On Thu, 10 Oct 2002, Stane Gruden wrote: > > from message: Partial solution for 'no sound problem' on flyvideo 3000 > > Of course, I supposed that this would work only for those who can > switch on the sound of the tv card by launching (and then > exiting) 'radio' (some people reported this, I tried it and it > worked for me). Apparently there are some slight differences on > the cards we have - or it is the other hardware of pc that makes > the difference, I do not know... For some the card works fully, > for some turning on the radio helps, for others there is no sound > at all. > > There is definitively a need for finding the real cause for this > problem and fix the drivers. I do not think Gerd has a possibility > to do it because his card works. I would try to do it, but I > need the doc. Gerd, Is it available...? > I don't think the saa7134 docs is available w/o a NDA (and I don't have it). In any case, it's more of a FV3K issue than a saa7134 issue at this point IMHO. After following this thread for the past few months, I'm beginning to wonder if there's different revisions of the FV3K board, and some differences between the FV3K-noradio and FV3K+FM boards. The reason that LINE2 works to enable the sound is (as I understand it) due to the presence of an external HEF4052BT analog mux on the FV3K board. This is also present in the FV2K-noradio (which I have), and is used for audio path switching (limitations of the saa7130 chip on the FV2K). Hence the FV2K-dervied settings worked for some folks. If your FV3K board works as is (using amux = TV), then I suspect it doesn't have the analog mux. Someone really needs to compile the board revision and chip list for these different boards to be sure. As far as I can tell, the HEF4052BT is controlled by two GPIO lines (the gpio_mask has two enabled bits), giving 4 input selections. When LINE2 is enabled, the gpio lines must have the correct setting in order for the proper audio path to be selected. So if LINE2 works intermittently, you may want to define the gpio_mask variable (follow FV2K card definitions), and use the correct gpio settings (0 = television audio out, I believe). Having said all this, I think there's some possible IRQ/PCI issues inherent in the I/O code of the saa7134 module. Sometimes selecting a new setting doesn't stick, and had to be performed several times before the card responds. This is just a guess since I'm not familiar with PCI bus programming. I only hacked the gpio controls part of the saa7134 module. > Another thing: is this the only mailing list on these subject, or > there exists some other more 'developer' list? This one looks > more like 'users' mailing list to mo. > I don't think there's a separate list. Everyone who's involved with the v4l2 modules seem to lurk here (even if they don't speak out). > Stane. > > > > ---- Wan Tat Chee (Lecturer) School of Computer Science, Univ. Science Malaysia, 11800 Minden, Penang, Malaysia. Ofc Ph: +604 657-7888 x 3617 Internet: tcwan@xxxxxxxxx Web: http://nrg.cs.usm.my/~tcwan GPG Key : http://nrg.cs.usm.my/~tcwan/tcw_gpg.asc F'print : FB0F CED7 85A5 ECF9 DEF0 50E8 A550 A0D2 8638 B1EB
