I haven't been part of this entire thread, so maybe I'm just adding noise here. Ton's comments about my suggested dual-BBD/one-clock approach never reaching below the shorter of the two delay times is absolutely spot-on accurate.
That being said, the challenge, as I see it at least, is that some folks want TZF, while others simply want what TZF capability also makes possible, and that is a sweep to near-zero. While buffering the clock (actually, more like buffering the clock inputs on the BBD) will permit much higher clock speeds from any of the Matsushita chips, and consequently shorter minimum delay times, there ARE still limits to how fast you can clock them. What I have attempted to suggest is a way of achieving the "near-zero" capabilities of true TZF units without either a) taxing individual BBDs too much, or b) introducing heterodyning risk by means of two individually-clocked BBDs.
The downside, of course, is that while one can tolerate the needed lowpass filtering of the delay path because the dry path is untampered with, inserting a BBD into the "clean/dry" path obligates one to use some lowpass filtering for noise control and that may erode some of the sparkle. Still, it may be an epxeriment worth trying out.
