PCTV 520e on Linux

PCTV 520e on Linux

[Claus Olesen]

I'm looking for a USB stick for DVB-C on Linux,
have good experience with the PCTV nanoStick T2 290e for DVB-T on
Linux (except for the replug issue)
http://www.pctvsystems.com/Products/ProductsEuropeAsia/Digitalproducts/PCTVnanoStickT2/tabid/248/language/en-GB/Default.aspx
and wonder if anyone know the status of support, if any, of the PCTV
QuatroStick nano 520e for DVB-C on Linux?
http://www.pctvsystems.com/Products/ProductsEuropeAsia/Hybridproducts/PCTVQuatroSticknano/tabid/254/language/en-GB/Default.aspx

Re: PCTV 520e on Linux

[Devin Heitmueller]

On Thu, Oct 13, 2011 at 10:49 AM, Claus Olesen <ceolesen@gmail.com> wrote:
> I'm looking for a USB stick for DVB-C on Linux,
> have good experience with the PCTV nanoStick T2 290e for DVB-T on
> Linux (except for the replug issue)

I believe the replug issue is probably fixed if you're using the
current media_build tree.

> http://www.pctvsystems.com/Products/ProductsEuropeAsia/Digitalproducts/PCTVnanoStickT2/tabid/248/language/en-GB/Default.aspx
> and wonder if anyone know the status of support, if any, of the PCTV
> QuatroStick nano 520e for DVB-C on Linux?
> http://www.pctvsystems.com/Products/ProductsEuropeAsia/Hybridproducts/PCTVQuatroSticknano/tabid/254/language/en-GB/Default.aspx

No support currently.  I have the stick, but haven't had any time to work on it.

Devin

-- 
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com

Re: PCTV 520e on Linux

[Antti Palosaari]

On 10/13/2011 06:57 PM, Devin Heitmueller wrote:
>> and wonder if anyone know the status of support, if any, of the PCTV
>> QuatroStick nano 520e for DVB-C on Linux?
>> http://www.pctvsystems.com/Products/ProductsEuropeAsia/Hybridproducts/PCTVQuatroSticknano/tabid/254/language/en-GB/Default.aspx
>
> No support currently.  I have the stick, but haven't had any time to work on it.

Is that EM28xx + DRX-K + TDA18217 ? And analog parts...

regards
Antti

-- 
http://palosaari.fi/

Re: PCTV 520e on Linux

[Devin Heitmueller]

On Thu, Oct 13, 2011 at 12:07 PM, Antti Palosaari <crope@iki.fi> wrote:
>> No support currently.  I have the stick, but haven't had any time to work
>> on it.
>
> Is that EM28xx + DRX-K + TDA18217 ? And analog parts...

You were close:  em2884, drx-k, xc5000, and for analog it uses the afv4910b.

Devin

-- 
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com

Re: PCTV 520e on Linux

[Antti Palosaari]

On 10/13/2011 07:15 PM, Devin Heitmueller wrote:
> On Thu, Oct 13, 2011 at 12:07 PM, Antti Palosaari<crope@iki.fi>  wrote:
>>> No support currently.  I have the stick, but haven't had any time to work
>>> on it.
>>
>> Is that EM28xx + DRX-K + TDA18217 ? And analog parts...
>
> You were close:  em2884, drx-k, xc5000, and for analog it uses the afv4910b.

Then it should be peace of cake at least for digital side.


Antti
-- 
http://palosaari.fi/

Re: PCTV 520e on Linux

[Devin Heitmueller]

On Thu, Oct 13, 2011 at 12:19 PM, Antti Palosaari <crope@iki.fi> wrote:
>> You were close:  em2884, drx-k, xc5000, and for analog it uses the
>> afv4910b.
>
> Then it should be peace of cake at least for digital side.

I don't think we've ever done xc5000 on an em28xx before, so it's
entirely possible that the xc5000 clock stretching will expose bugs in
the em28xx i2c implementation (it uncovered bugs in essentially every
other bridge driver I did work on).

That, and we don't know how much is hard-coded into the drx-k driver
making it specific to the couple of device it's currently being used
with.

But yeah, it shouldn't be rocket science.  I added support for the
board in my OSX driver and it only took me a couple of hours.

