blindscan on windows (and linux) is implemented by volunteers. You could encourage the ebspro
and streamreader developers. The best way is by a constructive attitude, e.g,. by reporting to them specific, clearly decribed
problems with the latest version of their software and/or making a financial donation to them.

PS. I am not one of them.
 
Back
Top