Mailing List archive

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

[linux-dvb] Re: CX22702 conexant/leadtek support



> > I'm also guessing that Chris Pascoe is one person on this list who can
> > really answer the question. Much extrapolation involved in that guess, but
> > reason being that in his linux driver additions for the MT352 (still not
> > committed to linux-dvb..?) for the FusionHDTV DVB-T1 (non-lite version),
> > there is a bit of code in the (LG) tuner config that goes:

Commit will be soon - waiting on CVS access so I can perform a number of
small commits to clean up..

> > #if IN_GERMANY
> >         tuner_regs[3] |= 0x08;  /* Activate IF SAW filter */
> > #endif
> >
> > (Presumably this define should be changed to bandwidth==8 and probably
> > already has, i haven't downlaoded it recently)

The notes I acquired from LG on their tuner indicated that the IF saw
filter was to be activated only on the UHF band, and only in Germany.  If
Germany only uses UHF for DVB-T then this change makes sense and I'll
update this to activate it on 8MHz bandwidth and UHF.  Is there a logical
reason why one would only want to activate said filter on a 8MHz
transponder but only on UHF (and not VHF) and only in one particular
country?  Is there some kind of co-channel interference that it helps with
in this particular environment?

> > So, when he releases his drivers for the DVB-T Lite we will hopefully all
> > find out! Or he might see this and answer (or, more likely, tell me to
> > stop my long-winded idle speculation :)
>
> I get it - there are two DIVCO cards, one of which has an LG-Z201 and one has
> a Thomson 7579. They both use the TUA6064 chip. However, the pins to control
> the SAW filter are GPIOs - so they could vary from tuner to tuner.
>
> Chris, I've CCed this your direct dvb address as well - would you be able to
> confirm if the Thomson 7579 has the same GPIO line configuration as the
> LG-Z201?

The 7579 and the Z201 have differing GPIO configurations.

For the Z201, it's:

	if (freq < 157500000)
                bs = 0x01;
        else if (freq < 443250000)
                bs = 0x02;
        else
                bs = 0x04;

From my notes for what I observed on the I2C bus talking to the
Thomson 7579, tuning a 7MHz channel in Australia:

        if (freq < 443250000)           /* mid band */
                bs = 0x12;
        else                            /* high band */
                bs = 0x18;

I'll admit I didn't open the 7579 module to determine that it used a
TUA6034, and I have just confirmed that it works here with 0x02 and 0x08,
too.

The TUA6034 PLL IC uses the low two bits of the band selection byte to
perform band selection byte internally, and they also drive two of the
external GPO pins.  The truth table for the low two bits of the band
select byte is:

0x00: High band
0x01: Low band
0x02: Mid band
0x03: PLL powered down

and most manufacturers will use P0 to activate whatever they need
externally for the low band and P1 to do the same for the mid band.  Then
they are free to select from the remaining 3 GPOs for the high band
selection and for SAW filter activation.

Andrew, where did you find the product notes on the Thomson 7579 to
determine that it has no SAW filter?  The recently released document was
for the 759x family; I havent' seen anything on the 757x.

Regards
Chris




Home | Main Index | Thread Index