Re: Fwd: Compro S300 - ZL10313

[Matthias Schwarzott <zzam@gentoo.org>]

On Samstag, 9. Januar 2010, JD Louw wrote:
> On Wed, 2010-01-06 at 22:17 +0200, Theunis Potgieter wrote:
> > 2010/1/2 JD Louw <jd.louw@mweb.co.za>:
> > > On Sat, 2010-01-02 at 09:39 +0200, Theunis Potgieter wrote:
> > >> 2010/1/1 JD Louw <jd.louw@mweb.co.za>:
> > >> > On Tue, 2009-12-29 at 23:23 +0200, Theunis Potgieter wrote:
> > >> >> Hi mailing list,
> > >> >>
> > >> >> I have a problem with my Compro S300 pci card under Linux 2.6.32.
> > >> >>
> > >> >> I cannot tune with this card and STR/SNRA is very bad compared to
> > >> >> my Technisat SkyStar 2 pci card, connected to the same dish.
> > >> >>
> > >> >> I have this card and are willing to run tests, tested drivers etc
> > >> >> to make this work.
> > >> >>
> > >> >> I currently load the module saa7134 with options: card=169
> > >> >>
> > >> >> I enabled some debug parameters on the saa7134, not sure what else
> > >> >> I should enable. Please find my dmesg log attached.
> > >> >>
> > >> >> lsmod shows :
> > >> >>
> > >> >> # lsmod
> > >> >> Module                  Size  Used by
> > >> >> zl10039                 6268  2
> > >> >> mt312                  12048  2
> > >> >> saa7134_dvb            41549  11
> > >> >> saa7134               195664  1 saa7134_dvb
> > >> >> nfsd                  416819  11
> > >> >> videobuf_dvb            8187  1 saa7134_dvb
> > >> >> dvb_core              148140  1 videobuf_dvb
> > >> >> ir_common              40625  1 saa7134
> > >> >> v4l2_common            21544  1 saa7134
> > >> >> videodev               58341  2 saa7134,v4l2_common
> > >> >> v4l1_compat            24473  1 videodev
> > >> >> videobuf_dma_sg        17830  2 saa7134_dvb,saa7134
> > >> >> videobuf_core          26534  3
> > >> >> saa7134,videobuf_dvb,videobuf_dma_sg tveeprom               12550 
> > >> >> 1 saa7134
> > >> >> thermal                20547  0
> > >> >> processor              54638  1
> > >> >>
> > >> >> # uname -a
> > >> >> Linux vbox 2.6.32-gentoo #4 Sat Dec 19 00:54:19 SAST 2009 i686
> > >> >> Pentium III (Coppermine) GenuineIntel GNU/Linux
> > >> >>
> > >> >> Thanks,
> > >> >> Theunis
> > >> >
> > >> > Hi,
> > >> >
> > >> > It's probably the GPIO settings that are wrong for your SAA7133
> > >> > based card revision. See
> > >> > http://osdir.com/ml/linux-media/2009-06/msg01256.html for an
> > >> > explanation. For quick confirmation check if you have 12V - 20V DC
> > >> > going to your LNB. The relevant lines of code is in
> > >> > ~/v4l-dvb/linux/drivers/media/video/saa7134/saa7134-cards.c:
> > >> >
> > >> > case SAA7134_BOARD_VIDEOMATE_S350:
> > >> > dev->has_remote = SAA7134_REMOTE_GPIO;
> > >> > saa_andorl(SAA7134_GPIO_GPMODE0 >> 2,   0x00008000, 0x00008000);
> > >> > saa_andorl(SAA7134_GPIO_GPSTATUS0 >> 2, 0x00008000, 0x00008000);
> > >> > break;
> > >>
> > >> Hi thanks for the hint. I changed it to the following:
> > >>
> > >>  case SAA7134_BOARD_VIDEOMATE_S350:
> > >>  dev->has_remote = SAA7134_REMOTE_GPIO;
> > >>  saa_andorl(SAA7134_GPIO_GPMODE0 >> 2,   0x0000c000, 0x0000c000);
> > >>  saa_andorl(SAA7134_GPIO_GPSTATUS0 >> 2, 0x0000c000, 0x0000c000);
> > >>  break;
> > >>
> > >> I now get the same SNR as on my skystar2 card, signal is still
> > >> indicating 17% where as the skystar2 would show 68%. At least I'm
> > >> getting a LOCK on channels :)
> > >>
> > >> Thanks!
> > >>
> > >> > Looking at your log, at least the demodulator and tuner is
> > >> > responding correctly. You can see this by looking at the i2c traffic
> > >> > addressed to 0x1c (demodulator) and 0xc0 (tuner). Attached is a
> > >> > dmesg trace from my working SAA7130 based card.
> > >> >
> > >> > Regards
> > >> > JD
> > >
> > > Hi,
> > >
> > > Just to clarify, can you now watch channels?
> > >
> > > At the moment the signal strength measurement is a bit whacked, so
> > > don't worry too much about it. I also get the 75%/17% figures you
> > > mentioned when tuning to strong signals. The figure is simply reported
> > > wrongly: even weaker signals should tune fine. If you want you can have
> > > a look in ~/v4l-dvb/linux/drivers/media/dvb/frontends/mt312.c at
> > > mt312_read_signal_strength().
> > >
> > > Also, if you have a multimeter handy, can you confirm that the
> > > 0x0000c000 GPIO fix enables LNB voltage? I'd like to issue a patch for
> > > this. I've already tested this on my older card with no ill effect.
> >
> > This is what happened when I started vdr.
> >
> > Vertical gave a Volt reading between 13.9 and 14.1, Horizontal Gave
> > 19.4 ~ 19.5. When I stopped vdr, the Voltage went back to 14V. I
> > thought that it would read 0V. What is suppose to happen?
> >
> > Theunis
> >
> > > Regards
> > > JD
> 
> Hi,
> 
> The newer revision cards should be able to shut down LNB power when the
> card is closed. This is what the Windows driver does; not yet
> implemented in Linux.

Do you know how this is done hardware-wise? Is this a gpio connected circuit?

If yes, I think it can be enabled in software by saving original pointer to 
set_voltage and overwriting it by some routine switching gpio and calling 
original function.

Regards
Matthias