[linux-dvb] STB6100, STB0899 programming information?

Manu Abraham abraham.manu at gmail.com
Tue Aug 29 14:36:48 CEST 2006


Robert Schlabbach wrote:
> From: "thomas schorpp" <t.schorpp at gmx.de>
>> used Your freemail adress in the support form?
>> no good idea, linear-tech e.g. say they dont accept gmx, etc.
> 
> Stupid companies. My GMX account is not even freemail, I'm paying for it...
> 
>> even *sales*@twinhan doesnt care about freemailings, too,
>> i asked about their new dvb-s2 ci card.
> 
> Maybe they don't have anything to say about it, because it only exists in
> their ads, but not as an actual product...?


Shortage of chips ? More than that, the Twinhan DVB-S2 card, VP-1040
uses the Mantis PCI bridge, which is a new chipset as well.


> 
>> You're working on the new TT dvb-s2 card, arent You?
> 
> Yes.
> 
>> Asked Technotrend?
> 
> Haven't had any responses from TT for a while. I wonder if my e-mail
> address is in their spam filter...
> 
>>> far is that the STB6100 seems to have 12 registers, while the STB0899
>>> has over 256 - but I just can't figure out how register addressing
>>> works on the STB0899. It seems to be neither an 8-bit subaddress nor
>>> a 16-bit subaddress. Stupid secrecy-ladden design... :(
>> how did You find out? linux/windows i2c sniffer/probing tools? which?
> 
> No, the TT Windows driver seems to have a register dump function (which I
> don't know how to activate) and contains a list of strings will all
> register names and their bits, and one can extract which bits in the
> register they are...
>


You can look at the STB0899 driver over here.
http://thadathil.net:8000/cgi-bin/hgwebdir.cgi/mp-stb0899

ST has been quite kind to provide help to put this driver together. They
did provide information that which was missing from the datasheets as
well, as they wanted to go for a very good driver.

The current driver, needs some more finishing touches.
To see the registers, you can take a look at stb0899_reg.h


>> what for should a demodulator need 256 registers? cant be, modern
>> fully automatic devices come with lesser configuration effort, normally
> 
> Wolfgang's post later in this thread seems to indicate that the STB0899's
> DVB-S2 function is actually somewhat unfinished and requires host CPU
> interaction to work. Maybe we should wait for a fully finished
> "STB0999"...?
> 
>> disassembling the windows driver or studying it with remote kernel
>> debugging would be a hard long way... :(
> 
> Especially if it requires quite some software overhead to run this stupid
> chip... :(

It requires some additional loops like checking for the Pilot symbols
(BC mode) etc
Carrier Synchronization, Checking for Unique Words etc. Hence it looks a
bit complex.


Manu




More information about the linux-dvb mailing list