1 hour ago, ich777 said:

Please post your Diagnostics after you've installed the LibreELEC driver and a reboot.

Done, thanks

tower-diagnostics-20260603-1714.zip

4 hours ago, MihaiS said:

Done, thanks

tower-diagnostics-20260603-1714.zip

With which driver did it work or at least you saw that messages in dmesg?

Must be the LibreELEC one or was it the TBS one?

However your Tuner should show up as card=26 where yours does show as card=93

https://www.linuxtv.org/wiki/index.php/Hauppauge_WinTV-HVR-935_HD

44 minutes ago, ich777 said:

With which driver did it work or at least you saw that messages in dmesg?

Must be the LibreELEC one or was it the TBS one?

However your Tuner should show up as card=26 where yours does show as card=93

https://www.linuxtv.org/wiki/index.php/Hauppauge_WinTV-HVR-935_HD

With LibreELEC was the only one that showed a message, but nothing more than that.
Also it manages to mount /dev/dvb but it is empty.

12 minutes ago, MihaiS said:

With LibreELEC was the only one that showed a message, but nothing more than that.
Also it manages to mount /dev/dvb but it is empty.

Strangely enough I don't see this message in your Diagnostics that you've sent.

I've found a few threads where some users reported that on newer Tuners that they don't work with Linux, not sure if the hardware changed or what the issue here is.

1 hour ago, ich777 said:

Strangely enough I don't see this message in your Diagnostics that you've sent.

I've found a few threads where some users reported that on newer Tuners that they don't work with Linux, not sure if the hardware changed or what the issue here is.

Is there any other log/diag that i can send in order to narrow down the issue ?

8 hours ago, MihaiS said:

Is there any other log/diag that i can send in order to narrow down the issue ?

Can you please try to create a file in:

/boot/config/modprobe.d/em28xx.conf

with the contents:

options em28xx card=92

and reboot.

2 hours ago, ich777 said:

Can you please try to create a file in:

/boot/config/modprobe.d/em28xx.conf

with the contents:

options em28xx card=92

and reboot.

I have done as requested:

root@Tower:~# ls -ltrh /boot/config/modprobe.d/em28xx.conf

-rw------- 1 root root 23 Jun 4 07:41 /boot/config/modprobe.d/em28xx.conf

root@Tower:~# cat /boot/config/modprobe.d/em28xx.conf

options em28xx card=92

root@Tower:~# ls -ltr /dev/dvb/

total 0

root@Tower:~# dmesg | grep -i -E "em28xx|dvb|firmware|2040"

[ 8.025448] rc.S: Verifying boot/bzfirmware checksum ...

tower-diagnostics-20260604-1055.zip

3 minutes ago, MihaiS said:

I have done as requested:

It seems that this is some kind of new hardware revision, if you search for your dmesg output from above you'll find several posts online with similar issues but all with other card ID's I'm not too sure what's going on here.

10 minutes ago, ich777 said:

It seems that this is some kind of new hardware revision, if you search for your dmesg output from above you'll find several posts online with similar issues but all with other card ID's I'm not too sure what's going on here.

Yea I have seen some posts related to this but the bad part is none of them have a solution unfortunately.

I have tried to debug this issue with Gemini and the answer i got so far is the following :

"I tested the card=92 parameter, and while it successfully forced the driver to change its board profile from KWorld to PCTV, it still crashed in the exact same way during the USB transfer buffer pre-allocation.

Here is the log after forcing the device with card=92:

[ 2563.147430] em28xx 1-3:1.0: unknown em28xx chip ID (148)

[ 2563.582550] em28xx 1-3:1.0: unknown eeprom format or eeprom corrupted !

[ 2563.646359] em28xx 1-3:1.0: Identified as PCTV DVB-S2 Stick (461e) (card=92)

[ 2563.654866] em28xx 1-3:1.0: Binding DVB extension

[ 2563.654869] em28xx 1-3:1.0: no endpoint for DVB mode and transfer type 0

[ 2563.654870] em28xx 1-3:1.0: failed to pre-allocate USB transfer buffers for DVB.

