[linux-dvb] Re: LGDT3302 stuff

[Michael Krufky]

Mac Michaels wrote:

>On Tuesday 28 June 2005 07:03 pm, Michael Krufky wrote:
>  
>
>>To be honest with you, I am just guessing here on how to
>>give dvb support to my card.  I don't think that this is
>>a problem with your driver, rather a matter of
>>implementing it on a different board.  For all I know, it
>>might be working fine except that I might even be testing
>>it incorrectly.  I dont have any other dvb card and thus
>>have no experience with dvb.  
>>    
>>
>I have no dvb experience either. Testing was a problem for 
>me too. I started by getting the scan program to work 
>reliably and detect my local channels. Fortunately the scan 
>program had just been modified for OTA ATSC so it just 
>worked. I examined the v4l2 tuner code to see how to make 
>it work in the dvb code. I know you will need a different 
>file for channel frequencies than I use as cable channels 
>above 13 are allocated different frequencies from OTA 
>channels. 
>  
>
Looks like we're on the same page right here...

>For testing you might consider reducing the number of 
>independent variables by using OTA VSB. The scan program is 
>known to work with that.
>
>You might just try OTA with a piece of wire attached to the 
>center conductor or the RF connector. I live on top of a 
>hill in the country so multipath interference is not a 
>problem for me. A $15 RadioShack UHF aimed at the 
>transmitter really reduced the error rate.
>  
>
I tried using the FM Antennae from my bttv radio card... Of cource I 
didnt get any results, but I didnt get anything running the same test 
under Windows, either... This leads me to conclude that OTA VSB isn't 
available in Brooklyn, New York :-(

>>I believe that I have my 
>>patch set up correctly, and my only room for error is in
>>dvb-pll.c, or maybe in cx88-mpeg - but I doubt that.  
>>    
>>
>
>I put in some extra stuff that I found in the spec to make 
>the MircoTune tuner change channels faster. It may not work 
>with your tuner. Can you point me to a spec for your tuner?
>  
>
I am unable to locate specs for the Thomson DDT 7611.  What is the extra 
stuff for the Microtune tuner that you speak of?  I can try removing 
these and see if it helps...

>>But 
>>yes, I would love for you to help me, if possible.  I am
>>currently in the irc chat room #v4l and also in #linuxtv.
>> (8:05pm EST) Hopefully we can meet in there.
>>    
>>
>I got there at 10:02 CDT and said Hi. You left at 10:20 CDT 
>without responding.   
>  
>
I didn't see you... It could have been the same irc netsplit problem 
that we had that first day we met on irc... Oh well... Thank you for 
trying to meet me, I appreciate it.

>I usually watch tv from 5:30 to 11:30 skipping the 10:00 
>news. Then I check email and go to bed. When I get my 
>problem solved, video card runs out of resources trying to 
>decode a HD stream, I'll spend less time watching TV during 
>prime time. 
>  
>
I went to bed (relatively) early last night as well, after watching two 
hours of HBO's Carnivale... What a great show!

>>If not, 
>>what information can I give you that would enable you to
>>help me?
>>    
>>
>Does anything work?
>  
>
I got result codes 0x08 and 0x09 while scanning QAM256 from cable.  I 
told you about these in a previous email, where you explained to me the 
following:

	FE_HAS_SIGNAL     = 0x01,   /*  found something above the 
noise level */
	FE_HAS_CARRIER    = 0x02,   /*  found a DVB signal  */
	FE_HAS_VITERBI    = 0x04,   /*  FEC is stable  */
	FE_HAS_SYNC       = 0x08,   /*  found sync bytes  */
	FE_HAS_LOCK       = 0x10,   /*  everything's working... */
	FE_TIMEDOUT       = 0x20,   /*  no lock within the last ~2 
seconds */
	FE_REINIT         = 0x40    /*  frontend was reinitialized,  

...so this means that in many cases I am finding something above the 
noise level, and in some cases I am also finding sync bytes.... This is 
the best I got so far...

It has just occurred to me that I need to change the scanning 
frequencies in the atscscan program in order for it to handle US 
ATSC/QAM256 ... Can you advise on how to do so?

-- 
Michael Krufky