In an effort to thermally stabilize the output RF Frequency of the Propeller that I use for WSPR, QRSS, and OPERA Beacons, I attached a external 100MHz +/-50ppm/C Master Clock Chip in place of the vendor supplied 5MHz Crystal, see previous post.
The Propeller has options to multiply the supplied source as necessary to obtain the desired operating clock frequency. The supplied 5MHz crystal x 16 = 80MHz, for normal use. My external 100MHz oscillator is not multiplied and over-clocks the processor by 25%. It works and it is fast, but . . .
The 100MHz Clock is accurate for average Frequency, but it produces a lot of jitter. The jitter shows up in the output RF Frequency as noise. For a Beacon, which typically uses a band width of only a few Hertz, the jitter is unacceptable. The jitter appears as fuzz on the received signal.
Another Approach is Necessary
Another type of oscillator is the; Temperature Compensated Crystal Oscillator (TCXO). The spec's on the TCXO's look very good. A high precision TCXO is typically +/-1ppm/C, with some parts rated as low as +/-0.5ppm/C. These parts are typically $5 to $10. But only a few limited frequencies are available, from 10MHz to 52MHz. They would work with Propeller but the internal PLL could not be configured to multiply the frequency. That is, the Prop internal PLL (1x, 2x, 4x, 8x, or 16x) could not be used, because the Prop PLL has an input limit of 4 to 8MHz.
So, I am looking for a low cost, high precision ~1ppm/C oscillator
- 4 to 8MHz TCXO (to be used with Prop internal PLL)
- Or, a 80MHz to 100MHz TCXO used un-multiplied in direct connection
Another type of oscillator is the; Oven Controlled Crystal Oscillators (OCXO), they are available with much better spec's (e.g, +/-0.005ppm), but they are very expensive (and large), > $150.
So far, a suitable part has NOT been found, . . . the search goes on :-)