MATT NEWPORT: I've got three different clocks here. These are master clocks. This is the one that we stick in the 002. It's our current production clock, and it actually is comprised of two clocks. One handles the 44 kHz sampling rate that's multiples of 44 and 88, and then the other one handles the 48 and 96 kHz sample rates.

Within pro audio, my experience has been that a lot of people in this industry, a lot of engineers in this industry seem to prefer what we call fundamental crystals. In fact, you can find stuff from some fairly well known designers online talking about how much better fundamental-order crystals are than other crystal oscillators. We don't use fundamental-order crystals. We actually use a third-order crystal which means that the third harmonic of the fundamental is what is the signal that's actually put out by the clock, and the reason that we do that is because it has wider bandwidth, and because of the wider bandwidth, it has inherently less jitter so it gives a wider, more analog-like sound to the audio reproduction. And what we do with this is because we're running a higher-order crystal and it's a higher frequency, we have to divide it down, and that's what this portion of the circuit does. It divides the signal into and then it goes out into the digital board of the 002 or 003.

Alright. So these are crystal oscillators. They're manufactured for us. Specifically for audio frequencies, you have to get them custom made, and we order them from a company here in the U.S. who makes -- typically makes crystals for aerospace applications like NASA. They're based in Florida. And we found that by far these are the best sounding crystals. They have the lowest intrinsic jitter, which is just jitter that's inherently there in the crystal, and it's about 1 picosecond. And then, once it comes out here, we divide it in this little integrated circuit here. This section here is the power supply, and we use something called a shunt regulated supply. It has less noise than traditional power supplies that are often used in pro audio.

This is the word clock. Same basic concept. Third-order crystal oscillators that have 1 picosecond of jitter, and then they're run into a division network that divides it down to word clock rate which is 44 kHz or 88 kHz or 192, and whatever you happen to need.

There's a couple of things that play. First of all, whenever we have to divide something down, there's so much switching that's going on, so much transistor switching that we have to figure out a way to limit the noise because it will show up as a harmonic within the clock signal. So we use a couple of tricks that I don't often discuss with people, and I'm not going to discuss them here.

One of the things that we like is, well, for this particular application, we used discrete transistor-based inverter instead of one that's built on a semiconductor chip, and then we drive everything into transformer coupled output. Now, a lot of people think that a clock is just a clock is just a clock. The key to a clock's sound, especially a word clock, is the transformer, and how much current you're driving into that transformer, because what happens is the harmonics within the clock become a little exaggerated, shall we say, in that it imparts a nice mid range characteristic, so that's kind of the goal, and you'll find that in pretty much any mass-produced word clock.