Blindscan on linux with tbs cards

[deeptho]

Hello.
I'm having a problem with the drivers of 6909x.
This is the stack trace. Let me know if you need something more.
Thanks!

You managed to exclude all the essential information, such as
-which driver version you were using
-what you were doing that made the problem appear
-if this is a one time error or occurs always
-which user program version you used
-if other programs work with these drivers

Even the provided piece of the kernel log does not contain the line that shows the problem.
Why not send the full kernel log?


Most likely cause is incorrect driver (version) installation, or perhaps some hardware problem.
6909x is the most tested card... Of coruse bugs are also possible.

 

[tasoss]

Hello.
I am using all the latest versions from your repositories and using your branches.
Of course i can attach the full log.
The error occurs during the system's booting.
The adapters are not appearing in /dev/dvb.
I have attached the log here
And now some more strange facts:
1.Right now after after logging in and doing ls -alt /dev/dvb i can only see one adapter. If i ls again after 1 second there are no adapters available.
2.The first time i installed the card , and after compiling etc it was working as expected. After rebooting it stopped working. Then i change the pci used and started working again.Then stopped again. Right now , after 20 reboots it may work one time.
Thanks again and let me know whatever you need.

ps:btw using the official drivers from tbs the adapters appear correctly

 

[deeptho]

According to your log, it is an i2c problem

I2cReadWrite:347 BUG: i2c_transfer returned -5.

Sounds like a hardware problem, but some people have that when running in a virtual machine.

What type of system are you running this on?
And instead of rebooting, also try powering off and rebooting. Sometimes that makes a difference.

Changing i2c related settings in bios can also help.

As an experiment, you can also try changing tbsecp3-cards.c so that line 1694
in the entry "[TBSECP3_BOARD_TBS6909X]"
becomes .i2c_speed =9 (instead of 4)
Then recompile, reinstall, en reboot
This will slow down access to the i2c bus.

 

[tasoss]

Ok thank you.
I will give it a try and let you know.
Thanks again for your work!

ps: It is not a VM

 

[tasoss]

@deeptho it is working as expected after changing to 9 (i even tried 5 and it worked).
1) How does it hurt performance ?
2) How can i fix it ? For example by changing motherboard or something like that ? Is this a tbs hardware problem which may not occur if i purchase another card?
Thanks again!

 

[deeptho]

@deeptho it is working as expected after changing to 9 (i even tried 5 and it worked).
1) How does it hurt performance ?

Yes it does. Leave it at 5 if that works.

2) How can i fix it ? For example by changing motherboard or something like that ? Is this a tbs hardware problem which may not occur if i purchase another card?
Thanks again!

Difficult to know, as you have not even told us what hardware you are using. It is not a tbs problem as they use value 9 for their drivers.

 

[tasoss]

The specs are as following:
CPU:12th Gen Intel(R) Core(TM) i5-12400 (12 cores)
Motherboard ASUSTeK COMPUTER INC. PRIME B660M-K D4
RAM:16GB DDR4 2400
HDD:KINGSTON SNV2S500G ( 500G NV2 M.2 2280 PCIe 4.0 NVMe SSD )

BTW 6903x work(in the same machine) without the need to change the i2c value.

On more question please.
3) While blscan2 finds

S2 12506000 H 1585000 1/4 20 QPSK 0
S2 12533000 H 10830000 3/4 20 32APSK 0
S1 12563000 H 3702000 7/8 35 QPSK
S2 12639000 H 10830000 3/5 20 8PSK
S1 12674000 H 2928000 3/4 35 QPSK
S2 12682000 H 7423000 2/3 20 8PSK
S2 12693000 H 2221000 4/5 20 QPSK
S1 12700000 H 9878000 2/3 35 QPSK
S2 12718000 H 4999000 3/4 20 QPSK
S2 12724000 H 4099000 3/4 20 QPSK
S2 12730000 H 1697000 2/3 20 8PSK 0
S2 12734000 H 3779000 4/5 20 QPSK
S2 12742000 H 999000 2/3 20 QPSK

stid135-blindscan (both 6903x and 6909x) max frequency is

S2 12528174 H 29999133 3/4 AUTO 8PSK

Any ideas ? ( more info here )
Thanks again!

