Oscillators may use crystals as precise tuned circuits. If you have a vector network analyzer (VNA) — or even some basic test equipment — you can use it to learn the parameters of a crystal. [All Electronics Channel] has the details, and you can see how in the video below.
There was a time when a VNA was an exotic piece of gear, but these days they are relatively common. Crystal parameters are important because crystals have a series resonance and a parallel resonance and they are not at the same frequency. You also may need to know how much loading capacitance you have to supply to get the crystal at the right frequency.
Sometimes, you want to pull the crystal frequency, and the parameters will help you figure that out, too. It can also help if you have a crystal specified as series in a parallel-mode oscillator or vice versa.
If you don’t have a VNA, you can use a tracking signal generator, as [Grégory] shows towards the middle of the video. The quality of a tuned circuit depends on the Q factor, and crystals have a very high Q factor.
We did something similar in 2018. The other way to pull a crystal frequency is a bit extreme.
This was a pretty cool watch from a new hardware engineer viewpoint with an electronics lab that looks like a starter version of his. I wish he had gone more in depth about how to perform the S21 measurement on a spectrum analyzer w/ tracking gen. Otherwise a great intro to crystal and how to model simple systems with just transmission parameters
You can also build a VNA using a function generator, and oscilloscope and some python to control them. See https://github.com/jp3141/Vector-Network-Analyzer/blob/master/4%20MHz%20XTAL%20with%201%20kohm%20load.png for an example of a crystal measurement.
Sorry, it’s not a VNA, it’s a dodgy Bode plotter.
For analyzing crystals, the reflection parameters don’t have much use. But yeah, a bode plotter is at most half of a typical VNA.
Who guards the guardians? Or in this case, how accurate is the clock in your VNA?
Where I worked, Agilent / Keysight gear was certified calibrated once a year. If you replicate this test with a humble nanoVNA, your accuracy is suspect. And this is before you throw in the bandwidth setting.
All this to say: adjusting the frequency of a crystal can lead you down a rabbit hole, in the name of accuracy.
Well, getting a reference frequency from GPS is possible.
And if you’re in a position where you want to do such a fine adjustment on the frequency, you probably already have some sort of reference you’re measuring against, rather than just trying to hit a magic number.