[linux-dvb] Re: VES1820.c question

[Holger Waechtler <holger@convergence.de>]

Hi,

Dennis Noordsij wrote:
> Hi,
> 
> This is a question in response to Marko Kenttälä's patch for VES1820.c to 
> allow QAM_128 tuning using a Siemens DVB-C card (QAM_128 is needed for 
> Finnish channels).
> 
> Specifically these lines (NEWSTRUCT branch by the way):
> 
> 246 QAM_SETTING QAM_Values[] = {
> 247         {QAM_16 ,  16, 145, 164, 162, 145},
> 248         {QAM_32 ,  32, 150, 120, 116, 150},
> 249         {QAM_64 ,  64, 106,  70,  67, 106},
> 250         {QAM_128, 128, 126,  54,  52, 126},
> 251         {QAM_256, 256, 107,  38,  35, 107}
> 252 };
> 
> The last 4 values of each setting refer to registers 1, 5, 8 and 9.
> 
> However, the same values are used for registers 1 and 9, whereas the datasheet 
> specifies the following values for register 1:
> QAM_16: 140
> QAM_32: 140
> QAM_64: 106
> QAM_128: 120
> QAM_256: 92
> 
> Now, I understand there can be a million reasons why the sourcecode does 
> something else, and I wouldn't know any of them, but the reason I am asking 
> is that in order to tune to QAM_128 (remember Siemens DVB-C) Marko discovered 
> register 1 should be set to 120 instead of 126 for QAM_128.

(Hmm, I'd suppose a typo which did not showed up in our tests:)

Unfortunally we have only QAM_64 transmissions here, and there the value 
was correct...


> Which brings the 
> question, why does a different value than the datasheet work for for example 
> QAM_64, but not for QAM_128 ? 

For QAM_64 values in registers 1 and 9 were equal (106).

Holger



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