Urgent: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

Urgent: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Eyal Lebedinsky]

I have just upgraded to f24. I am now using the standard dvb_usb_rtl28xxu fe
which logs messages suggesting all is well (I get the /dev/dvb/adapter? etc.)
but I get no channels tuned when I run mythfrontend or scandvb.

Is anyone using this combination?
Is this the correct way to use this tuner?

BTW:

Until f22 I was using the out of kernel driver from
	https://github.com/jaredquinn/DVB-Realtek-RTL2832U.git
but I now get a compile error.

TIA

-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)

f26: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Eyal Lebedinsky]

On 18/09/17 14:26, Eyal Lebedinsky wrote:
> I have just upgraded to f24. I am now using the standard dvb_usb_rtl28xxu fe

I have upgraded to f26 and this driver still fails to tune the "Leadtek Winfast DTV2000 DS PLUS TV".

> which logs messages suggesting all is well (I get the /dev/dvb/adapter? etc.)
> but I get no channels tuned when I run mythfrontend or scandvb.
> 
> Is anyone using this combination?
> Is this the correct way to use this tuner?

Is this the wrong list? If so then please suggest a more suitable one.

> BTW:
> 
> Until f22 I was using the out of kernel driver from
>      https://github.com/jaredquinn/DVB-Realtek-RTL2832U.git
> but I now get a compile error.
> 
> TIA

-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)

Re: f26: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Vincent McIntyre]

On Sat, Sep 23, 2017 at 10:48:34PM +1000, Eyal Lebedinsky wrote:
> On 18/09/17 14:26, Eyal Lebedinsky wrote:
> >I have just upgraded to f24. I am now using the standard dvb_usb_rtl28xxu fe
> 
> I have upgraded to f26 and this driver still fails to tune the "Leadtek Winfast DTV2000 DS PLUS TV".
> 
> >which logs messages suggesting all is well (I get the /dev/dvb/adapter? etc.)
> >but I get no channels tuned when I run mythfrontend or scandvb.
> >
> >Is anyone using this combination?
> >Is this the correct way to use this tuner?
> 
> Is this the wrong list? If so then please suggest a more suitable one.

It's the right list. The problem is nobody seems to care.
I have one of these too, I was able to get it tune at one time
but there were some problems that I never ended up running down.

I was planning to dig it out and have a play with it again.

Just to confirm - you're building the media_build git tree on f26
and those drivers are the ones that are not working, yes?
If not, you need to try that to get any help here.
Have a look at https://git.linuxtv.org/media_build.git/about/
and let me know if you need further help with that.

It may be possible to get the driver into debug mode and get more
information logged. I'm not sure this will work but give it a go.

First set up the dynamic debug filesystem (may already be there)
# cat >> /etc/fstab
debugfs /sys/kernel/debug debugfs defaults 0 0
^D
# mount -av

Turn on debug printing for the modules of interest
# echo 'module rtl2832 +p' > /sys/kernel/debug/dynamic_debug/control
# echo 'module dvb_usb_rtl28xxu +p' > /sys/kernel/debug/dynamic_debug/control


Vince

Re: f26: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Eyal Lebedinsky]

Thanks for writing Vince,

On 24/09/17 19:04, Vincent McIntyre wrote:
> On Sat, Sep 23, 2017 at 10:48:34PM +1000, Eyal Lebedinsky wrote:
>> On 18/09/17 14:26, Eyal Lebedinsky wrote:
>>> I have just upgraded to f24. I am now using the standard dvb_usb_rtl28xxu fe
>>
>> I have upgraded to f26 and this driver still fails to tune the "Leadtek Winfast DTV2000 DS PLUS TV".
>>
>>> which logs messages suggesting all is well (I get the /dev/dvb/adapter? etc.)
>>> but I get no channels tuned when I run mythfrontend or scandvb.
>>>
>>> Is anyone using this combination?
>>> Is this the correct way to use this tuner?
>>
>> Is this the wrong list? If so then please suggest a more suitable one.
> 
> It's the right list. The problem is nobody seems to care.

This is really sad.
Seeing so many patches on this list, I hoped there is another list more suitable
for issues that users encounter.

> I have one of these too, I was able to get it tune at one time
> but there were some problems that I never ended up running down.
> 
> I was planning to dig it out and have a play with it again.

Thanks, this may help. I also hoped to get a response from people who have this card
working on f26. Maybe it will work for you (fingers crossed).

> Just to confirm - you're building the media_build git tree on f26
> and those drivers are the ones that are not working, yes?

No, but I did try this before (on f24).

However I did build the media_build git tree two days ago (on f26) but did not
install it yet. Just to confirm, when testing on f24, after an install I find
duplicated drivers:

/lib/modules/4.11.12-100.fc24.x86_64/kernel/drivers/media/usb/dvb-usb-v2/dvb-usb-rtl28xxu.ko.xz
/lib/modules/4.11.12-100.fc24.x86_64/kernel/drivers/media/usb/dvb-usb-v2/dvb-usb-rtl28xxu.ko

The second one is from my build. I assume it has precedence over the original?