Devin

-- 
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com

Re: PCTV 520e on Linux

[Benjamin Larsson]

On 10/13/2011 07:48 PM, Devin Heitmueller wrote:
> On Thu, Oct 13, 2011 at 12:19 PM, Antti Palosaari <crope@iki.fi> wrote:
>>> You were close:  em2884, drx-k, xc5000, and for analog it uses the
>>> afv4910b.
>> Then it should be peace of cake at least for digital side.
> I don't think we've ever done xc5000 on an em28xx before, so it's
> entirely possible that the xc5000 clock stretching will expose bugs in
> the em28xx i2c implementation (it uncovered bugs in essentially every
> other bridge driver I did work on).
>
> That, and we don't know how much is hard-coded into the drx-k driver
> making it specific to the couple of device it's currently being used
> with.
>
> But yeah, it shouldn't be rocket science.  I added support for the
> board in my OSX driver and it only took me a couple of hours.
>
> Devin
>

Eddi De Pieri has patches for the HVR-930C that works somewhat. The
hardware in that stick is the same.

MvH
Benjamin Larsson

Re: PCTV 520e on Linux

[Devin Heitmueller]

On Thu, Oct 13, 2011 at 7:06 PM, Benjamin Larsson <benjamin@southpole.se> wrote:
> On 10/13/2011 07:48 PM, Devin Heitmueller wrote:
>> On Thu, Oct 13, 2011 at 12:19 PM, Antti Palosaari <crope@iki.fi> wrote:
>>>> You were close:  em2884, drx-k, xc5000, and for analog it uses the
>>>> afv4910b.
>>> Then it should be peace of cake at least for digital side.
>> I don't think we've ever done xc5000 on an em28xx before, so it's
>> entirely possible that the xc5000 clock stretching will expose bugs in
>> the em28xx i2c implementation (it uncovered bugs in essentially every
>> other bridge driver I did work on).
>>
>> That, and we don't know how much is hard-coded into the drx-k driver
>> making it specific to the couple of device it's currently being used
>> with.
>>
>> But yeah, it shouldn't be rocket science.  I added support for the
>> board in my OSX driver and it only took me a couple of hours.
>>
>> Devin
>>
>
> Eddi De Pieri has patches for the HVR-930C that works somewhat. The
> hardware in that stick is the same.
>
> MvH
> Benjamin Larsson

While the basic chips used are different, they are completely
different hardware designs and likely have different GPIO
configurations as well as IF specs.

Devin

-- 
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com

Re: PCTV 520e on Linux

[Mauro Carvalho Chehab]

Em 13-10-2011 20:19, Devin Heitmueller escreveu:
> On Thu, Oct 13, 2011 at 7:06 PM, Benjamin Larsson <benjamin@southpole.se> wrote:
>> On 10/13/2011 07:48 PM, Devin Heitmueller wrote:
>>> On Thu, Oct 13, 2011 at 12:19 PM, Antti Palosaari <crope@iki.fi> wrote:
>>>>> You were close:  em2884, drx-k, xc5000, and for analog it uses the
>>>>> afv4910b.
>>>> Then it should be peace of cake at least for digital side.
>>> I don't think we've ever done xc5000 on an em28xx before, so it's
>>> entirely possible that the xc5000 clock stretching will expose bugs in
>>> the em28xx i2c implementation (it uncovered bugs in essentially every
>>> other bridge driver I did work on).
>>>
>>> That, and we don't know how much is hard-coded into the drx-k driver
>>> making it specific to the couple of device it's currently being used
>>> with.
>>>
>>> But yeah, it shouldn't be rocket science.  I added support for the
>>> board in my OSX driver and it only took me a couple of hours.
>>>
>>> Devin
>>>
>>
>> Eddi De Pieri has patches for the HVR-930C that works somewhat. The
>> hardware in that stick is the same.
>>
>> MvH
>> Benjamin Larsson
> 
> While the basic chips used are different, they are completely
> different hardware designs and likely have different GPIO
> configurations as well as IF specs.

The IF settings for xc5000 with DRX-K are solved with this patch:
	http://patchwork.linuxtv.org/patch/7932/

Basically, DRX-K will use whatever IF the tuner uses.

I've sent to Eddi to get some feedback, but he never returned back.

