TBS6982SE - card or driver problem?

[Wonderingaboutit]

Hi,

I have been having a bit of difficulty scanning known transponders on AMC21 with both scan-s2 and w_scan using a new install of Debian Stretch x86_64, and as V4l modules required a some headers not in Debian, a newly compiled vanilla kernel (4.4.7 originally, now with 4.4.8). The tuner is on loan from a friend and while it seemed to work in windows it did not pick up all the transponders. A MicroHD receiver does work well on the feed reporting signal strength at 90% and good quality on all desired FTA transponders.

Frequencies from Lyngsat (or the MicroHD scanning) did not match with what the card is reporting using a 4.4.7 kernel and the open source drivers. I thought this was potentially was an open source driver issue, so tried with the TBS 160405 drivers on the 4.4.7 kernel with no success.

I then went back to TBS drivers on freshly compiled and installed 4.4.8 kernel again but again with no success. I then removed the TBS modules and installed the open source drivers. Both scan-s2 and w_scan report services on the frequencies which do not correspond to those which should be found. For example this w_scan vdr result is for signals that should have been found at 12180 but are found at 11186.

Code: Select All Code

HD01;PBS:11186:vC56M2O35S1:??:22000:49=27:0;52,53,54:0:0:3:1:1:0
HD02;PBS:11186:vC56M2O35S1:??:22000:65=27:0;68,69,70:0:0:4:1:1:0
HD03;PBS:11186:vC56M2O35S1:??:22000:81=27:0;84,85,86:0:0:5:1:1:0
SD01;PBS:11186:vC56M2O35S1:??:22000:177=27:0;180,181,182,183:0:0:11:1:1:0
SD02 Create;PBS:11186:vC56M2O35S1:??:22000:193=27:0;196,197,198:0:0:12:1:1:0
SD03 VME;PBS:11186:vC56M2O35S1:??:22000:209=27:0;212,213,214:0:0:13:1:1:0
FNX SD08;PBS:11186:vC56M2O35S1:??:22000:241=27:244@3:0:0:15:1:1:0
SD04 World;PBS:11186:vC56M2O35S1:??:22000:225=27:0;228,229,230:0:0:14:1:1:0
Done, scan time: 05:30.190

1. Is it possible that the card is not working as it should? I have read elsewhere that frequencies are a bit off when scanned with different cards, but this seems a bit more than I would expect?

2. The frequency offset does not seem to be a constant, as trying for the other wanted AMC21 transponders with that offset still has not been successful. Is this frequency offset a driver issue?

3. Are the TBS 160405 drivers compatible with a 4.4.7 or 4.4.8 64-bit kernel?

Any suggestions would be appreciated.

Thanks

Wayne

PS With those with multiple kernels getting wrong symbol versions messages in the logs - use

Code: Select All Code

$locate media.ko

and remove or moving the located files as recommended on the Tvlinux site (and/or the whole located media directories). Using “uname -r” does not pick up those non-running kernels and on my set-up those files from non-running kernels also had to be moved to stop the errors.

[Wonderingaboutit]

I received an email from "support" two business days after my posting and my email to them reporting the posting. The red font is from the original message.

Code: Select All Code

hi ,

Installed the driver for tbs card
1.extract  tbs-linux-drivers_*.zip archive:
#unzip  tbs-linux-drivers_*.zip
2.extract linux-tbs-drivers.tar.bz2 archive:
# tar xjvf linux-tbs-drivers.tar.bz2
3. go to driver package directory:
# cd linux-tbs-drivers
4. set platform , use " [color=#BF0000]getconf LONG_BIT[/color] " to get the bit of the system. use "[color=#FF0000]uname -r[/color] " to get the version of the kernel.
4.1  x86 32 bit
     4.1.1  for x86 kernel 2.6.x (x86 32 bit installations of kernel 2.6.x):
                #./v4l/tbs-x86.sh
    4.1.2  for x86 kernel 3.x (x86 32 bit installations of kernel 3.x):
               #./v4l/tbs-x86_r3.sh
4.2  x86 64 bit
                #./v4l/tbs-x86_64.sh
                #uname -r
                3.10.0-327.13.1.el7.x86_64
                #rm -rf /lib/modules/3.10.0-327.13.1.el7.x86_64/kernel/drivers/media
5. [color=#FF0000]if install for DVBC (if install for DVBS/S2 or DVBT/T2 , skip this step, go to step 6)

- for x86 kernel 2.6.x (x86 32 bit installations of kernel 2.6.x):

# ./v4l/tbs-dvbc-x86.sh

- for x86 kernel 3.x (x86 32 bit installations of kernel 3.x):

# ./v4l/tbs-dvbc-x86_r3.sh

- for any x86_64 kernel (x86 64 bit installations of Linux):

# ./v4l/tbs-dvbc-x86_64.sh
[/color]
6. compile driver source.
# make && make install

7. reboot pc
 # shutdown -r now

best regards!

There clearly is a language issue and the response contradicts the instructions/manuals/wikis/linuxtv guides, as I read them, but I did try it anyway....without that script but the compile stops with a "no rule to make" at least on my 64 bit system...
I did not get a followup response to my reply.

My tentative conclusions are:
-v160405 drivers are probably broken, and are based on an old version of the DVB_API;
-the TBS6982SE I have is probably defective; and,
-support appears spotty-something seen on other threads here.