> If not, you need to try that to get any help here.

Sure, I am comfortable building from git/source.

> Have a look at https://git.linuxtv.org/media_build.git/about/
> and let me know if you need further help with that.

No worries. What I did before was simple:

	git clone git://linuxtv.org/media_build.git
	cd media_build
	make distclean
	./build
	make install

> It may be possible to get the driver into debug mode and get more
> information logged. I'm not sure this will work but give it a go.
> 
> First set up the dynamic debug filesystem (may already be there)
> # cat >> /etc/fstab
> debugfs /sys/kernel/debug debugfs defaults 0 0
> ^D
> # mount -av
> 
> Turn on debug printing for the modules of interest
> # echo 'module rtl2832 +p' > /sys/kernel/debug/dynamic_debug/control
> # echo 'module dvb_usb_rtl28xxu +p' > /sys/kernel/debug/dynamic_debug/control

This evening there are a few recordings scheduled that I do not want to risk (I am now
using a few USB tuners I had laying around) but I will try this in the morning and
report the results to the list.

> Vince

cheers

-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)

Re: f26: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Eyal Lebedinsky]

[-- Attachment #1: Type: text/plain, Size: 7745 bytes --]

On 24/09/17 19:04, Vincent McIntyre wrote:
> On Sat, Sep 23, 2017 at 10:48:34PM +1000, Eyal Lebedinsky wrote:
>> On 18/09/17 14:26, Eyal Lebedinsky wrote:
>>> I have just upgraded to f24. I am now using the standard dvb_usb_rtl28xxu fe
>>
>> I have upgraded to f26 and this driver still fails to tune the "Leadtek Winfast DTV2000 DS PLUS TV".
>>
>>> which logs messages suggesting all is well (I get the /dev/dvb/adapter? etc.)
>>> but I get no channels tuned when I run mythfrontend or scandvb.
>>>
>>> Is anyone using this combination?
>>> Is this the correct way to use this tuner?
>>
>> Is this the wrong list? If so then please suggest a more suitable one.
> 
> It's the right list. The problem is nobody seems to care.
> I have one of these too, I was able to get it tune at one time
> but there were some problems that I never ended up running down.
> 
> I was planning to dig it out and have a play with it again.
> 
> Just to confirm - you're building the media_build git tree on f26

I now installed from media_build with no improvement. These are the boot messages:
[included are the two USB tuners: 'Realtek RTL2832U reference design' 'AVerMedia Twinstar (A825)']

