Mailing List archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[linux-dvb] Re: DVB-C tuning problems caused by inittab changesin ves1820.c since 2003-05-24



First thank you for fast reply.

--On Saturday, August 16, 2003 03:10:39 PM +0200 Robert Schlabbach
<robert_s@gmx.net> wrote:

> From: "Peter Bieringer" <pb@bieringer.de>
>>  u8 ves1820_inittab [] =
>>  {
>> -       0x69, 0x6A, 0x9B, 0x0A, 0x52, 0x46, 0x26, 0x1A,
>> +       0x69, 0x6A, 0x9B, 0x12, 0x12, 0x46, 0x26, 0x1A,
>> 
>> Rolling back this diff, all works fine again here with newest driver
>> 
>> I'm sure not familiar with this driver/chip, but sure the developers
>> can describe the reason for this change.
> 
> That's what CVS comments are for. The corresponding CVS comment is:
> 
> "use Robert's suggestions for CLKCONF (0x03) and CARCONF (0x04)
> please test in your local networks and report problems"
> 
> The old values meant:
> 
> CLKCONF = 0x0A = DYN | CLK_C[2:0]=2
> CARCONF = 0x52 = CAR_P[2]=1 | CAR_P[1:0]=1 | CAR_C[3:0]=2
> 
> The new values are the default values taken from the VES1820 data sheet:
> 
> CLKCONF = 0x12 = GAIN3 | CLK_C[2:0]=2
> CARCONF = 0x12 = CAR_P[2]=0 | CAR_P[1:0]=1 | CAR_C[3:0]=2
> 
> The reason for this change was that I found my Siemens Cable card often
> had problems locking after a channel change, due to the DYN and CAR_P[2]
> bits, which seem to cause the VES1820 to get stuck. These bits extend the
> carrier acquisition range, which is typically not necessary, since the
> frequencies in a cable system are usually correct.
> 
> But the GAIN3 bit may be causing a problem. Could you try a few variations
> of these two register settings and report back which work for you? For
> CARCONF, I'd prefer to leave the CAR_P[2] bit off, so there'd only be 0x12
> to test. For CLKCONF, we should test variations of DYN, so please try:
> 
> 0x69, 0x6A, 0x9B, 0x02, 0x12, 0x46, 0x26, 0x1A,
> (No DYN, No GAIN3, No CAR_P[2])
> 
> 0x69, 0x6A, 0x9B, 0x0A, 0x12, 0x46, 0x26, 0x1A,
> (DYN, No GAIN3, No CAR_P[2])
> 
> If the first of these two works, I'd suggest that one, i.e. only change
> the 0x12 in the fourth position to 0x02 (disable the GAIN3 bit).

Both not working, but following is working:

0x69, 0x6A, 0x9B, 0x12, 0x52, 0x46, 0x26, 0x1A,

means

CLKCONF = 0x12
CARCONF = 0x52 

I'm not sure about your comment that frequencies in cable systems are
usually correct. Earlier versions display AFC information somewhere and I
saw a not very small difference.

Is it possible to enhance the module ves1280 to modify this parameters with
module options?

        Peter
-- 
Dr. Peter Bieringer                     http://www.bieringer.de/pb/
GPG/PGP Key 0x958F422D               mailto: pb at bieringer dot de 
Deep Space 6 Co-Founder and Core Member  http://www.deepspace6.net/


-- 
Info:
To unsubscribe send a mail to ecartis@linuxtv.org with "unsubscribe linux-dvb" as subject.



Home | Main Index | Thread Index