Author Topic: Why is my 20Mhz output a little sawtooth'ed?  (Read 1691 times)

0 Members and 1 Guest are viewing this topic.

Offline jnzTopic starter

  • Frequent Contributor
  • **
  • Posts: 593
Why is my 20Mhz output a little sawtooth'ed?
« on: January 09, 2019, 12:57:07 am »
I have an 20Mhz crystal going to a micro's XTAL1 and XTAL2. The leads to the clock are 5mm long each. There is a 8pF cap to ground at each crystal terminal. Four layer board with power and ground internal layers under the clock (forgot I might want to remove one or both). I have a bridged 0402 resistor on the XTAL2 line that I could cut and put a resistor on. The scope shows a peak to peak at about 1.2V with a 5V VCC. The Specs for the clock show 4pF load, but I this is a 6pF or 8pF model. The load caps are 8pF.

What causes the clock waveform to sawtooth a bit and how do I fix it?

The scope pic is measuring on the XTAL2 line. I thought I could try a resistor or different load caps, but wanted to ask first. Thoughts?
 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 11905
  • Country: us
    • Personal site
Re: Why is my 20Mhz output a little sawtooth'ed?
« Reply #1 on: January 09, 2019, 01:14:43 am »
This is normal for a crystal. In fact you want it to be close to sine.
Alex
 

Offline jnzTopic starter

  • Frequent Contributor
  • **
  • Posts: 593
Re: Why is my 20Mhz output a little sawtooth'ed?
« Reply #2 on: January 09, 2019, 02:12:29 am »
Oh yea yea yea, I know I want it to be close to sin. I'm just interested in how much it's skewd/shifted... sawtoothed... I don't know the right work - away from sin.

The dev kit for my micro had a 500ohm (iirc) on the XTAL2 line. I'm not sure if I could correct this with that, or if this is a capacitor loading "problem".
« Last Edit: January 09, 2019, 02:14:20 am by jnz »
 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 11905
  • Country: us
    • Personal site
Re: Why is my 20Mhz output a little sawtooth'ed?
« Reply #3 on: January 09, 2019, 02:14:38 am »
In that case I don't know what to tell you. This looks exactly like all the Xtal waveforms I've seen. They all are slightly misshaped.

Why do you need it to be sine?
Alex
 

Offline jnzTopic starter

  • Frequent Contributor
  • **
  • Posts: 593
Re: Why is my 20Mhz output a little sawtooth'ed?
« Reply #4 on: January 09, 2019, 05:20:23 am »
In that case I don't know what to tell you. This looks exactly like all the Xtal waveforms I've seen. They all are slightly misshaped.

Why do you need it to be sine?

I figured slightly misshapen should/could be corrected?
 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 11905
  • Country: us
    • Personal site
Re: Why is my 20Mhz output a little sawtooth'ed?
« Reply #5 on: January 09, 2019, 05:21:33 am »
There is no need to correct it if you want it for clock generation reasons. It looks perfectly normal.
Alex
 
The following users thanked this post: jnz

Offline chrisl

  • Regular Contributor
  • *
  • Posts: 90
  • Country: us
Re: Why is my 20Mhz output a little sawtooth'ed?
« Reply #6 on: January 09, 2019, 06:01:43 am »
In that case I don't know what to tell you. This looks exactly like all the Xtal waveforms I've seen. They all are slightly misshaped.

Why do you need it to be sine?

I figured slightly misshapen should/could be corrected?

It is a  perfectly normal waveform and needs no fix at all.
 
The following users thanked this post: jnz

Offline Bud

  • Super Contributor
  • ***
  • Posts: 7276
  • Country: ca
Re: Why is my 20Mhz output a little sawtooth'ed?
« Reply #7 on: January 09, 2019, 06:37:43 am »
Distorted waveform is because of the harmonics, which may be because the resonator is overdriven. The datasheet you posted says the drive level should be between 10 and 100 uW. You measured 1.2V of the drive voltage. The datasheet also says a 20MHz resonator equivalent resistance is around 50 Ohm. That gives 1.2^2/50= 28.8mW which is two orders of magnitude higher than recommended drive level. Try inserting say a 10 Ohm resistor in series and measure voltage across it, calculate the current and then calculate the drive level. I am not sure how to reduce the drive level, perhaps a attenuator may need to be consructed. The 500 Ohm resistor seems to be in the kit to serve that purpose, but i'd think a series resistor will reduce the resonator Q. You still can try it. A better way is perhaps to replace the resonator with one with appropriate drive level. But as other said the micro will work as it is. There may be technically a risk of damaging the resonator with time if it is overdriven too much.
If you will be measuring RF voltage across the series resistor with an oscilloscope make sure your circuit ground is floating, i.e. not connected to the scope ground, or use a 2-channel oscilloscope with math to emulate a differential probe.
Facebook-free life and Rigol-free shack.
 
The following users thanked this post: jnz

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Why is my 20Mhz output a little sawtooth'ed?
« Reply #8 on: January 09, 2019, 06:49:15 am »
Hi jnz

Also, what you see on the scope is not the actual signal, for two reasons:
  • The capacitance/inductance of the scope probe will affect the waveform (20MHz is a relatively high frequency).
  • Unless you have been extremely particular about the earth connection of the scope probe, the scope will not display just the waveform you are interested in. This is just a fact of life.
As the other members say though, the waveform you show looks pretty normal. :)
« Last Edit: January 09, 2019, 06:54:56 am by spec »
 
The following users thanked this post: jnz


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf