1
\$\begingroup\$

What is the correct way to connect a crystal or ceramic resonator to a network analyzer so as to be able to measure the series and parallel resonate frequencies?

Should non-attenuated probes be used? Assuming that the input and output of the analyzer is 50ohms, can the crystal be connected as follows?

schematic

simulate this circuit – Schematic created using CircuitLab

If the Analyzer has a high-impedance input or if I were to manually plot this with a signal generator and an oscilloscope, should there be a 50ohm resistor between the right side of the crystal and ground?

The series and parallel resonant frequencies will be used to calculate the motional perimeters (Lm and Cm) of the crystal.

This appears to be a new question. I searched here and Google.

\$\endgroup\$
7
  • \$\begingroup\$ Do you know the equivalent circuit at least? It's just a two port with Cs,Cp,Ls,Rs \$\endgroup\$ Oct 26, 2016 at 16:58
  • \$\begingroup\$ Yes, the equivalent circuit of a crystal is just an RLC circuit with the capacitance of the crystal plates. My question then is, do I need to match the impedance of that circuit to 50ohms in order to accurately measure it's properties? \$\endgroup\$ Oct 26, 2016 at 17:04
  • \$\begingroup\$ You need to take the maximum drive level into account or you could damage the crystal. \$\endgroup\$ Oct 26, 2016 at 17:07
  • \$\begingroup\$ You leave the crystal alone if you want to measure its properties. If you mess about trying to match it, you will hide its properties behind the transformation and the uncertainty of the matching networks. A 50ohm analyser will measure them with respect to 50 ohms. If that doesn't give you the dynamic range you need for that particular crystal, then perhaps use a high impedance scope shunted by (say) 1k or 10k to give it a well defined input impedance. \$\endgroup\$
    – Neil_UK
    Oct 26, 2016 at 17:37
  • \$\begingroup\$ It depends how accurate you need to measure Rs, Cs, Cp, Ls or whether it is more important to measure \$R_s, f_s, f_p error\$ or even better AT cut slope fp(T) to bin for TCXO \$\endgroup\$ Oct 26, 2016 at 17:53

3 Answers 3

2
\$\begingroup\$

schematic

simulate this circuit – Schematic created using CircuitLab

Follow OEM Mfg test specs for power levels and load Cp

N.B. Define purpose of Tests and test acceptance criteria.

or more details read IEC 444 on standard \$\pi\$ network.

For google search consider most important keywords... parallel series resonance crystal ATE network analyzer frequency test Rs etc... All your answers will be out there. I learnt from MIL-STD HDBK on Crystals ( entire book ) circa'75

https://www.google.ca/search?q=crystal+motion+capacitance+test+jig&ie=utf-8&oe=utf-8&gws_rd=cr&ei=duEQWOORI8zyUpyKvdAE

\$\endgroup\$
0
\$\begingroup\$

schematic

simulate this circuit – Schematic created using CircuitLab Alternative method using scope

\$\endgroup\$
3
  • \$\begingroup\$ In last above picture, is there an error value 5 Mohm ? \$\endgroup\$
    – Antonio51
    Jul 1, 2021 at 13:01
  • \$\begingroup\$ @Antonio51 For testing a high Q equivalent cct. I chose 5M to 50 which would attenuate without much loading for a passive transfer function. \$\endgroup\$ Jul 1, 2021 at 22:32
  • \$\begingroup\$ Ok. Thanks for this info. I will try making a simulation. I used for years a Q-meter (not very precise, but usefull, difficult use with XTAL), then a network analyser ... but always with 50 Ohm load only as first picture. Will try also with 5Meg. \$\endgroup\$
    – Antonio51
    Jul 2, 2021 at 8:36
0
\$\begingroup\$

I think this paper summarizes several methods to measure Xtal parameters. Five different test fixtures were used, six different measuring techniques, and the results were compared.

\$\endgroup\$
1

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.