[vdr] Re: Problems while switching channels -> seems to be a diseqc problem

[Oliver Endriss <o.endriss@gmx.de>]

Reinhard Walter Buchner wrote:

> > I'm using a full DiSEqc sequence without any voltage/toneburst
> > commands. Even with this setup switching is not very reliable,
> > unless I add some retransmissions. It's just a work-around.
>
> Strange, while I don't have a switch, I do have a motor. After
> my workaround and later Klaus's new tuning routine, it runs
> w/o a hitch at all. Even rapid switching of channels (lots of
> DiSEqc commands on the line) between satellites doesn't
> throw the motor or the LNB reception off (my card is a Rev

Lucky you, DVB-S and motor speak the same dialect. :-)

> 1.6). The voltage on my card is slightly off spec (12.5V and
> 17.1V), but it works. One of my mods to the VDR code is the
> possibilty to add 0.5V to the LNB line.

Normally this is used to compensate for long cables.
In my case this is rather pointless because I don't switch the 
voltage at all. In full DiSEqC mode it should be ignored by the 
switch anyway.

> > *sometimes* the *first* DiSEqC command sequence [...] doesn't
> >work. So I added a dummy DiSEqC command at the begining of
> >the line and now it seem to work...
>
> Interesting. Did you try to increase the Wtime before and after
> sending the DiSEqC command?

Of course, but it didn't help. It just makes tuning slower...
Since the Wnn commands are processed by vdr, not the firmware,
this was not a big surprise. :-(

Meanwhile I have removed all delays (Wnn). Using some retransmissions,
I have a very fast and perfectly working solution.

> >With the full-featured cards this is (unfortunately) done by
> > the AV7110 firmware...
>
> Yes, true. It seems (just like CAM support ;-() to be a weak
> point in the FW. However, it is most likely a timing problem
> with DiSEqC command sequence which prevents Convergence
> from implementing it directly into the driver code (and taking
> it out of the FW).

A much better solution for STV0299 based cards would be to use
the DiSEqC capabilities of the STV0299 chip. It has a complete
DiSEqC sequencer, so timing is not an issue at all. Ironically, 
this can be done by adding two or three components to the 2.1 PCB.

Before I found the retransmission workaround I did this with the 
old driver: I patched the driver to send the DiSEqC comamnds over 
the I2C bus to the STV0299. It worked perfectly.

Oliver



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