It looks like because Chip ID 148 is completely missing from the em28xx source code (specifically in em28xx.h and em28xx-cards.c), the driver doesn't know how to query the physical endpoints for this newer revision of the Empia silicon. It panics and fails to allocate the USB buffers before the card= override can even be utilized.

Would it be possible to patch the kernel source for the LibreELEC package to recognize Chip ID 148? It likely just needs to be mapped to one of the modern Empia chip architectures (like the em28174 or em28178) so it stops failing the initial hardware validation check."

2 hours ago, MihaiS said:

Would it be possible to patch the kernel source for the LibreELEC package to recognize Chip ID 148? It likely just needs to be mapped to one of the modern Empia chip architectures (like the em28174 or em28178) so it stops failing the initial hardware validation check."

Sorry, I'm not a huge fan from what you are proposing since this indeed could work but please also do note that this could go horribly wrong. I would rather see that this gets actually mode upstream since I really don't know if that's the way to go.

41 minutes ago, ich777 said:

Sorry, I'm not a huge fan from what you are proposing since this indeed could work but please also do note that this could go horribly wrong. I would rather see that this gets actually mode upstream since I really don't know if that's the way to go.

Totally understandable, thanks for your info. I guess i will have to wait then, or buy another device that is oficially supported

On 5/30/2026 at 11:02 AM, ich777 said:

Or write the support from TBS.

I'm not 100% sure if this card is supported because I can't find any information which devices are supported in the closed source driver.

So, this was the conversation with the TBS Support:

now that the linux driver has gone closed source, will there ever be support for the TBS6982 card again, with a current non abandoned driver, or am i stuck with an old linux kernel?

TBS6982 have open source driver you can check here

https://github.com/tbsdtv/linux_media/wiki

Yes, but they are abandoned and can't be compiled against the newest kernel, right?

which kernel you want to use ?

Currently I am using 6.12.90, but also newer kernels in the future.

You can try this command 

curl -sSf https://www.tbsdtv.com/download/document/linux/drv-tbs.sh | sh

some guide you can find from here :

https://www.tbsiptv.com/help_center/tbs_driver_installation/linux.html 

This script installs the latest closed source driver, right?

But is the TBS6982 even supported by the closed source driver?

closed driver already discontinued support many years , we are using the open source driver now .

The script is use the latest open source driver 

So... I am now highly confused as to what is the newest driver, and if my card is supported at all. Please help. 🫠

1 hour ago, Greyberry said:

So... I am now highly confused as to what is the newest driver, and if my card is supported at all. Please help. 🫠

The GitHub page says it supports up to Kernel 6.14 (Open Source) and the script that they've linked includes only their closed source drivers, not the open source ones from GitHub.

Download the script and see for yourself, the script downloads a tar archive from their site.

Ok i'll try running the script, but i need the build tools on unraid. How can i obtain them?

V4L drivers building...
./install.sh: line 58: make: command not found
V4L drivers installing...
sudo: make: command not found

Install system utilities to build drivers from the source code:

apt -y install gcc\
    build-essential \
    patchutils \
    libproc-processtable-perl \
    linux-headers-$(uname -r) \

Will probably have to build on another machine/VM and transfer the compiled result

I still do not fully grasp the whole picture of this issue.

• tbsdtv/linux_media has a commit from last week, so the open-source tree isn't really abandoned, is it? (https://github.com/tbsdtv/linux_media).

• TBS support told me (discussion above) the closed-source driver was discontinued years ago, and that the drv-tbs.sh script now installs the latest open-source driver; the opposite of what we'd assumed here, so there may be some confusion about which is which.

• Their install page lists "Auto (kernel >= 6.8)" via the script vs "Manual (kernel <= 6.14)" via git clone, which implies the script's package supports kernels above 6.14.

• I'm on 7.2.7 (kernel 6.12.90), which is under the 6.14 cap, yet TBS-OpenSource stopped working. so the plugin's build may just be on a stale snapshot?!

Would you consider updating the TBS-OpenSource build to the current tree, and/or adding two separate entries (open-source linux_media vs the script package) so older cards can try both? Happy to test on the 6982 and post diagnostics. Thanks!

Edited June 17Jun 17 by Greyberry