> 
> Devin
> 

Re: PCTV 520e on Linux

[Devin Heitmueller]

On Fri, Oct 14, 2011 at 9:19 AM, Mauro Carvalho Chehab
<mchehab@redhat.com> wrote:
>> While the basic chips used are different, they are completely
>> different hardware designs and likely have different GPIO
>> configurations as well as IF specs.
>
> The IF settings for xc5000 with DRX-K are solved with this patch:
>        http://patchwork.linuxtv.org/patch/7932/
>
> Basically, DRX-K will use whatever IF the tuner uses.

While I fundamentally disagree with this change, I'm not going to nack
it.  That said, this wasn't the issue I was concerned with.  My
suggestion was simply that you cannot assume that all devices that
happen to have a particular demod and tuner combo will always use the
same IF configuration.  The PCB layout can effect the optimal IF.

This is one of those things that (like many tuners in the LinuxTV
tree) will probably work good enough to get a signal lock for whoever
added the board profile, but will result in poor tuning performance
(and a failure to work in less-than-ideal reception conditions).

All that said, if somebody actually intends to hack on it, I can look
up what the correct IF is for the 520e.

Devin

-- 
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com

AW: PCTV 520e on Linux

[Sönke Brandt]

 Just a quick note: The 520e does use the TDA18271 tuner, not an XC5000.

 Soenke.

-----Ursprüngliche Nachricht-----
Von: linux-media-owner@vger.kernel.org [mailto:linux-media-owner@vger.kernel.org] Im Auftrag von Devin Heitmueller
Gesendet: Freitag, 14. Oktober 2011 15:28
An: Mauro Carvalho Chehab
Cc: Benjamin Larsson; linux-media@vger.kernel.org; Eddi De Pieri
Betreff: Re: PCTV 520e on Linux

On Fri, Oct 14, 2011 at 9:19 AM, Mauro Carvalho Chehab
<mchehab@redhat.com> wrote:
>> While the basic chips used are different, they are completely
>> different hardware designs and likely have different GPIO
>> configurations as well as IF specs.
>
> The IF settings for xc5000 with DRX-K are solved with this patch:
>        http://patchwork.linuxtv.org/patch/7932/
>
> Basically, DRX-K will use whatever IF the tuner uses.

While I fundamentally disagree with this change, I'm not going to nack
it.  That said, this wasn't the issue I was concerned with.  My
suggestion was simply that you cannot assume that all devices that
happen to have a particular demod and tuner combo will always use the
same IF configuration.  The PCB layout can effect the optimal IF.

This is one of those things that (like many tuners in the LinuxTV
tree) will probably work good enough to get a signal lock for whoever
added the board profile, but will result in poor tuning performance
(and a failure to work in less-than-ideal reception conditions).

All that said, if somebody actually intends to hack on it, I can look
up what the correct IF is for the 520e.

Devin

-- 
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com
--
To unsubscribe from this list: send the line "unsubscribe linux-media" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: PCTV 520e on Linux

[Mauro Carvalho Chehab]

Em 14-10-2011 10:28, Devin Heitmueller escreveu:
> On Fri, Oct 14, 2011 at 9:19 AM, Mauro Carvalho Chehab
> <mchehab@redhat.com> wrote:
>>> While the basic chips used are different, they are completely
>>> different hardware designs and likely have different GPIO
>>> configurations as well as IF specs.
>>
>> The IF settings for xc5000 with DRX-K are solved with this patch:
>>        http://patchwork.linuxtv.org/patch/7932/
>>
>> Basically, DRX-K will use whatever IF the tuner uses.
> 
> While I fundamentally disagree with this change, I'm not going to nack
> it.  That said, this wasn't the issue I was concerned with.  My
> suggestion was simply that you cannot assume that all devices that
> happen to have a particular demod and tuner combo will always use the
> same IF configuration.  The PCB layout can effect the optimal IF.
> 
> This is one of those things that (like many tuners in the LinuxTV
> tree) will probably work good enough to get a signal lock for whoever
> added the board profile, but will result in poor tuning performance
> (and a failure to work in less-than-ideal reception conditions).

This patch doesn't prevent customizing the IF. It will just avoid the
need of setting the IF on both xc5000 and drx-k. Basically, (some) DRX-K
based boards use different IF's depending on the bandwidth and delivery
system type. Instead of adding a complex logic that would allow such
kind of IF adjustments on both, drx-k will simply inquire the tuner about
what IF is currently used.

