[vdr] Re: burned DVB-s

[Roberto Deza <rdeza@unav.es>]

Hi Thomas and all,

At 17:16 18/10/2002 +0200, Thomas Breuer wrote:

>I have some hardware info + questions on the AV7x.

Very interesting your mail!.

>... I already asked
>around for the technical data sheet of the chip. Do you know if this is
>available? I only have the PDF with the "Functional Specifications".
>Actually,

I only have the same "common" datashet, available in many places; ie:

http://www.cdaniel.de/download/AV711x_3_1.pdfs

On the thread below, I guy ('xanti') commented to have "a pair of PDF's 
of the DSP pulled, unfortunately password protected" (note that I don't 
read German :-( ), but I don't have contacted with him:

<http://x-satboard.irde.to/ubb/ultimatebb.php?ubb=get_topic;f=30;t=000018#00
0000>

>... I am only interested in the schematic of the output stages.
>Any info on the outputs only would be great as well.

I'd like to know its (and many other details :-), but the most close
approach that I can think is to see how other Digital Video Encoders
(maybe from Texas) do it.

>Some time ago I had a rev 1.3 card here which had been killed by
>applying 12 volts to the composite output. The result was a short
>between the supply pins of the AV7111, so the card was useless.
>However, by applying a current of several amps into the dead output, I
>was able to remove the short. The output is still dead, of course, but
>the DSP is functional again and the card can still be used as a
>secondary card now.

Wooooow!!!, this is as the "recovery" of shorted NiCa baterys, but much
more miraculous :-).

>Most likely there are clamping diodes from each output to the supply
>rails, one of which had been molten when the 12 volts had been applied.

Very likely.

>I can not understand why even the original composite cinch output goes
>directly to the chip (well, there are some caps plus one coil in
>between, but there is *no* DC discoupling).

The analog electronic isn't my preffered, but I think that would not be
easy to block the DC (without disturb the bandwith and levels) in a video 
output and anyway, a big capacitor could allow to pass enought current/
voltage to break the chip. Perhaps, a "emiter follower" transistor on 
each output could have been "wise", but the image quality should not 
improve. Anyway, the *fact* is that the outputs are not protected and the 
common ground is a good protection.

>I read the postings of you and some others regarding the attempts of
>identifying the RGB outputs pins for rev 2.1 cards back at the beginning
>of this year. Well, the pins *are* available! There was a request in the
>German newsgroup de.science.electronics back in July which triggered me
>to take a close look at my rev 1.3 card. The outputs are located
>as follows at the outside of the thin PCB which carries the chip
>(take the wide gold track at one of the edges as a reference as seen
>here: http://www.christoph-roell.de/dvb/vergl.html)
>
>O = via, | = track without via
>
>O||O|||O||O|||O  O| GOLD
>R    G    B

Many thanks by the info! and your latter pictures.

>So R and B are available at vias, while G is only available at a track
>between two other tracks. Soldering connections for R and B should
>be possible, but a making a connection for G will be very difficult.
>I don't know if this is of any practical use, but if I had a rev 2.1
>card only I would give it a try ...

On addition to tap the RGB signals on the BGA mounting_PCB tracks or my
crazy idea of "in the balls" (drilling a hole, with big precision, on the
board PCB), it could be another way that already I commented:

The AV7111x is having four Bias resistors (BIAS_RC, BIAS_GY, BIAS_B &
BIAS_COMP) and four Compensation capacitors (COMP[3:0]) wich are wired
and mounted on all boards that I've seen (included the v2.1 and v2.2 that
don't have RGB outputs availables!).

Could be that this Bias and Compensation components have some usefull
signals to recover a "normalized RGB", perhaps adding a few more parts
(as OpAmps, etc). For now, I don't have had the need to investigate it 
nor the time :-(, but it also could be investigated on a <v2.1 card, with 
the good point that then one have the propper outputs to compare with the
"reconstructed" ones (with an osciloscope and a "colours bars" test picture) .

@Sergei:

Depending of the *internal* [1] damage on your AV7111x, the previous 
paragraph could be of use for your card with more chances of sucess as
it would be enought to recover a digital SYNC signal (from the BIAS_COMP
or COMPn signals) and use the RGB outputs with it.

[1] IMO, there is not any external protection diode on this cards.


Best regards:
--
Roberto Deza Asensio
Universidad de Navarra
Data Procesing Center
rdeza@unav.es


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