On Friday 11 July 2003 11:24, Gregor Lawatscheck wrote: > At 06:24 11/07/2003, you wrote: > >On Friday 11 July 2003 01:21, Gregor Lawatscheck wrote: > > > At 22:20 10/07/2003, you wrote: > > > > > Either your revision of the card makes a difference or you're > > > > > in the lucky position that somehow vdr or the driver just > > > > > never touches the tuning part of the dvb-s. > > > >Add a printk to SetPIDs() (av7110.c) and verify that the pids are > > set to zero when there is no signal. av7110_before_after_tune() > > calls SetPIDs(0,0,0,0,0) if the frontend looses signal. > >AFAIK this code was added to avoid arm crashes... > > OK, thanks I'll check that once I have the box up running again.
I assume that you have already disabled - EPG scan timeout and - Set system time in EPG setup, right? Just to simplify things. > > > > > What revision is you full dvb-s card? I've got a 1.3 here. > > > > > Which version of vdr are you using? How's your EPG setup for > > > > > vdr - standard or altered so it doesn't scan at night? > > > > > > > >I think the DVB-s is a 2.1 - no J2 connector > > > > > > Most interesting. I'm getting a 1.6 from ebay with broken tuner > > > soon. Let's see whether there's any difference there. It it even > > > feasible to assume it could be the revision of the card that > > > makes a difference here? > > > >The 1.3. has a different frontend chip. > >I tested my Rev. 2.1 Nexus (stv0299-based) without sat signal: It > > did not crash within several hours. How long does it take in your > > setup until the arm crashes? > > It takes around 4-6 hours. Before that the load goes to around 40-50% > if tuned to a sat channel without having signal. It only recovers on > reloading the drivers. The tuner of my 1.3 says BSRV2-301A. Ok, I'll check again tonight for a longer period. FYI, I'm using this setup: - AMD K6-II 400 MHz - linux kernel 2.4.21 from kernel.org - single DVB-S Nexus 2.1 - DVB driver CVS 2003-06-28 - vdr-1.2.1 + remote control plugin - no other plugins, no patches I simply removed the sat cable from the box. Then the [kdvb-fe-0:0] frontend process consumes approx. 30% cpu while it tries to retune. AFAIK this is normal behaviour. Then I played some recordings. Nothing happened. > So it looks like it could be the revision and together with Andy > Carter there are now two people who don't appear have this problem > with a Rev 2.1. Good news for everyone thinking about using a full > featured card without a signal for decoding. The only thing a Rev 2.1 > is lacking is RGB output, right? Yes. > I suppose a lot of people don't intend to use it anyway? Maybe. RGB is a nice feature if you have a high-quality TV set. AFAIK the Rev. 1.6 has the same frontend as the 2.1 _and_ it has the J2 jumper block.. Anyway, I think it's possible to fix the problem when the problem has been identified. Oliver -- Info: To unsubscribe send a mail to [EMAIL PROTECTED] with "unsubscribe linux-dvb" as subject.