Sep 25 09:12:04 e7 kernel: dvb_usb_af9035: unknown parameter 'remote' ignored
Sep 25 09:12:04 e7 kernel: usb 3-1: dvb_usb_v2: found a 'Leadtek WinFast DTV2000DS Plus' in warm state
Sep 25 09:12:04 e7 kernel: dvb_usb_af9035 2-1.5:1.0: prechip_version=00 chip_version=03 chip_type=3802
Sep 25 09:12:04 e7 kernel: usb 2-1.5: dvb_usb_v2: found a 'AVerMedia Twinstar (A825)' in cold state
Sep 25 09:12:04 e7 kernel: usb 2-1.5: dvb_usb_v2: downloading firmware from file 'dvb-usb-af9035-02.fw'
Sep 25 09:12:04 e7 kernel: usb 3-1: dvb_usb_v2: will pass the complete MPEG2 transport stream to the software demuxer
Sep 25 09:12:04 e7 kernel: dvbdev: DVB: registering new adapter (Leadtek WinFast DTV2000DS Plus)
Sep 25 09:12:04 e7 kernel: usb 3-1: DVB: registering adapter 0 frontend 0 (Realtek RTL2832 (DVB-T))...
Sep 25 09:12:05 e7 kernel: rc rc0: lirc_dev: driver ir-lirc-codec (dvb_usb_rtl28xxu) registered at minor = 0
Sep 25 09:12:05 e7 kernel: usb 3-1: dvb_usb_v2: schedule remote query interval to 200 msecs
Sep 25 09:12:05 e7 kernel: usb 3-1: dvb_usb_v2: 'Leadtek WinFast DTV2000DS Plus' successfully initialized and connected
Sep 25 09:12:05 e7 kernel: usb 3-2: dvb_usb_v2: found a 'Leadtek WinFast DTV2000DS Plus' in warm state
Sep 25 09:12:05 e7 kernel: usb 3-2: dvb_usb_v2: will pass the complete MPEG2 transport stream to the software demuxer
Sep 25 09:12:05 e7 kernel: dvbdev: DVB: registering new adapter (Leadtek WinFast DTV2000DS Plus)
Sep 25 09:12:05 e7 kernel: usb 3-2: DVB: registering adapter 1 frontend 0 (Realtek RTL2832 (DVB-T))...
Sep 25 09:12:05 e7 kernel: rc rc1: lirc_dev: driver ir-lirc-codec (dvb_usb_rtl28xxu) registered at minor = 1
Sep 25 09:12:05 e7 kernel: usb 3-2: dvb_usb_v2: schedule remote query interval to 200 msecs
Sep 25 09:12:05 e7 kernel: usb 3-2: dvb_usb_v2: 'Leadtek WinFast DTV2000DS Plus' successfully initialized and connected
Sep 25 09:12:05 e7 kernel: usb 4-1: dvb_usb_v2: found a 'Leadtek WinFast DTV2000DS Plus' in warm state
Sep 25 09:12:05 e7 kernel: usb 4-1: dvb_usb_v2: will pass the complete MPEG2 transport stream to the software demuxer
Sep 25 09:12:05 e7 kernel: dvbdev: DVB: registering new adapter (Leadtek WinFast DTV2000DS Plus)
Sep 25 09:12:05 e7 kernel: usb 4-1: DVB: registering adapter 2 frontend 0 (Realtek RTL2832 (DVB-T))...
Sep 25 09:12:05 e7 kernel: rc rc2: lirc_dev: driver ir-lirc-codec (dvb_usb_rtl28xxu) registered at minor = 2
Sep 25 09:12:05 e7 kernel: usb 4-1: dvb_usb_v2: schedule remote query interval to 200 msecs
Sep 25 09:12:05 e7 kernel: usb 4-1: dvb_usb_v2: 'Leadtek WinFast DTV2000DS Plus' successfully initialized and connected
Sep 25 09:12:05 e7 kernel: usb 4-2: dvb_usb_v2: found a 'Leadtek WinFast DTV2000DS Plus' in warm state
Sep 25 09:12:05 e7 kernel: usb 4-2: dvb_usb_v2: will pass the complete MPEG2 transport stream to the software demuxer
Sep 25 09:12:05 e7 kernel: dvbdev: DVB: registering new adapter (Leadtek WinFast DTV2000DS Plus)
Sep 25 09:12:05 e7 kernel: usb 4-2: DVB: registering adapter 3 frontend 0 (Realtek RTL2832 (DVB-T))...
Sep 25 09:12:05 e7 kernel: rc rc3: lirc_dev: driver ir-lirc-codec (dvb_usb_rtl28xxu) registered at minor = 3
Sep 25 09:12:05 e7 kernel: usb 4-2: dvb_usb_v2: schedule remote query interval to 200 msecs
Sep 25 09:12:05 e7 kernel: usb 4-2: dvb_usb_v2: 'Leadtek WinFast DTV2000DS Plus' successfully initialized and connected
Sep 25 09:12:05 e7 kernel: usb 1-1.6.4: dvb_usb_v2: found a 'Realtek RTL2832U reference design' in warm state
Sep 25 09:12:05 e7 kernel: usb 1-1.6.4: dvb_usb_v2: will pass the complete MPEG2 transport stream to the software demuxer
Sep 25 09:12:05 e7 kernel: dvbdev: DVB: registering new adapter (Realtek RTL2832U reference design)
Sep 25 09:12:05 e7 kernel: usb 1-1.6.4: DVB: registering adapter 4 frontend 0 (Realtek RTL2832 (DVB-T))...
Sep 25 09:12:05 e7 kernel: rc rc4: lirc_dev: driver ir-lirc-codec (dvb_usb_rtl28xxu) registered at minor = 4
Sep 25 09:12:05 e7 kernel: usb 1-1.6.4: dvb_usb_v2: schedule remote query interval to 200 msecs
Sep 25 09:12:05 e7 kernel: usb 1-1.6.4: dvb_usb_v2: 'Realtek RTL2832U reference design' successfully initialized and connected
Sep 25 09:12:05 e7 kernel: usbcore: registered new interface driver dvb_usb_rtl28xxu
Sep 25 09:12:07 e7 kernel: dvb_usb_af9035 2-1.5:1.0: firmware version=12.13.15.0
Sep 25 09:12:07 e7 kernel: usb 2-1.5: dvb_usb_v2: found a 'AVerMedia Twinstar (A825)' in warm state
Sep 25 09:12:07 e7 kernel: dvb_usb_af9035 2-1.5:1.0: Device may have issues with I2C read operations. Enabling fix.
Sep 25 09:12:07 e7 kernel: usb 2-1.5: dvb_usb_v2: will pass the complete MPEG2 transport stream to the software demuxer
Sep 25 09:12:07 e7 kernel: dvbdev: DVB: registering new adapter (AVerMedia Twinstar (A825))
Sep 25 09:12:07 e7 kernel: usb 2-1.5: DVB: registering adapter 5 frontend 0 (Afatech AF9033 (DVB-T))...
Sep 25 09:12:07 e7 kernel: usb 2-1.5: dvb_usb_v2: will pass the complete MPEG2 transport stream to the software demuxer
Sep 25 09:12:07 e7 kernel: dvbdev: DVB: registering new adapter (AVerMedia Twinstar (A825))
Sep 25 09:12:07 e7 kernel: usb 2-1.5: DVB: registering adapter 6 frontend 0 (Afatech AF9033 (DVB-T))...
Sep 25 09:12:07 e7 kernel: usb 2-1.5: dvb_usb_v2: schedule remote query interval to 500 msecs
Sep 25 09:12:07 e7 kernel: usb 2-1.5: dvb_usb_v2: 'AVerMedia Twinstar (A825)' successfully initialized and connected
Sep 25 09:12:07 e7 kernel: usbcore: registered new interface driver dvb_usb_af9035

> and those drivers are the ones that are not working, yes?

