SiTime now has a low-frequency oscillator, the SiT8503, that supplies clock signals from 200 kHz to 1 MHz. Like all our oscillators, the 8503 comes in a range of packages and is configurable for frequency, voltage, and enable pin function. This is the first and presently the only MEMS oscillator available in this frequency range. In an industry where higher and higher frequencies are vital, it might seem surprising that a low frequency oscillator is important. But it is very important, particularly for audio equipment. Clock quality is vital to good audio reproduction, and high-end audio companies select their oscillators very carefully.
One of our customers, a large and well respected audio company, declared the 8503 to be a significant improvement over their existing options. Their expert listeners told us that the sound reproduction in test systems with our 8503 was “substantially better” than their reference systems. What about low frequency quartz oscillators? They tend to be large, expensive, and difficult to manufacture.
The reason is because standard AT-cut quartz crystals cannot easily be built for low frequencies. To move the frequency down with these, one needs to grind a thicker blank, and that requires a larger blank to keep the resonant energy away from the edges where it causes problems. A larger blank requires a larger package, and all that adds up to an expensive and bulky part. Okay, then what about running a higher frequency crystal with an electronic divider in the circuit? Good idea, but most quartz companies don’t think that way. As surprising as it may seem, they are experts at grinding quartz and they tend not to select electronic solutions. Instead they tend to develop specially shaped crystals for low frequencies. And while that works, it is still not cheap and it is definitely not versatile since each frequency requires individual engineering. SiTime, on the other hand, is an analog semiconductor company and we generate our low output frequencies by dividing down higher internal frequencies. We solve the problem with transistors, not with fancy grinding.
This is an illustration of why SiTime is transforming the industry – we think in new and modern ways. It seems like a small thing, how one approaches a technical challenge, but really it is a very big thing. How we think defines what we are. At SiTime we think like semiconductor people. We live under Moore's Law. What was impossible last year becomes only difficult this year. What is difficult this year becomes commonplace next year.