[linux-dvb] PATCH: New driver for Zarlink ZL10313 based Compro S350/S300

[Jan D. Louw]

>> * "Jan D. Louw" <jd.louw at mweb.co.za> wrote:
>>> On Sunday 30 December 2007 11:36:12 Tino Reichardt wrote:
>>> > * "Jan D. Louw" <jd.louw at mweb.co.za> wrote:
>>> > > Hi List,
>>> > >
>>> > > Attached is a new driver for the Zarlink ZL10313/ZL10039 based 
>>> Compro
>>> > > S350/S300 budget DVB-S cards. Everything except 2 way diseqc 
>>> should work.
>>> >
>>> > I will test it on my Avermedia DVB-S Pro, which is ZL10313/ZL10036
>>> > based.
>>> >
>>> > Is it a good idea to rename the the driver to zl1031x.c / zl1003x.c ?
>>> >
>>> > zl1031x.c: zl100312/zl100313/intel6313
>>> > zl1003x.c: zl10036/zl10037/zl10038/zl10039 tuner
>>> >
>>> > Thanks for the driver, I will try it out ;)
>>>
>>> Many have suggested the possibility of combining the 
>>> vp310,mt312,zl313 and intel 6313 demodulators into a single driver. 
>>> Now that the 313 is working this will be easier.
>>>
>>> Unfortunately the zl10036 tuner register map differs from the zl10039 
>>> tuner, so this driver will definitely not work for you. However, 
>>> writing s zl10036 driver would be trivial, as a full datasheet is 
>>> available.
>>
>> Hello again,
>>
>> I have updated my old zl10036 patch, which is located @
>> http://www.mcmilk.de/projects/dvb-card/patches/
>>
>> It uses the whole patch from "Jan D. Louw" and adds support for the
>> zarlink zl10036/38 tuners. But there is some bug in zl1003x_set_params()
>> function ... so It doesn't get tuned.
>>
>> There is also a file called "everything.log.txt" - this is some test
>> tuning to german channels.
>>
>> Jan D. Louw, have you an idea, what is wrong in that function ?
>>
>> -- 
>> regards, TR
> 
> Hi Tino,
> 
> I had a quick look at the zl10036 datasheet and I can't see anything 
> obviously wrong. The formulas look OK. There is one wrong assignment:
> 
> if (!priv->has_loopthrough) buf[0] |= 0x01;
> 
> and some wrong comments:
> 
>         buf[6] = 0xd3;   /*  6: from datasheet */
>         buf[7] = 0x40;   /*  7: from datasheet */
> 
> but this shouldn't cause it not to work. I suspect this is some obscure 
> writing order bug. The zl10038 datasheet does provide more info on 
> registers 8-10 which might prove useful.
> 

Also, maybe some hardware problem (like pin SLEEP asserted) or wrong 
crystal freq?