Yes.

> If not, you need to try that to get any help here.
> Have a look at https://git.linuxtv.org/media_build.git/about/
> and let me know if you need further help with that.
> 
> It may be possible to get the driver into debug mode and get more
> information logged. I'm not sure this will work but give it a go.
> 
> First set up the dynamic debug filesystem (may already be there)
> # cat >> /etc/fstab
> debugfs /sys/kernel/debug debugfs defaults 0 0
> ^D
> # mount -av

Already had debugfs.

> Turn on debug printing for the modules of interest
> # echo 'module rtl2832 +p' > /sys/kernel/debug/dynamic_debug/control
> # echo 'module dvb_usb_rtl28xxu +p' > /sys/kernel/debug/dynamic_debug/control

Have done this. Attached are the messages from a (failed) scandvb that fails for all multiplexes.

The messages at the end continued at a high rate after the test finished (until I disabled debug
with '-p') and there was no user of the tuners. Maybe IR RC is active?

> Vince

cheers

-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)

-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)

[-- Attachment #2: debug.dmesg.xz --]
[-- Type: application/x-xz, Size: 12700 bytes --]

Re: f26: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Vincent McIntyre]

On Mon, Sep 25, 2017 at 10:16:23AM +1000, Eyal Lebedinsky wrote:

> >Turn on debug printing for the modules of interest
> ># echo 'module rtl2832 +p' > /sys/kernel/debug/dynamic_debug/control
> ># echo 'module dvb_usb_rtl28xxu +p' > /sys/kernel/debug/dynamic_debug/control
> 
> Have done this. Attached are the messages from a (failed) scandvb that fails for all multiplexes.
> 
> The messages at the end continued at a high rate after the test finished (until I disabled debug
> with '-p') and there was no user of the tuners. Maybe IR RC is active?
 
This looks like progress, now we can see more of the rather odd behaviour.

The tuned frequency does not progress monotonically,
but wanders up and down the band.
 Frequency      Delta
 219166666
 184833334  -34333332
 191833334    7000000
 191166666    -666668
 177833334  -13333332
 184166666    6333332
 177166666   -7000000
 191500000   14333334
 184500000   -7000000
 191833334    7333334
 191166666    -666668
 177500000  -13666666
 219500000   42000000
 184833334  -34666666
 191500000    6666666
 177833334  -13666666
 184166666    6333332
 191833334    7666668
 191166666    -666668
 177166666  -14000000
 191500000   14333334
 184500000   -7000000
 177500000   -7000000
 177833334     333334
 177166666    -666668

The bandwidth and inversion type seem to be set correctly, at least.

Also:
 - every instance of if_frequency=0 pset_iffreq=00000000
   shows the same numbers - zero. Surely that can't be right.
 - there seems to be no correlation at all between the AGC level
   (automatic gain control, I assume) and the 'cnr raw' which I'm
   guessing is some measure of the signal level.

I don't know where to start with decoding the rtl28xxu_ctrl_msg
I'm afraid. It's quite possible the remote control is active.
You can run
   ir-keytable -v
to show which remotes the system knows about.

We might get more clues to rub together from looking at where
scandvb falls over, if you run it under 'strace'
  strace -t -s 2048 -o ./scandvb.strace scandvb <arguments here>
This will show the system calls being made. The -t will add timestamps
to the output that could be correlated with the dmesg output.

I'm curious - did you try scandvb with one of your other tuners?

I'm actually not familiar with scandvb, I can't find a program
with that name in the ubuntu repositories.
What I've used before is dvbv5-scan, which is part of:
 git://linuxtv.org/v4l-utils.git

That repo also includes the v4l2-compliance tool,
which might be useful here. Something like:
 ./v4l2-compliance -d /dev/dvb/adapter3 -a -T -v
What I am not sure about is whether the tuner needs to be
'zapped' to a nice strong TV channel before you can use this.

Anyway, hope some of these ideas are helpful.
Cheers
Vince

Re: f26: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Eyal Lebedinsky]

Thanks Vincent,

On 25/09/17 23:24, Vincent McIntyre wrote:
> On Mon, Sep 25, 2017 at 10:16:23AM +1000, Eyal Lebedinsky wrote:
> 
>>> Turn on debug printing for the modules of interest
>>> # echo 'module rtl2832 +p' > /sys/kernel/debug/dynamic_debug/control
>>> # echo 'module dvb_usb_rtl28xxu +p' > /sys/kernel/debug/dynamic_debug/control
>>
>> Have done this. Attached are the messages from a (failed) scandvb that fails for all multiplexes.
>>
>> The messages at the end continued at a high rate after the test finished (until I disabled debug
>> with '-p') and there was no user of the tuners. Maybe IR RC is active?
>   
> This looks like progress, now we can see more of the rather odd behaviour.
> 
> The tuned frequency does not progress monotonically,
> but wanders up and down the band.
>   Frequency      Delta
>   219166666
>   184833334  -34333332
>   191833334    7000000
>   191166666    -666668
>   177833334  -13333332
>   184166666    6333332
>   177166666   -7000000
>   191500000   14333334
>   184500000   -7000000
>   191833334    7333334
>   191166666    -666668
>   177500000  -13666666
>   219500000   42000000
>   184833334  -34666666
>   191500000    6666666
>   177833334  -13666666
>   184166666    6333332
>   191833334    7666668
>   191166666    -666668
>   177166666  -14000000
>   191500000   14333334
>   184500000   -7000000
>   177500000   -7000000
>   177833334     333334
>   177166666    -666668
> 
> The bandwidth and inversion type seem to be set correctly, at least.
> 
> Also:
>   - every instance of if_frequency=0 pset_iffreq=00000000
>     shows the same numbers - zero. Surely that can't be right.
>   - there seems to be no correlation at all between the AGC level
>     (automatic gain control, I assume) and the 'cnr raw' which I'm
>     guessing is some measure of the signal level.
> 
> I don't know where to start with decoding the rtl28xxu_ctrl_msg
> I'm afraid. It's quite possible the remote control is active.
> You can run
>     ir-keytable -v
> to show which remotes the system knows about.
> 
> We might get more clues to rub together from looking at where
> scandvb falls over, if you run it under 'strace'
>    strace -t -s 2048 -o ./scandvb.strace scandvb <arguments here>
> This will show the system calls being made. The -t will add timestamps
> to the output that could be correlated with the dmesg output.
> 
> I'm curious - did you try scandvb with one of your other tuners?

The other 3 tuner on the "Leadtek Winfast DTV2000 DS PLUS TV" are failing in the same way.

> I'm actually not familiar with scandvb, I can't find a program
> with that name in the ubuntu repositories.
> What I've used before is dvbv5-scan, which is part of:
>   git://linuxtv.org/v4l-utils.git
> 
> That repo also includes the v4l2-compliance tool,
> which might be useful here. Something like:
>   ./v4l2-compliance -d /dev/dvb/adapter3 -a -T -v
> What I am not sure about is whether the tuner needs to be
> 'zapped' to a nice strong TV channel before you can use this.

I will look at these tools.

> Anyway, hope some of these ideas are helpful.

Another data point that may help:

While playing with the tuners, I tried to change the signal strength
(there is an active 4-way splitter feeding the cards) and it made no
difference.

However, at one point I powered off the splitter and the cards started
tuning. However, the USB tuners (now attached as a temporary solution)
had difficulty tuning (naturally). I think that I need the amplifier
but this driver cannot handle the strong signal.

So this made me think: is it possible that something in the dvb_usb_rtl28xxu
setup for the DTV2000 sets some internal gain that makes the  signal too
strong for the tuner? I did not see an option to turn off an LNA or such.

I know that the off kernel driver I was using successfully earlier
	https://github.com/jaredquinn/DVB-Realtek-RTL2832U
had no such issue. It does not build on recent kernels. Looking at
the source I see that at least usb_control_msg() changed.

BTW, If I get the "Hauppauge Nova-TD Stick" working then I can give up
on the DTV2000, they are PCI and will not suit the next mobo upgrade
that is long overdue. I started another thread for the Hauppauge a
few days ago (did not get a response yet).

> Cheers
> Vince

cheers

-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)

Re: f26: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Eyal Lebedinsky]

On 25/09/17 23:24, Vincent McIntyre wrote:
> On Mon, Sep 25, 2017 at 10:16:23AM +1000, Eyal Lebedinsky wrote:
> 
>>> Turn on debug printing for the modules of interest
>>> # echo 'module rtl2832 +p' > /sys/kernel/debug/dynamic_debug/control
>>> # echo 'module dvb_usb_rtl28xxu +p' > /sys/kernel/debug/dynamic_debug/control
>>
>> Have done this. Attached are the messages from a (failed) scandvb that fails for all multiplexes.
>>
>> The messages at the end continued at a high rate after the test finished (until I disabled debug
>> with '-p') and there was no user of the tuners. Maybe IR RC is active?
>   
> This looks like progress, now we can see more of the rather odd behaviour.
> 
> The tuned frequency does not progress monotonically,
> but wanders up and down the band.
>   Frequency      Delta
>   219166666
>   184833334  -34333332
>   191833334    7000000
>   191166666    -666668
>   177833334  -13333332
>   184166666    6333332
>   177166666   -7000000
>   191500000   14333334
>   184500000   -7000000
>   191833334    7333334
>   191166666    -666668
>   177500000  -13666666
>   219500000   42000000
>   184833334  -34666666
>   191500000    6666666
>   177833334  -13666666
>   184166666    6333332
>   191833334    7666668
>   191166666    -666668
>   177166666  -14000000
>   191500000   14333334
>   184500000   -7000000
>   177500000   -7000000
>   177833334     333334
>   177166666    -666668
> 
> The bandwidth and inversion type seem to be set correctly, at least.
> 
> Also:
>   - every instance of if_frequency=0 pset_iffreq=00000000
>     shows the same numbers - zero. Surely that can't be right.
>   - there seems to be no correlation at all between the AGC level
>     (automatic gain control, I assume) and the 'cnr raw' which I'm
>     guessing is some measure of the signal level.
> 
> I don't know where to start with decoding the rtl28xxu_ctrl_msg
> I'm afraid. It's quite possible the remote control is active.
> You can run
>     ir-keytable -v
> to show which remotes the system knows about.
> 
> We might get more clues to rub together from looking at where
> scandvb falls over, if you run it under 'strace'
>    strace -t -s 2048 -o ./scandvb.strace scandvb <arguments here>
> This will show the system calls being made. The -t will add timestamps
> to the output that could be correlated with the dmesg output.
> 
> I'm curious - did you try scandvb with one of your other tuners?
> 
> I'm actually not familiar with scandvb, I can't find a program
> with that name in the ubuntu repositories.
> What I've used before is dvbv5-scan, which is part of:
>   git://linuxtv.org/v4l-utils.git