 

[deeptho]

The specs are as following:
CPU:12th Gen Intel(R) Core(TM) i5-12400 (12 cores)
Motherboard ASUSTeK COMPUTER INC. PRIME B660M-K D4
RAM:16GB DDR4 2400
HDD:KINGSTON SNV2S500G ( 500G NV2 M.2 2280 PCIe 4.0 NVMe SSD )

BTW 6903x work(in the same machine) without the need to change the i2c value.

On more question please.
3) While blscan2 finds


stid135-blindscan (both 6903x and 6909x) max frequency is

Any ideas ? ( more info here )
Thanks again!

Have you tried putting the 6909x in a different pci-e slot?
And is this connected directly or through a riser cable?

Note that the two cards are almost the same internally

 

[tasoss]

I have tried on different pci slot with the same results.
It is connected directly.
I will try it with another hardware and let you know.
Thanks.

 

[tasoss]

I would like to ask one final time because maybe you haven't seen my question.

While blscan2 finds

S2 12506000 H 1585000 1/4 20 QPSK 0
S2 12533000 H 10830000 3/4 20 32APSK 0
S1 12563000 H 3702000 7/8 35 QPSK
S2 12639000 H 10830000 3/5 20 8PSK
S1 12674000 H 2928000 3/4 35 QPSK
S2 12682000 H 7423000 2/3 20 8PSK
S2 12693000 H 2221000 4/5 20 QPSK
S1 12700000 H 9878000 2/3 35 QPSK
S2 12718000 H 4999000 3/4 20 QPSK
S2 12724000 H 4099000 3/4 20 QPSK
S2 12730000 H 1697000 2/3 20 8PSK 0
S2 12734000 H 3779000 4/5 20 QPSK
S2 12742000 H 999000 2/3 20 QPSK

stid135-blindscan (both 6903x and 6909x) max frequency is

S2 12528174 H 29999133 3/4 AUTO 8PSK

Any ideas ?
Thanks again!

 

[deeptho]

I would like to ask one final time because maybe you haven't seen my question.

While blscan2 finds


stid135-blindscan (both 6903x and 6909x) max frequency is


Any ideas ?
Thanks again!

And the question is?

 

[tasoss]

I don't get some transponders as you can see.
Why does it happen ?

 

[deeptho]

I don't get some transponders as you can see.
Why does it happen ?

I don't know, but they look like having very large symbol rates and I may have excluded those, because there are
almost none that can be received here, so there are not many test cases. and what is the point if the stream cannot be decoded?
These streams are probably not for broadcast.

Also, these high symbol rates require combining multiple demods to work together.
It is still in the code and probably still works (tuning and maybe even blind tuning),
but does not play well with the concept of one demod per adapter. So it probably only works if you do not
tune anything else in parallel.

Can you tune them with the drivers or not (blind tune, or after inserting all parameters)?
And can they be tuned (after inserting all parameters) with the official tbs drivers?

Are you sure you can tune them under windows? If the signal is too weak for tuning, neumo does not consider them.

It also seems that blscan2 cannot detect the
S2 12528174 H 29999133 3/4 AUTO 8PSK?

 

[satesco]

...

It also seems that blscan2 cannot detect the
S2 12528174 H 29999133 3/4 AUTO 8PSK?

The parameters are probably without the required punctuation (?).

I think they would be: mux 12528.175 H, symbol rate: 29999.133

 

[deeptho]

The parameters are probably without the required punctuation (?).

I think they would be: mux 12528.175 H, symbol rate: 29999.133

Then they are not high symbolrates. Maybe he is just scanning another satellite in both cases?

 

[tasoss]

I don't know, but they look like having very large symbol rates and I may have excluded those, because there are
almost none that can be received here, so there are not many test cases. and what is the point if the stream cannot be decoded?
These streams are probably not for broadcast.

Some of the transponders contain valid streams which i can tune using a tbs 6908(on linux) and 6903 on windows(blind scanning).

And can they be tuned (after inserting all parameters) with the official tbs drivers?

