Hi, Mark
Mark Callaghan wrote:
> Hi Nick, Hartmut,
>
> I read your posts recently on some "interesting" behaviour of the Compro
> DVB-T300 with different input signal levels. I suspect I may be having
> the same problem that Nick had with his external antenna.
>
> I have an external antenna that gives good pictures on my analog TV,
> with some minor ghosting on a UHF channel around 500MHz. I also had good
> reception from a Twinhan (Visionplus, I think) for all local channels,
> from 177.5 MHz to 226.5 MHz and 536.5MHz. The Compro with its Windows
> software also works well on all channels, reporting signal strength of
> 85% for some, up to 100% (at 536.5MHz).
>
> I'm still in the process of getting my Myth box going properly, but I'm
> at the stage where I have problems with DTV reception with the DVB-T300.
> Under Linux, with scan, the card picks up all channels but the 536.5MHz.
> During the scan on 536.5, the "signal quality" jumps between ~10% to
> ~90%, going in and out of lock. The final result is no lock, and the
> signal at 536.5 is not recognized.
>
> I think I saw this behaviour on one of the ~200MHz channels as well, and
> I think it was one that reported 100% signal strength under the windows app.
>
> It is making me wonder if there is something wrong with the AGC settings
> for the tuner/driver? Or perhaps the signal quality or signal level
> indication is being misinterpreted by the driver?
>
> I thought about having a look at it myself, but I can't find a data
> sheet for either the TD1316 or TDA10046. I presume Philips are
> restricting distribution, with an NDA or similar. Do you have datasheets
> Hartmut? Have you tried to get them directly from Philips?
>
Well, nor the original author, neither me spent much effort on the signal
strength readout. The value reported is the AGC level and its not renormalized
to the tuner needs. And the driver works with different tuners. More important
is the raw ber.
But the point is that these values are not reliable until the decoder is locked.
I suspect you have another problem i observed here as well.
I checked in an update of the driver beginning of november. Please download and
install it and try again.
I changed 3 things:
- It uses different clock frequencies depending on the tuner configuration. This
improves the sensitivity.
- I modified the initialization sequence. This makes the DSP boot more reliable.
- I changed a configuration that makes the decoder work with a higher crystal
frequency deviation. This is important at high input frequencies and bad signal
quality. The penalty is a longer lock in time.
Another possible cause is that your initial config file is not ok. Opposite to
some other drivers, it honours the OFDM parameters and will not lock if i.e.
the guard interval or the code rate is not set properly.
Best regards
Hartmut