Grabbed the repo and built it. After some tries (I did not find enough doco)
I ran:

$ cat scan.in
T 177500000 7MHz 3/4    3/4    QAM64 8k                1/16           NONE
T 184500000 7MHz 2/3    1/2    QAM64 8k                1/8            NONE
T 191500000 7MHz 3/4    2/3    QAM64 8k                1/16           NONE
T 219500000 7MHz 3/4    3/4    QAM64 8k                1/16           NONE
T 226500000 7MHz 3/4    1/2    QAM64 8k                1/16           NONE

# the above is the usual frequencies table that I usually use successfully.

$ utils/dvb/dvbv5-scan -I CHANNEL --lna=0 scan.in

Lock   (0x1f) Signal= 93.33% C/N= 16.62dB postBER= 17.5x10^-3
ERROR    dvb_read_sections: no data read on section filter
ERROR    error while waiting for PAT table
Scanning frequency #2 184500000
Lock   (0x1f) Signal= 93.73% C/N= 20.09dB postBER= 65.5x10^-3
ERROR    dvb_read_sections: no data read on section filter
ERROR    error while waiting for PAT table
Scanning frequency #3 191500000
        (0x00)
Scanning frequency #4 219500000
Lock   (0x1f) Signal= 96.86% C/N= 20.29dB postBER= 65.5x10^-3
ERROR    dvb_read_sections: no data read on section filter
ERROR    error while waiting for PAT table
Scanning frequency #5 226500000
Lock   (0x1f) Signal= 93.33% C/N= 19.66dB postBER= 19.6x10^-3
ERROR    dvb_read_sections: no data read on section filter
ERROR    error while waiting for PAT table

I also tried lna=0 and -1, same output.

> That repo also includes the v4l2-compliance tool,
> which might be useful here. Something like:
>   ./v4l2-compliance -d /dev/dvb/adapter3 -a -T -v

Did not manage to run this:

$ utils/v4l2-compliance/v4l2-compliance -d /dev/dvb/adapter0 -a -T -v
Failed to open /dev/dvb/adapter0: Is a directory

$ ls -l /dev/dvb/adapter0/
total 0
crw-rw----+ 1 root video 212, 0 Sep 25 23:30 demux0
crw-rw----+ 1 root video 212, 1 Sep 25 23:30 dvr0
crw-rw----+ 1 root video 212, 3 Sep 25 23:30 frontend0
crw-rw----+ 1 root video 212, 2 Sep 25 23:30 net0

I need to find more doco on how to use this utility.

> What I am not sure about is whether the tuner needs to be
> 'zapped' to a nice strong TV channel before you can use this.

> Anyway, hope some of these ideas are helpful.
> Cheers
> Vince

cheers

-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)

Re: f26: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Eyal Lebedinsky]

