SATPI icon indicating copy to clipboard operation
SATPI copied to clipboard

Only 2 streams works on my vuduo4kse with 2 x FBC DVB-S2X

Open ddarek opened this issue 3 years ago • 34 comments

Hi, I have Vu+ Duo 4K SE with 2 x FBC DVB-S2X.

In web interface of satpi I see that only frontend 0 and 1 are "FBC-Root". I think that also frontend 8 and 9 should be "FBC-Root".

When I connect to SATPI with Tvheadend only two frontends can be used.

I have 2xFBC DVB-S2X phisical tuners. I see Frontends 0,1,2,3,4,5,6,7,8,9,10,11,12,14,15 0,1 - Root of first DVB-S2X 8,9 - Root of second DVB-S2x 2,3,4,5,6,7 - virtuals 10,11,12,13,14,15 - virtuals

In Tvheadend only 2 stream at the same time works ok, if more then streams are interrupted, no tuned.

Do you know the reason why it don't work? Maybe my configuration is wrong? Should I see 4 root frontends in case of my device ?

Best regards Dariusz

ddarek avatar Feb 06 '22 16:02 ddarek

Hello @ddarek ,

I only own receivers with only 1 FBC tuner DVB-S2 (So only 0 - 7). I have a GB quad which can fit one extra, but currently empty. So sorry, I have not tested this situation myself.

A few questions:

  • Is enigma2 still runing
  • Which tuners are used for the streams?
  • Do you use Unicable/Jess LNBs

Barracuda09 avatar Feb 06 '22 17:02 Barracuda09

Hi, Thanks for reply, I have tested without running enigma2. In enigma2:

  • I have 4 frontends configured: a,b, i, j (simple, diseqcx4 with a and b set for astra and hb).
  • The rest of frontends (c,d,e,f,g,,k,l,m,n,o,p) are configured automatically.

I have legacy Quad LNB connected with 4 cables to those 2fbc modules.

Satpi web int. says that only frontend1 and frontend2 are Fbc-root.

Only 2 streams works stable (a and b). If I try to use more than 2, then they are unstable.

It looks like frontend 8 and 9 are treated like virtual for first dvbs2x card. In enigma2 frontend 8 and 9 are treated like a roots( can be configured) for virtual 10,11,12,13,14,15.

ddarek avatar Feb 06 '22 17:02 ddarek

In my case I have: in /proc/stb/frontend i have 15 entries. first 8 (0-7) has /proc/stb/frontend/x/fbc_set_id with value 0 (I think it means: dvbs2x card 1) second 8 (8-15) has /proc/stb/frontend/x/fbc_set_id with value 1 (I think it means: dvbs2x card 2)

in /dev/dvb/ I have only adapter0

ddarek avatar Feb 06 '22 17:02 ddarek

Yes, I did not foresee this, with 2 FBC tuners.

Barracuda09 avatar Feb 06 '22 17:02 Barracuda09

Could you add support for that case ?

ddarek avatar Feb 06 '22 18:02 ddarek

Yes, I could do that. But first I need to purchase a second tuner for my GB, to test it myself.

Edit: But i am not sure that is the same, because GB do not support second FBC tuners (as I understand it now)

Barracuda09 avatar Feb 06 '22 18:02 Barracuda09

@ddarek

Could you please provide me the contents of dir (recursive):

  • /proc/stb/frontend

Could you give me the contents of these files:

  • /proc/stb/frontend/xx/fbc_connect_choices
  • /proc/stb/frontend/xx/fbc_set_id
  • /proc/stb/frontend/xx/fbc_link
  • /proc/stb/frontend/xx/fbc_connect

Barracuda09 avatar Feb 07 '22 08:02 Barracuda09

Sure, Here's the full, recursive content of dir /proc/stb/frontent frontend.tar.gz

ddarek avatar Feb 07 '22 12:02 ddarek

Did you change manually contents of these files:

  • /frontend/8/fbc_connect to 8
  • /frontend/9/fbc_connect to 9

Barracuda09 avatar Feb 07 '22 12:02 Barracuda09

No, I didn't change anything in /proc. Only enigma2 is working with configured 0,1,8,9 adapters.

ddarek avatar Feb 07 '22 13:02 ddarek

Ok, that nice. So there is just on offset for the second FBC tuner.

Barracuda09 avatar Feb 07 '22 13:02 Barracuda09

Yes, I think so.

ddarek avatar Feb 07 '22 13:02 ddarek

Did you build from source or from image repo

Barracuda09 avatar Feb 07 '22 13:02 Barracuda09

image repo

ddarek avatar Feb 07 '22 13:02 ddarek

Ok, possible thats quite an old version. Could you try to build it from source?

