Mailing List archive

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

[linux-dvb] Re: Patch for cards with the Philips CD1516 tuner box (spXXXX?)



> --- DVB/driver/frontends/ves1820.c.bak  Tue Oct 15 00:38:03 2002
> +++ DVB/driver/frontends/ves1820.c      Tue Oct 15 01:30:58 2002
> @@ -151,7 +151,7 @@
>  int tuner_set_tv_freq (struct dvb_frontend *frontend, u32 freq, u8 pwr)
>  {
>          u32 div;
> -       static int freq_mod [] = { 36750000, 36125000 };
> +       static int freq_mod [] = { 36125000, 36125000 };
>
> After this, the locking frequencies on my Siemens rev1.3 board are
> shifted exactly 625 kHz upwards. (i.e. what once used to lock at
> 150000 kHz now locks at 150625 kHz.) I know for certain that
> 150000 kHz should be the right frequency, since this is what both the
> cable operator and a Hauppauge DVB-C card tells me. Maybe we might
> try with different freq_mod values depending on the revision.

Can I ask where you found the 36750000 value ? The tuner datasheet first 
somewhere claims the IF frequency is 36 MHz, and on another page 42.25 MHz. 
Though I am no HAM buff so maybe I don't know what to look for :-)

The original (before NEWSTRUCT) drivers used:

tuner.c (line 213):
        {"SPXXXX", MITEL, DVBC,
             40000000UL, 870000000UL, 62500, 1, 36125000,
            174000000UL, 454000000UL, 0xa1, 0x92, 0x34, 0x8e},

i.e. 36125000.

Note that I am not saying you are wrong :-) I am just curious :)

On a related note (not specifically for you Juha, though feel free to answer 
:)), does the zigzagging code use the minimum step size of 62.5 kHz ? There's 
no point in calling set_tv_tuner_freq with steps smaller than that because 
they all get converted to a 15 bit 62.5 kHz step anyway. 

Cheers,
Dennis


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



Home | Main Index | Thread Index