On 25/09/17 23:24, Vincent McIntyre wrote:
> On Mon, Sep 25, 2017 at 10:16:23AM +1000, Eyal Lebedinsky wrote:
> 
>>> Turn on debug printing for the modules of interest
>>> # echo 'module rtl2832 +p' > /sys/kernel/debug/dynamic_debug/control
>>> # echo 'module dvb_usb_rtl28xxu +p' > /sys/kernel/debug/dynamic_debug/control
>>
>> Have done this. Attached are the messages from a (failed) scandvb that fails for all multiplexes.
>>
>> The messages at the end continued at a high rate after the test finished (until I disabled debug
>> with '-p') and there was no user of the tuners. Maybe IR RC is active?
>   
> This looks like progress, now we can see more of the rather odd behaviour.
> 
> The tuned frequency does not progress monotonically,
> but wanders up and down the band.
>   Frequency      Delta
>   219166666
>   184833334  -34333332
>   191833334    7000000
>   191166666    -666668
>   177833334  -13333332
>   184166666    6333332
>   177166666   -7000000
>   191500000   14333334
>   184500000   -7000000
>   191833334    7333334
>   191166666    -666668
>   177500000  -13666666
>   219500000   42000000
>   184833334  -34666666
>   191500000    6666666
>   177833334  -13666666
>   184166666    6333332
>   191833334    7666668
>   191166666    -666668
>   177166666  -14000000
>   191500000   14333334
>   184500000   -7000000
>   177500000   -7000000
>   177833334     333334
>   177166666    -666668
> 
> The bandwidth and inversion type seem to be set correctly, at least.
> 
> Also:
>   - every instance of if_frequency=0 pset_iffreq=00000000
>     shows the same numbers - zero. Surely that can't be right.
>   - there seems to be no correlation at all between the AGC level
>     (automatic gain control, I assume) and the 'cnr raw' which I'm
>     guessing is some measure of the signal level.
> 
> I don't know where to start with decoding the rtl28xxu_ctrl_msg
> I'm afraid. It's quite possible the remote control is active.
> You can run
>     ir-keytable -v
> to show which remotes the system knows about.
> 
> We might get more clues to rub together from looking at where
> scandvb falls over, if you run it under 'strace'
>    strace -t -s 2048 -o ./scandvb.strace scandvb <arguments here>
> This will show the system calls being made. The -t will add timestamps
> to the output that could be correlated with the dmesg output.
> 
> I'm curious - did you try scandvb with one of your other tuners?
> 
> I'm actually not familiar with scandvb, I can't find a program
> with that name in the ubuntu repositories.
> What I've used before is dvbv5-scan, which is part of:
>   git://linuxtv.org/v4l-utils.git
> 
> That repo also includes the v4l2-compliance tool,
> which might be useful here. Something like:
>   ./v4l2-compliance -d /dev/dvb/adapter3 -a -T -v
> What I am not sure about is whether the tuner needs to be
> 'zapped' to a nice strong TV channel before you can use this.
> 
> Anyway, hope some of these ideas are helpful.
> Cheers
> Vince

I have now confirmed (tested) that the DTV2000 tunes properly when fed a signal
directly from the antenna. The other USB tuners need a stronger signal so they
are fed through a powered splitter (as it always was).

The issue left open is: can the tuner (is it the FC0013?) be set to handle the
stronger input? I have too many antenna cables attached right now...

cheers
-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)

Re: [progress]: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Eyal Lebedinsky]

On 18/09/17 14:26, Eyal Lebedinsky wrote:
> I have just upgraded to f24. I am now using the standard dvb_usb_rtl28xxu fe
> which logs messages suggesting all is well (I get the /dev/dvb/adapter? etc.)
> but I get no channels tuned when I run mythfrontend or scandvb.
> 
> Is anyone using this combination?
> Is this the correct way to use this tuner?
> 
> BTW:
> 
> Until f22 I was using the out of kernel driver from
>      https://github.com/jaredquinn/DVB-Realtek-RTL2832U.git
> but I now get a compile error.
> 
> TIA

This is an FYI in case anyone else stumbles on this thread.

While the problem persists, I managed to find a way around it for now.

I changed the antenna input.

Originally I used a powered splitter to feed all the tuners, and it worked
well with the out-of-kernel driver. This driver does not build or work with
a more modern kernel so I shifted to using dvb_usb_rtl28xxu which fails to
tune.

The new wiring splits (passive) the antenna in two, feeds one side directly
to the two "Leadtek Winfast DTV2000 DS PLUS TV" cards (through another passive
2-way) and the other side goes to the old powered splitter that feeds a few
USB tuners.

Now all tuners are happy. It seems that the "Leadtek Winfast DTV2000 DS PLUS TV"
cannot handle the amplified input while the USB tuners require it.

I hope that there is a way to set a profile in dvb_usb_rtl28xxu to attenuate
the input to an acceptable level thus unravelling the antenna cables rat-nest.

cheers

-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)

Re: [progress]: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Eyal Lebedinsky]

On 26/09/17 14:32, Eyal Lebedinsky wrote:
> On 18/09/17 14:26, Eyal Lebedinsky wrote:
>> I have just upgraded to f24. I am now using the standard dvb_usb_rtl28xxu fe
>> which logs messages suggesting all is well (I get the /dev/dvb/adapter? etc.)
>> but I get no channels tuned when I run mythfrontend or scandvb.
>>
>> Is anyone using this combination?
>> Is this the correct way to use this tuner?
>>
>> BTW:
>>
>> Until f22 I was using the out of kernel driver from
>>      https://github.com/jaredquinn/DVB-Realtek-RTL2832U.git
>> but I now get a compile error.
>>
>> TIA
> 
> This is an FYI in case anyone else stumbles on this thread.
> 
> While the problem persists, I managed to find a way around it for now.
> 
> I changed the antenna input.
> 
> Originally I used a powered splitter to feed all the tuners, and it worked
> well with the out-of-kernel driver. This driver does not build or work with
> a more modern kernel so I shifted to using dvb_usb_rtl28xxu which fails to
> tune.
> 
> The new wiring splits (passive) the antenna in two, feeds one side directly
> to the two "Leadtek Winfast DTV2000 DS PLUS TV" cards (through another passive
> 2-way) and the other side goes to the old powered splitter that feeds a few
> USB tuners.
> 
> Now all tuners are happy. It seems that the "Leadtek Winfast DTV2000 DS PLUS TV"
> cannot handle the amplified input while the USB tuners require it.
> 
> I hope that there is a way to set a profile in dvb_usb_rtl28xxu to attenuate
> the input to an acceptable level thus unravelling the antenna cables rat-nest.

