Re: [SI-LIST] : Dear XTAL expert...

Sandy Taylor (
Wed, 13 Oct 1999 13:17:20 -0700

> Had someone measured the signal amplitude on both sides of the
> XTAL. I always thought it to be VDD, but was told its not so on both
> sides. Whats the trues...?
> Is there any benefit/limit from connecting the internal shmit
> buffer before or after the inverting amplifier..?


A typical crystal oscillator looks like a weak inverter with a crystal
across it and a capacitor to ground at each end. The capacitors
are part of a series RLC network. A very high value resistor from
the input to the output biases the circuit in the high gain region.

The voltages present on the two pins (for an oscillator that is not overdriven)
will be will be inversely proportional to the total capacitance on each pin.
Be careful when trying to measure these voltages with a typical scope probe,
you may significantly affect the total capacitance and reduce the voltage.

Typically the capacitors are equal in value. In order to have precise frequency
control, the parasitic and load capacitance must be included.
The external caps are typically 10s of PF, much larger than the
equivalent internal series cap of the crystal, but changes in the total
external capacitances will still shift the frequency.

But the capacitors might deliberately be skewed to alter the start up gain of the
oscillator. If the input capacitor were large compared to the output capacitor
(with parasitics), then the input voltage swing would be smaller than the output
voltage swing.

Subsequent gain stages are typically connected to the output node.

Sandy Taylor
CMOS Solutions

(If you are having trouble building an oscillator, try building an amplifier instead)

**** To unsubscribe from si-list: send e-mail to In the BODY of message put: UNSUBSCRIBE si-list, for more help, put HELP. si-list archives are accessible at ****