> 
> All that said, if somebody actually intends to hack on it, I can look
> up what the correct IF is for the 520e.
> 
> Devin
> 

Re: PCTV 520e on Linux

[Devin Heitmueller]

On Fri, Oct 14, 2011 at 10:01 AM, Sönke Brandt <SBrandt@pctvsystems.com> wrote:
>  Just a quick note: The 520e does use the TDA18271 tuner, not an XC5000.
>
>  Soenke.

Wow, how the hell did I screw that up?  Of course Sönke is correct.  I
momentarily got the 520e confused with the HVR-930c (I've done work on
both in the past).

Regards,

Devin

-- 
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com

Re: PCTV 520e on Linux

[Mauro Carvalho Chehab]

Em 14-10-2011 11:04, Devin Heitmueller escreveu:
> On Fri, Oct 14, 2011 at 10:01 AM, Sönke Brandt <SBrandt@pctvsystems.com> wrote:
>>  Just a quick note: The 520e does use the TDA18271 tuner, not an XC5000.

The tda18271-dd/drx-k/em28xx combination works fine, provided that the GPIO
initialization enables both tuner and demod during probe time. Currently, the
device I used to add support for it (a Terratec H5) has a hack to enable
the devices: it just replies whatever initialization the original driver does.

When I have some time, I'll fix that, but I'm not urging doing so, because it
just works ;)

In order to add support for PCTV 520e, it is probably a matter of just set the
GPIO's.

>>
>>  Soenke.
> 
> Wow, how the hell did I screw that up?  Of course Sönke is correct.  I
> momentarily got the 520e confused with the HVR-930c (I've done work on
> both in the past).
> 
> Regards,
> 
> Devin
> 

Re: PCTV 520e on Linux

[Devin Heitmueller]

On Fri, Oct 14, 2011 at 12:14 PM, Mauro Carvalho Chehab
<mchehab@redhat.com> wrote:
> The tda18271-dd/drx-k/em28xx combination works fine, provided that the GPIO
> initialization enables both tuner and demod during probe time. Currently, the
> device I used to add support for it (a Terratec H5) has a hack to enable
> the devices: it just replies whatever initialization the original driver does.
>
> When I have some time, I'll fix that, but I'm not urging doing so, because it
> just works ;)
>
> In order to add support for PCTV 520e, it is probably a matter of just set the
> GPIO's.

Complements of our friends at PCTV:

520e:
GPIO02: Decoder Reset, active-low
GPIO04: Decoder Suspend, active-low
GPIO06: Demod Reset, active-low
GPIO07: LED on, active-high

Devin

-- 
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com

Re: PCTV 520e on Linux

[Mauro Carvalho Chehab]

Em 14-10-2011 13:41, Devin Heitmueller escreveu:
> On Fri, Oct 14, 2011 at 12:14 PM, Mauro Carvalho Chehab
> <mchehab@redhat.com> wrote:
>> The tda18271-dd/drx-k/em28xx combination works fine, provided that the GPIO
>> initialization enables both tuner and demod during probe time. Currently, the
>> device I used to add support for it (a Terratec H5) has a hack to enable
>> the devices: it just replies whatever initialization the original driver does.
>>
>> When I have some time, I'll fix that, but I'm not urging doing so, because it
>> just works ;)
>>
>> In order to add support for PCTV 520e, it is probably a matter of just set the
>> GPIO's.
> 
> Complements of our friends at PCTV:
> 
> 520e:
> GPIO02: Decoder Reset, active-low
> GPIO04: Decoder Suspend, active-low
> GPIO06: Demod Reset, active-low
> GPIO07: LED on, active-high

What are the USB ID's for the device? I may try to do a patch for it during this
weekend, if I found time to add support for a few other devices that Terratec
gently donated me.

Thanks!
Mauro
> 
> Devin
> 

Re: PCTV 520e on Linux

[Devin Heitmueller]

On Fri, Oct 14, 2011 at 1:39 PM, Mauro Carvalho Chehab
<mchehab@redhat.com> wrote:
> What are the USB ID's for the device? I may try to do a patch for it during this
> weekend, if I found time to add support for a few other devices that Terratec
> gently donated me.

510e 2304:0242
520e 2013:0251, 2013:0252

Of course, you shouldn't just blindly check anything in.  That said,
seems there are people on this list who have at least some of these
variants.

Devin

-- 
Devin J. Heitmueller - Kernel Labs
http://www.kernellabs.com