On Tue, 2011-04-19 at 13:12 -0400, Eric B Munson wrote: > On Tue, 19 Apr 2011, Andy Walls wrote: > > > > > Have you used v4l2-ctl or ivtv-tune to tune to the proper tv channel > > > (the driver defaults to US channel 4)? > > > > > > emunson@grover:~$ v4l2-ctl -F > > > Frequency: 0 (0.000000 MHz) > > > emunson@grover:~$ v4l2-ctl -f 259.250 > > > Frequency set to 4148 (259.250000 MHz) > > > emunson@grover:~$ v4l2-ctl -F > > > Frequency: 0 (0.000000 MHz) > > > > OK, that doesn't look good. The tda18271 tuner and/or tda8290 demod > > drivers may not be working right. I looked into this. Apparently the tda18271 tuner driver doesn't report back analog frequency setting. So reading back 0.00000 MHz is OK. > > I'll have to look into that later this week. > > > > BTW, Mike Krufky just submitted some patches that may be relevant: > > > > http://kernellabs.com/hg/~mkrufky/tda18271-fix > > > > I have applied these patches and I am still seeing the same problem (frequency > still report 0 after being set) and mplayer still closes immediately. I don't have those patches applied. I just tested my 74351 HVR-1600 again tonight, and analog channel 3 works for me. I don't know what to say.... > > > > > > > Does v4l2-ctl --log-status still show no signal present for the '843 core in the CX23418? > > > > > > Yeah, > > > [94465.349721] cx18-0 843: Video signal: not present > > > > The tuner or demod isn't tuning to a channel or getting a signal. > > > > Can you try channel 3 (61.250 MHz)? That one works for me. > > Still shows not present on channel 3. OK. Tonight was the first time I have *ever* been able to reproduce the "red screen" out of a CX23418. I am fairly sure the red screen happens because of some kernel bug or PCI bus problem that ends up corrupting the CX23418 registers responsible for configuring the CX25843 A/V core inside of the CX23418. The '843 core is responsible for digitizing the analog video, so when it stops functioning, the result is a red screen. I guess the root cause could also happen due to some thermal or power condition causes the '843 core to stop, but I think that is less likely. Anyway, when you're going to get a red screen, the '843 core will always indicate "cx18-N 843: Video signal: not present", even if there is a signal present from the analog tuner and demodulator chips. > > > > Does mplayer /dev/videoN -cache 8192 have a tv station when set to the rf analog input with v4l2-ctl? > > > > > > emunson@grover:~$ mplayer /dev/video0 -cache 8192 > > > MPlayer 1.0rc4-4.4.5 (C) 2000-2010 MPlayer Team > > > > > > Playing /dev/video0. > > > Cache fill: 0.00% (0 bytes) > > > > > > > > > Exiting... (End of file) > > > > Hmmm. I would have expected at least a black picture with snow, if not > > tuned to a channel. > > > > Does analog S-Video or Composite work? > > Unfortunately, I do not have anything I can use to test these. The card only > takes coaxial or S-Video input and I don't have any sort of S-Video cables or > converters. With new HVR-1600 from Hauppauge, an adapter for plugging a Composite RCA cable into the S-Video jack is in the box. It looks very much like the one at the end of this thread, but it is grey in color: http://www.hauppauge.co.uk/board/showthread.php?t=22115 If this is a second hand card, or one bought from e-Bay, etc. have you ever tested it in a Windows machine with the Hauppauge Windows drivers? (In other words, are you reasonably confident the card is not defective or broken.) When testing tonight, I unloaded all the drivers, performed the following commands to get verbose debugging, and saved the dmesg: # modinfo tuner # modprobe tuner debug=7 # modinfo tda8290 # modprobe tda8290 debug=7 # modinfo tda18271 # modprobe tda18271 debug=31 # modinfo cx18 # modprobe cx18 debug=255 # ivtv-tune -d /dev/video1 -c3 # v4l2-ctl -d /dev/video1 --log-status # mplayer /dev/video1 -cache 8192 I'll send you the 138 kB dmesg file off list, so you can see the messgaes generated by a functioning 73451 HVR-1600 setup. Note that the card of interest in the dmesg output was 'cx18-1'. You'll want to look to see that the TDA18271 and TDA8290 related messages look the same, to get reasonable assurance that the tuner and demodulator chip are being set up properly and repsonding. If you consistently get "cx18-N 843: Video signal: not present" in your v4l2-ctl --log-status output but the tuner chip configurations look good (i.e. closely match the debug output from my machine), then you really have a "red screen" problem versus a problem related to the analog tuner/demod chips. The "red screen" will be a tough one to track down. It's a system level issue, likely involving one or more of: 1. a kernel bug in a driver for some other hardware in your system 2. a hardware problem with your PCI chipset when the PCI bus is heavily loaded 3. a power or thermal condition that causes problems for the HVR-1600 hardware I usually start by recommending that all non-critical linux driver modules be blacklisted (or not loaded), and booting the system into run-level 3 or lower (no X windows), when trying to isolate the cause in any one system. Regards, Andy > Eric -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
