[linux-dvb] QAM 256 starts kicking out VDR widespread TT/FuSi FF cards / on AGC, etc, IMPL and Filter Optimisation

thomas schorpp t.schorpp at gmx.de
Sat Aug 5 15:47:56 CEST 2006


thomas schorpp wrote:
> thomas schorpp wrote:
> 
>> Marco Skambraks wrote:
>>
>>> hi,
>>>
>>
>>>
>>> any ideas?
>>>
>>> marco
>>>
>>
>> no.
>>
>> and the other devs seem to be in holidays or have no more interest in 
>> older FF cards.
>>
>> ill do trial'n'error with the filter registers or implement 
>> sysfs/procfs support in ves1820 for a trimmer GUI until I get access 
>> to professional mesurement equipment.
>>
> 
> ves1820 driver has DFE (decision feedback equalizer) set to off, however 
> web literature states this should be the better method than just 
> transversal.
> ill try it.
> and there more ideas from the stv driver code.
> 
> tom

DFE brings no qam256 relevevant improve, too:

static u8 ves1820_inittab[] = {
	0x69, 0x6A, 0x93, 0x12, 0x12, 0x46, 0x27, 0x0A, // 06 46 //07 1A
	0x43, 0x6A, 0xAA, 0xAA, 0x1E, 0x85, 0x43, 0x20,
	0xE0, 0x00, 0xA1, 0x00, 0x00, 0x00, 0x00, 0x00,
	0x00, 0x00, 0x00, 0x01, 0x32, 0x00, 0x00, 0x00, // 1C 00, 1E 00
	0x02, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
	0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
	0x00, 0x00, 0x00, 0x00, 0x40
};

no more much software optimization left. the stv code's "software filter regulation" 
cannot be easily adapted for now, too, the better DFE is hard to configure without using 
hardware testmode and lab equipment, and i dont believe in "matlab" simulations.

next trying to hardware chain missing ves1820 VAGC -> IF chip's extern AGC input 
or bypass if chip completely (at cost of analogue tv support).

CD1516 tuner is also suspect to degradation, design mismatch and so replacement 
with a more modern device. 



More information about the linux-dvb mailing list