[linux-dvb] Re: DVB signal

[wolfgang@leila.ping.de (Wolfgang Wegner)]

Hi,
On Tue, Jun 01, 2004 at 09:47:31PM +0200, Robert Schlabbach wrote:
> From: "Wolfgang Wegner" <wolfgang@leila.ping.de>
> > Furthermore, percentage readings are only one possibility.
> The _only_ possibility, I'd say.
hmm... :)
> 
> > For comparison, dBµV oder dBm values would be best IMHO,
> Name a receiver card from which you actually can get such a reading. So far
> I've only seen two demodulator data sheets which gave a formula to
> calculate dB from the SNR reading. All other data sheets won't tell you how
[...]
I do not know if it is really a demodulator data sheet that helps here.
The demodulator can give an estimation about SNR value (an example here
is TDA10046), but for signal strength, tuner and (if present) IF amp are
in between, such that the tuner module manufacturer would have to provide
the conversion table between AGC value(s) and input signal level.

As long as it is hard to find really needed information in the data sheets
(crystal frequency, use of different AGC loops the demodulator chip offers)
these are, of course, dreams of the future.

I tried to implement a conversion by piece-wise linear approximation for
a thomson frontend (using STV0360), which even gave quite good results
(as i only had one evalution board, i do not know how good the results
are reproduced with different production charges). As far as I remember,
the TU1216 (Philips, TDA10046) also gave quite stable register values,
but I did not have time to make a table and implement a conversion yet.
The cable version (CU1216, using TDA10021) seemed to have only a limited
range of useful readings, but I do not know about cable levels anyways...

I do not say the hardware we have gives the right values, but what I
would ask for is an interface that does not limit future hardware
because of buggy (or at least not very comfortable) current hardware.

> Regards,
> --
> Robert Schlabbach

Regards,
Wolfgang