I will be giving up on this card very shortly unless I can get it to work. To save duplication for any others with the same problems, here are my findings:

W_scan with either TBS or Opensource drivers reports that this is a good choice for a satellite card based on the identification of the chips from the drivers. Unfortunately that has not been my experience.

Both w_scan and dvbv-5 scan warn about use of "outdated drivers" and potential errors in calls when I use the TBS drivers. They do not report that with the Opensource drivers.

TBS drivers use from the version.h files:
#define DVB_API_VERSION 5
#define DVB_API_VERSION_MINOR 3

The linux 4.4 kernel and the opensource drivers use:
#define DVB_API_VERSION 5
#define DVB_API_VERSION_MINOR 10

http://www.tbsdtv.com/forum/viewtopic.php?f=69&t=8314 suggests this was a problem in 2013. I found that on a search after I looked at those files on my computer to explore the warnings from w_scan and dvbv-5 scan.

Advice not to upgrade to the TBS6905 Linux Driver v160405:
http://rickcaylor.websitetoolbox.com/post/latest-tbs-linux-driver-package-doesnt-appear-to-work-with-tbs6905-card-8071021
I saw the file names/sizes are exactly the same as the 6982 links so assume they are the same. I did a compile/install of the older v160126 (obtained via a 6982 link a while back if that would have made any difference), but that did not help me.

Given this card didn't fully work in "windows" I looked at that side of this forum: viewtopic.php?f=147&t=9389 "This product is broken", shows there are cards with bad chips out there (board V13).
Pulled the card as it was originally acquired in the same time frame as that posting, it is a V12. Computer's mother board does have the most recent bios and the PCI interrupts are fixed as per linuxTV wiki instructions for TBS cards.

Scan-s2 and w_scan and dvbv5-scan do not report any channels using TBS drivers. Dvbv5-scan does report a lock on one frequency with the TBS drivers (not an expected one and no useable signal). It reports one message ERROR FE_GET_FRONTEND: Remote IO error followed by time outs. Given that some dvbv5.10 calls from older APIs may be misinterpreted I am not sure about that one.

While using the Open source drivers, scans do lock on multiple frequencies showing a stream reporting channel names and data from one of the transponders at a different frequency. I have found 5 different freqency repeats similar to those in my first posting. I did not find streams from the two other transponders of interest. I haven't suceeded in using/seeing any of the streams, which do show up on the scans, but that is not all that unexpected as Lyngsat reports those channel names should be at a different (single) freqency. (During scanning the scanning programs parse the stream's tables showing station names and pids etc, although in the case of dvbv-5 it also reports some out of range values for the pilot freqency.)

It would be good to know if anyone else has the 6982SE cards working with the 160405 drivers and a 4.4.7or.8 kernel with the updated api?

Thanks to all who read this message, however, I expect unless the company comes up with new drivers fairly quickly that work for me I shall be moving on fairly soon.

W


lspci -v

Code: Select All Code

05:00.0 Multimedia controller: Philips Semiconductors SAA7160 (rev 03)
   Subsystem: Device 6982:0002
   Flags: bus master, fast devsel, latency 0, IRQ 50
   Memory at fe800000 (64-bit, non-prefetchable) [size=1M]
   Capabilities: [40] MSI: Enable+ Count=1/32 Maskable- 64bit+
   Capabilities: [50] Express Endpoint, MSI 00
   Capabilities: [74] Power Management version 2
   Capabilities: [80] Vendor Specific Information: Len=50 <?>
   Capabilities: [100] Vendor Specific Information: ID=0000 Rev=0 Len=088 <?>
   Kernel driver in use: SAA716x TBS
   Kernel modules: saa716x_tbs_dvb

Here are extracts from dmesg- I have tried to upload the whole file but apparently the extension ".txt" and "the extension" are not allowed. And its length prevents posting it here.

Code: Select All Code

[    5.466563] TBS version 160405
[    5.508761] [drm] Internal thermal controller without fan control
[    5.509884] [drm] radeon: dpm initialized
[    5.519569] tbs6290fe: module license 'TurboSight Proprietary' taints kernel.
[    5.519574] Disabling lock debugging due to kernel taint

[    6.565814] Registered IR keymap rc-tbs-nec
[    6.565947] input: saa716x IR (TurboSight TBS 6982SE) as /devices/pci0000:00/0000:00:15.3/0000:05:00.0/rc/rc0/input19
[    6.566032] rc0: saa716x IR (TurboSight TBS 6982SE) as /devices/pci0000:00/0000:00:15.3/0000:05:00.0/rc/rc0

[    6.566306] DVB: registering new adapter (SAA716x dvb adapter)

[    7.032465] TurboSight TBS 6982SE Frontend Attaching...

[    7.481869] TurboSight TBS6982SE DVB-S2 card port0 MAC=00:22:ab:f0:76:e8
[    7.481875] DVB: registering adapter 0 frontend 0 (TurboSight TBS 6982SE DVBS/S2 frontend)...
[    7.482019] DVB: registering new adapter (SAA716x dvb adapter)

[    7.945796] TurboSight TBS 6982SE Frontend Attaching...
[    8.016591] TurboSight TBS6982SE DVB-S2 card port1 MAC=00:22:ab:f0:76:e9
[    8.016598] DVB: registering adapter 1 frontend 0 (TurboSight TBS 6982SE DVBS/S2 frontend)...