Hi, Andrea Venturi wrote: >>The DiB7000 was at first just used for DVB-H and currently DiBcom is >>also using this demod for DVB-T devices (because power consumption is >>lower with this design and there is also the DiB7700 and DiB7710 which >>is demod and USB/PCI bridge in one chip - devices are really small). > > so this kind of DVB-T card are not actually supported by the linux driver? Yes. Afaik, the dib7000 is not register compatible with the dib3000, so there has to be another new demod-driver for that. > anyway, are there DVB-T devices ALREADY on the market using DiB7000, in > the enanced 4K version, or they are just the old 2k-8k only version? Just for reference designs for OEMs I think. DVB-T won't use 4k FFT. As far as I can see from now, as soon as the first dib7000-device will hit the market, a linux-driver for that will be ready... - Problem then will be if the silicon-tuner-manufacturer wants to reveal information about that (see current problem with Microtune and their MT2060). > do you have a dump of a complete TS of such a "DVB-H like" transmission? No, sorry. >>The main differences/additions (from memory): >> >>- every data transmitted with DVB-H is encapsulated in IPv6 MPE sections >>- a sophisticated RS FEC is used for error correction of the MPE-sections >>- 4k-mode >> > > yes, i read somewhere, and there's the sychronization stuff etc.., but i > was oversimplyfing just to give an idea Yeah, right, I forgot the time-slicing, which is used to save battery on devices. time-slicing was used for the test-signal in Berlin - so every 2 seconds I saw bursts of data coming in on different PIDs. Patrick.