Looking further, I think that it is the fc0013 tuner that should manage the lna
but does not. Compare it to the module in here
	https://github.com/jaredquinn/DVB-Realtek-RTL2832U/blob/master/src/tuner_fc0013.c
which I was using successfully (unfortunately this repo does not build or work
with current linux kernels).

Is this possible? If so, can it be fixed?

cheers

> cheers

-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)

Re: [progress]: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Vincent McIntyre]

On Tue, Sep 26, 2017 at 02:32:26PM +1000, Eyal Lebedinsky wrote:
> 
> While the problem persists, I managed to find a way around it for now.
> 
> I changed the antenna input.
> 
> Originally I used a powered splitter to feed all the tuners, and it worked
> well with the out-of-kernel driver. This driver does not build or work with
> a more modern kernel so I shifted to using dvb_usb_rtl28xxu which fails to
> tune.
> 
> The new wiring splits (passive) the antenna in two, feeds one side directly
> to the two "Leadtek Winfast DTV2000 DS PLUS TV" cards (through another passive
> 2-way) and the other side goes to the old powered splitter that feeds a few
> USB tuners.
> 
> Now all tuners are happy. It seems that the "Leadtek Winfast DTV2000 DS PLUS TV"
> cannot handle the amplified input while the USB tuners require it.
> 
> I hope that there is a way to set a profile in dvb_usb_rtl28xxu to attenuate
> the input to an acceptable level thus unravelling the antenna cables rat-nest.

Glad you had some success.

I did some more rummaging in v4l-utils. It may help you to know about
dvb-fe-tool, which gives information about the frontend device
(eg /dev/dvb/adapter0/frontend0). In particular you can monitor the
s/n as you make changes:

 # dvb-fe-tool --femon -a 0  #here doing adapter0/frontend0

It displays info about the signal quality and the carrier/noise (C/N) ratio,
which might help any investigation of where the driver fails to cope as
you change what you are feeding it.

I noticed your dvbv5-scan showed C/N around 20dB but the manpage shows
'good signal' with C/N of 36dB which suggests the device should be
expected to deal with higher signal levels.

Once you figured out the signal level, did a dvbv5-scan work with no
errors? In the example you showed I saw the channels getting 'lock'
but then some kind of error occurred.

Regards
Vince

Re: [progress]: dvb_usb_rtl28xxu not tuning "Leadtek Winfast DTV2000 DS PLUS TV"

[Eyal Lebedinsky]

On 26/09/17 22:45, Vincent McIntyre wrote:
> On Tue, Sep 26, 2017 at 02:32:26PM +1000, Eyal Lebedinsky wrote:
>>
>> While the problem persists, I managed to find a way around it for now.
>>
>> I changed the antenna input.
>>
>> Originally I used a powered splitter to feed all the tuners, and it worked
>> well with the out-of-kernel driver. This driver does not build or work with
>> a more modern kernel so I shifted to using dvb_usb_rtl28xxu which fails to
>> tune.
>>
>> The new wiring splits (passive) the antenna in two, feeds one side directly
>> to the two "Leadtek Winfast DTV2000 DS PLUS TV" cards (through another passive
>> 2-way) and the other side goes to the old powered splitter that feeds a few
>> USB tuners.
>>
>> Now all tuners are happy. It seems that the "Leadtek Winfast DTV2000 DS PLUS TV"
>> cannot handle the amplified input while the USB tuners require it.
>>
>> I hope that there is a way to set a profile in dvb_usb_rtl28xxu to attenuate
>> the input to an acceptable level thus unravelling the antenna cables rat-nest.
> 
> Glad you had some success.
> 
> I did some more rummaging in v4l-utils. It may help you to know about
> dvb-fe-tool, which gives information about the frontend device
> (eg /dev/dvb/adapter0/frontend0). In particular you can monitor the
> s/n as you make changes:
> 
>   # dvb-fe-tool --femon -a 0  #here doing adapter0/frontend0
> 
> It displays info about the signal quality and the carrier/noise (C/N) ratio,
> which might help any investigation of where the driver fails to cope as
> you change what you are feeding it.
> 
> I noticed your dvbv5-scan showed C/N around 20dB but the manpage shows
> 'good signal' with C/N of 36dB which suggests the device should be
> expected to deal with higher signal levels.
> 
> Once you figured out the signal level, did a dvbv5-scan work with no
> errors? In the example you showed I saw the channels getting 'lock'
> but then some kind of error occurred.

I had very good reception and scandvb tuned to all channels. I did not test
dvbv5-scan again but I expect it would work too (will try in the morning).

> Regards
> Vince

-- 
Eyal Lebedinsky (eyal@eyal.emu.id.au)