Barracuda09 avatar Feb 07 '22 13:02 Barracuda09

Ok, but later this evening. I have to refresh my memory about toolchain... :)

ddarek avatar Feb 07 '22 13:02 ddarek

Try to look here: https://github.com/Barracuda09/SATPI/wiki/Cross-Compiling-SatPI

Barracuda09 avatar Feb 07 '22 13:02 Barracuda09

Hi, Sorry for a delay but I had no time yesterday. I have :

  • killed enigma2 process.
  • started satpi with enabled 0,1,8,9 frontends
  • set diseqc for all those frontends
  • set channel a frontend connections : 0=A,1=B,8=A,9=B
  • restarted satpi process
  • made copy of /proc/stb/frontend
  • saved log file.

frontend_and_log.tar.gz

ddarek avatar Feb 08 '22 18:02 ddarek

Did you change in the webUI the frontend 8 and 9 option FBC Connection

Barracuda09 avatar Feb 08 '22 19:02 Barracuda09

yes Frontend 8 : FBC Connection=A Frontend 9 : FBC Connection=B

ddarek avatar Feb 08 '22 19:02 ddarek

I have found that frontends 0-7 have fbc_set_id=0 (/proc/stb/frontend/X/fbc_set_id) frontends 8-15 have fbc_set_id=1

ddarek avatar Feb 08 '22 19:02 ddarek

Yes, you mentioned that frontend 8,9 seem to be connected to 0,1?

Barracuda09 avatar Feb 08 '22 19:02 Barracuda09

Yes, Frontends 0-7 are for FBC-1 Frontends 8-15 are for FBC-2

I managed to work 4 streams on different server app. 4 streams use 0,1,8,9 frontends (because I use in legacy lnb configuration with diseqc). Frontend : 0 is a FBC-0-A 1 is a FBC-1-B 8 is a FBC-8-A 9 is a FBC-9-B

I'm testing it on tvheadend connected to this app and it seems to work stable ( I managed to do this 1h ago ) :) So I'm almost sure that problem is that SATPI don't see that 8 and 9 are FBC-Roots.

ddarek avatar Feb 08 '22 19:02 ddarek

Yes the problem is that SatPI writes to

  • /proc/stb/frontend/8/fbc_connect a 0 that should be 8
  • /proc/stb/frontend/9/fbc_connect a 1 that should be 9

If you do pobably (without changing anything in the webUI:

  • echo 8 > /proc/stb/frontend/8/fbc_connect
  • echo 9 > /proc/stb/frontend/9/fbc_connect

Barracuda09 avatar Feb 08 '22 19:02 Barracuda09

Hmm, it looks that currently on other app only 3 streams are working, if I add 4th then it's getting unstable and last to are disrupting them selves :) With : echo 8 > /proc/stb/frontend/8/fbc_connect echo 9 > /proc/stb/frontend/9/fbc_connect and with echo 0 > /proc/stb/frontend/8/fbc_connect echo 1 > /proc/stb/frontend/9/fbc_connect it works the same without any change.

ddarek avatar Feb 08 '22 19:02 ddarek

It started to be unstable, but after restarting tvheadend it's ok now, so it was a tvheadend problem now (on different server app)

ddarek avatar Feb 08 '22 20:02 ddarek

I will have a look at this, to support the second FBC tuner (NOTES). It's probably the offset that needs to be added to selection A/B of FBC Connection when fbc_set_id = 1 (variable _fbcSetID). https://github.com/Barracuda09/SATPI/blob/6b90b138ec2903254be982793a485dd43d7aa05e/src/input/dvb/delivery/DVBS.cpp#L52-L65

Adapt this to include the offset depending on _fbcSetID https://github.com/Barracuda09/SATPI/blob/6b90b138ec2903254be982793a485dd43d7aa05e/src/input/dvb/delivery/DVBS.cpp#L291-L300

Barracuda09 avatar Feb 08 '22 21:02 Barracuda09

I have implemented the support for the second(Slot B) DVB-S2X FBC tuner

Barracuda09 avatar Feb 09 '22 19:02 Barracuda09

Hi, great!! Change looks promising. I will compile it tomorrow, test it and give you a feedback. Thanks

ddarek avatar Feb 09 '22 19:02 ddarek

I have compiled and run it. Unfortunately fix don't work :(

Frontend 0 = DVB-S(2) FBC-Root (Slot-A) Frontend 1 = DVB-S(2) FBC-Root (Slot-A) Frontend 2 - 15 = DVB-S(2) FBC-Root (Slot-A)

I didn't analyze the change - I'm java dev, I didn't work with c++ since long time. I think slot-b root0 should be the first frontend with fbc_set_id=1.

I attach the log SATPI log file and content of /proc/stb/frontend dir. proc.tar.gz

ddarek avatar Feb 09 '22 20:02 ddarek