0

enter image description here

I've seen many app notes and calculations that in the above circuit, the value of C1 and C2 are the same. I haven't seen in many places where the values of C1 and C2 can be different.

Can someone tell me whether C1 and C2 can be different or should be the same? I read that C1 and C2 should be the value that the crystal manufacturer mentions as load capacitance in their datasheet.

Suppose, the crystal manufacturer mentions that the load capacitance of the crystal is 20pF, then C1 and C2 should be 20pF. Am I correct?

  • 3
    Does this answer your question? [Crystal oscillator load capacitance, again](https://electronics.stackexchange.com/questions/250608/crystal-oscillator-load-capacitance-again) – Justme May 24 '22 at 18:29
  • 4
    The series combination of C1 and C2, in parallel with the input capacitance of the IC and the parasitic capacitance of the interconnect between them, should equal the crystal's designed load capacitance. – The Photon May 24 '22 at 18:29
  • It doesn't say whether C1 C2 can be different values. Please help –  May 24 '22 at 18:42
  • Why should it be equal to load capacitance? –  May 24 '22 at 18:43
  • @ThePhoton, could you please explain why and provide an answer in simple terms? –  May 24 '22 at 18:43
  • 1
    Because that's the capacitance that the crystal "sees" in parallel with itself. – The Photon May 24 '22 at 18:53
  • 2
    Related: "[Why use asymmetrical capacitor values for crystal?](https://electronics.stackexchange.com/q/111325/101852)" and the answers that it links to. – SamGibson May 24 '22 at 19:07

1 Answers1

2

Suppose, the crystal manufacturer mentions that the load capacitance of the crystal is 20pF, then C1 and C2 should be 20pF. Am I correct?

Because the two capacitors load the crystal as series elements each individual capacitor would need to be 40 pF because, 40 pF in series with 40 pF = 20 pF.

I haven't seen in many places where the values of C1 and C2 can be different.

No it isn't mentioned very much but they can be different to a certain degree providing that together, in series, they add up to the stated xtal manufacturer's load capacitance. So, if that value is 20 pF then it could be made from

  • 40 pF and 40 pF
  • 50 pF and 33 pF
  • 60 pF and 30 pF

Etc. etc..

But, don't push it too far and, if you want my opinion, I see little reason not to make them both identical.

If you want some backdrop to how a pierce oscillator works using a crystal, an article on my basic website might help. The bottom line is that the two capacitors do subtly different jobs; one provides near 180° phase shift (C2 in your diagram) and the other (C1 in your diagram along with Rs) create an extra circa 30 (ish) ° of phase shift to ensure that the crystal and capacitors produce a perfect phase inversion at a very unambiguous frequency: -

enter image description here

Without the extra capacitor (C1 in your diagram), the crystal circuit cannot produce sufficient phase angle to produce accurate oscillations at the manufacturer's predefined frequency: -

enter image description here

The above picture is with one capacitor and clearly, there isn't enough phase shift to get a pierce oscillator to work effectively.

If you want to test this out in a simulator, use this equivalent circuit (tried and tested): -

enter image description here

Results: -

enter image description here

So, with equal 20 pF capacitors (total load = 10 pF in this example), the oscillation frequency will be 10.001374 MHz. If you imbalanced the capacitors such as with using a 35 pF and a 14 pF capacitor, the oscillation frequency could be 10.001357 MHz in one direction or 10.001392 MHz in the other direction.

The change is small (+/- 18 Hz or, +/- 0.00018 %) but, if you want accuracy, stick to using the same value capacitor or, be prepared to do bench testing to get the exact frequency you desire.

Andy aka
  • 434,556
  • 28
  • 351
  • 777