|
|||||||||||||||||||
|
advertisement |
|
|
Pierce-gate oscillator crystal load calculation Jul 1, 2004 12:00 PM By Ramon Cerda
The Pierce-gate oscillator of Figure 1 is well recognized by most designers, but few understand how to specify the crystal correctly. The crystal used in the topology of Figure 1 can be either a fundamental AT-CUT or BT-CUT. A BT-CUT crystal has poor frequency stability over temperature compared to an AT-CUT. This topology uses a parallel crystal and not a series crystal. When a parallel crystal is specified, the crystal manufacturer will also require that you specify a load capacitance. To understand load capacitance, think of a series LC circuit where the crystal is the L and the load capacitance is the C. The resonance frequency of the LC circuit will vary as a function of L and C. But in the crystal case, the L is fixed (temperature not being a parameter). The parameter on the crystal data sheet that is controlled by the load capacitance is the tolerance or calibration of the center frequency at 25°C. If the oscillator circuit is not designed to match the load capacitance value, then the center frequency will not be within the tolerance limits of the data sheet. Interestingly enough, a so-called parallel crystal requires its capacitive load effectively be in series with its terminals. So what load is your Pierce-gate oscillator presenting to the crystal? A simple calculation illustrated with Figure 2 will tell you. C Example: Let C C Therefore, C Select C The most important fact in Figure 2 that most designers neglect is the internal input and output capacitance of the inverter gate. These are significant in value compared to the external (C Now that you know how to calculate the load capacitance the circuit presents to the crystal, what load capacitance should you choose? Before answering this question, you need to know the sensitivity of the crystal center frequency vs. load capacitance. This is known as the trim sensitivity S and is given by: in ppm/pF where C From the trim sensitivity equation you can see that the smaller you make C The C ABOUT THE AUTHOR
Ramon Cerda is director of engineering at Crystek Crystals Corp., Fort Myers, Fla. He can be reached at rcerda@crystek.com.
|
|
||||||||||||||||
| Back to Top |