Using the original drivers 6903x tunes correctly on 12693000(a frequency that blind scanning is not finding) but 6909x not.
neumo drivers don't tune on both cards.

Also, these high symbol rates require combining multiple demods to work together.
It is still in the code and probably still works (tuning and maybe even blind tuning),
but does not play well with the concept of one demod per adapter. So it probably only works if you do not
tune anything else in parallel.

Although you are talking technically(and i don't understand everything as i don't know how it works) i just want to say that i get the same results
with 6903x (non parallel).

Then they are not high symbolrates. Maybe he is just scanning another satellite in both cases?

It is the same satellite. It's output from blscan2 in vdr format ( Syntax of channels.conf - VDR Wiki )
It mentions

The values here will be multiplied by 1000 until it's bigger than 1000000

and they treat frequency as integer ( i suppose that's why they are multiplying)

Another problem that i have found is 'max-symbol-rate' parameter doesn't seem to work . In 6909x it's not used in the code and in 6903x although it is passed as parameter to other functions it doesn't seem to work(returns greater symbol rates than the one i pass to the parameter). Am i missing something?

Thanks again.

 

[deeptho]

Some of the transponders contain valid streams which i can tune using a tbs 6908(on linux) and 6903 on windows(blind scanning).


Using the original drivers 6903x tunes correctly on 12693000(a frequency that blind scanning is not finding) but 6909x not.
neumo drivers don't tune on both cards.

It is very unlikely that 6903x works and 6909x does not. So it is probably a configuration problem: using the wrong cable,
or wrong adapter or some diseqc problem.

Although you are talking technically(and i don't understand everything as i don't know how it works) i just want to say that i get the same results
with 6903x (non parallel).


It is the same satellite. It's output from blscan2 in vdr format ( Syntax of channels.conf - VDR Wiki )
It mentions

and they treat frequency as integer ( i suppose that's why they are multiplying)

Another problem that i have found is 'max-symbol-rate' parameter doesn't seem to work . In 6909x it's not used in the code and in 6903x although it is passed as parameter to other functions it doesn't seem to work(returns greater symbol rates than the one i pass to the parameter). Am i missing something?

Thanks again.

It is not needed and should be removed as an option. I can only guess what you are doing, but probably you are using one of the blindscan command line tools. These should work but are not much tested. Why not use neumodvb instead?

In any case you have several problems, even without the neumo drivers, so probably something is wrong with your setup.

Also, reporting "it does not work" without any further information provides no clues at all. So it is not really possible to help without some detialed information (log files and such)

 

[tasoss]

It is not needed and should be removed as an option. I can only guess what you are doing, but probably you are using one of the blindscan command line tools. These should work but are not much tested. Why not use neumodvb instead?

Indeed i am using stid135-blindscan and neumo-blindscan command line tools.
If you are using the max-symbol-rate option and it's working as expected then i am doing something wrong.
But please correct the documentation on stid135-blindscan about the parameter.
I will try to send a patch (pull request) for the stid135-blindscan.cc

It is very unlikely that 6903x works and 6909x does not. So it is probably a configuration problem: using the wrong cable,
or wrong adapter or some diseqc problem.

In any case you have several problems, even without the neumo drivers, so probably something is wrong with your setup.

I can give you access and see by yourself. Tell me how i can contact you.

Why not use neumodvb instead?

I want to have a solution as lightweight as possible.

Also, reporting "it does not work" without any further information provides no clues at all. So it is not really possible to help without some detialed information (log files and such)

I am available to provide whatever you want. Just ask me.
Thanks again.

 

[tasoss]

It is very unlikely that 6903x works and 6909x does not. So it is probably a configuration problem: using the wrong cable,
or wrong adapter or some diseqc problem.

I have written that using the tbs drivers 6903x can tune to the specific freq but 6909x does not.
But why neumo drivers don't work with 6903x at least ?

 

[vasi54]

Hi,

I have TBS6903X v1.2.
I use the neumo-blindscan program under Fedora40.
An error occurs when changing in H polarity, the band from 10700 - 11700 to 11700 - 12750 MHZ. This error is : " problem Setting voltage ".
Can you help me to understand what cause this?
Is software or hardware problem?

thanks,