Re: [PATCH] FM1216ME_MK3 some changes

[Dmitri Belimov <d.belimov@gmail.com>]

Hi Andy

> Dmitri,
> 
> 
> On Wed, 2009-04-22 at 17:48 +1000, Dmitri Belimov wrote:
> > Hi All
> > 
> > 1. Change middle band. In the end of the middle band the
> > sensitivity of receiver not good. If we switch to higher band,
> > sensitivity more better. Hardware trick.
> 
> This concerns me slightly as it does not match the datasheet (hence
> the design objectives) of the FM1236ME_MK3.

Yes, I know.

> How are you measuring sensitivity?  Do you know if it really is the
> middle-band preselector filter (and PLL and Mixer) or is it a problem
> with the input signal?  How do you know it is not manufacturing
> variations in the preselector filters with the particular tuner
> assembly you are testing?

Several years a go your customers write some messages about bad quality of TV
if frequency of TV is the end of band. It can be low band or middle. Our
hardware engeneer make some tests with hardware TV generator and our TV tuners.

If we set default frequency range for low and middle band, quality of TV signal 
on 159MHz and 442 MHz is bad. When we make our changes with moving end of bands
the quality of TV much better. And our system programmer for OS Windows use changed
bands for drivers. Customers be happy.

You can test it if in your placement available TV programm on 159MHz or 442MHz. This trick
can be usefull for other tuners.

> Also, as an alternative to using a different frequency for the
> bandswitch, have you considered setting the Auxillary Byte in the
> tuner chip (Infineon TUA6030?) to use external AGC and experimented
> with changing the tuner AGC take-over point (TOP) in the TDA9887?
> 
> By maximizing the gain in the tuner chip, but avoiding clipping, with
> the proper TOP setting, you minimize the contributions by the rest of
> the receive chain to the overall receiver Noise Figure:
> 
> http://en.wikipedia.org/wiki/Friis_formulas_for_noise
> 
> This may be a way to improve receiver sensitivity that does not
> conflict with the data sheet specification.
> 
> 
> 
> 
> > 2. Set correct highest freq of the higher band.
> 
> :)
> 
> This bothers me too; all the tuners in tuner-types.c have it set too
> high (999.0 MHz).  I think I rememeber at time when all the
> tuner_range definitions had a real value there.
> 
> It would be nice to have a real value there for all the tuners.  The
> function tuner-simple.c:simple_config_lookup() would then prevent
> attempts to tune to an unsupported frequnecy.
> 
> 
> 
> > 3. Set charge pump bit
> 
> This will improve the time to initially tune to a frequency, but will
> likely add some noise as the PLL continues to maintain lock on the
> signal.  If there is no way to turn off the CP after the lock bit is
> set in the tuner, it's probably better to leave it off for lower
> noise and just live with slower tuning.

We discuss with our windows system programmer about it. He sad that
in analog TV mode noise from PLL don't give any problem. But in digital TV mode
noise from PLL decreased BER.

> Leaving the CP bit set should be especially noticable ad FM noise when
> set to tune to FM radio stations.  From the FM1236ME_MK3 datasheet:
> "It is recommended to set CP=0 in the FM mode at all times."
> But the VHF low band control byte is also used when setting FM radio
> (AFAICT with a quick look at the code.)

Yes. You are right. We can swith CP off in FM mode.

With my best regards, Dmitry.

> 
> Regards,
> Andy
> 
> > diff -r 43dbc8ebb5a2 linux/drivers/media/common/tuners/tuner-types.c
> > --- a/linux/drivers/media/common/tuners/tuner-types.c	Tue
> > Jan 27 23:47:50 2009 -0200 +++
> > b/linux/drivers/media/common/tuners/tuner-types.c	Tue Apr 21
> > 09:44:38 2009 +1000 @@ -557,9 +557,9 @@ /* ------------
> > TUNER_PHILIPS_FM1216ME_MK3 - Philips PAL ------------ */ 
> >  static struct tuner_range tuner_fm1216me_mk3_pal_ranges[] = {
> > -	{ 16 * 158.00 /*MHz*/, 0x8e, 0x01, },
> > -	{ 16 * 442.00 /*MHz*/, 0x8e, 0x02, },
> > -	{ 16 * 999.99        , 0x8e, 0x04, },
> > +	{ 16 * 158.00 /*MHz*/, 0xc6, 0x01, },
> > +	{ 16 * 441.00 /*MHz*/, 0xc6, 0x02, },
> > +	{ 16 * 864.00        , 0xc6, 0x04, },
> >  };
> >  
> > 
> > 
> > Signed-off-by: Beholder Intl. Ltd. Dmitry Belimov
> > <d.belimov@gmail.com>
> > 
> > With my best regards, Dmitry.
> > --
> > video4linux-list mailing list
> > Unsubscribe
> > mailto:video4linux-list-request@redhat.com?subject=unsubscribe
> > https://www.redhat.com/mailman/listinfo/video4linux-list
>