Eric Ellison wrote: [ everything trimmed ]
It sounds like there are several groups doing a lot of interesting things with FGPAs that extend way beyond just frequency stability. I think that's really great, and thank Gerald for creating such a great base for creative people to play.
From my (and TAPR's) perspective, we want to do a Reflock-based design because it will have a broader use than just the SDR-1000. For example, with a 64MHz VCXO, it can work with the Ettus Research USRP software radio. I can also see it serving as an external clock for a sound card to eliminate that source of frequency uncertainty. So, I don't see any conflict at all between what we're doing and the other approaches.
The one SDR1k specific item I'd like to see, whether TAPR produces it or Gerald does, is a "civilized" board that will plug into the reference oscillator socket to allow a better interface to an external signal. I've started looking into that and we'll work with Gerald to figure out how best to offer that. It's mainly a matter of finding a buffer/level converter chip that will take a single-ended input and convert it to a differential signal at the right level for the DDS chip (suggestions welcome; I've started researching the chipmaker sites, but haven't found an obvious answer yet).
73, John
