Seeing crystall waveform with oscilloscope!

Go To Last Post
7 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi,

I recently bought a Agilent Oscilloscope and i wanted to see the crystal output on a AVR Board! The only thing i saw was a sinus-like waveform and the frequency i got was ~90MHz!!!

Shouldn't i see a square waveform of e.g. 8MHz (when i have a 8MHz crystal)?

I wanted to compare the UART outputs together with the clock rising/falling edge.

Thanks,

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

What was the amplitude of 90MHz sine?

Weird things can happen if you just poke something to crystal pins - but I would never have guessed 90MHz sine.

If your scope is not broken, use a 10x attennuating probe, and you set the timebase to something like 1us or 10us / div and voltage base to something like 1V or 100mV / div, you should see sine or square wave at the crystal frequency.

Whether it is sine or square depends on the AVR fuses, if the output is normal or full-swing.

Then again, you may see something random if the fuses are set to internal RC oscillator, so the crystal does not operate.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

You should be able to see a reasonable waveform on XTAL2 using a X10 probe.

I cannot see the point of inspecting the UART outputs. If you program the AVR correctly, it just works.

David.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Is the waveform clean and stable on the scope? In other words, is it triggered correctly on the scope? If you see a "jitter" in the waveform on the scope, it may not be triggered or stable in which case, the measured frequency will be wrong. Try changing the trigger level on the scope.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Which model of AVR? If it has a CKOUT fuse/pin set/monitor that and you'll see a nice, full-swing square wave at F_CPU.

Alternatively program the UART for a known baud rate then transmit 'U' repeatedly at 8N1. For example you should see 104us bits for 9600 baud if F_CPU is what you think it is (when programming UBRR)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Thanks to all.

I will be back at home in about 1hr. I will take some images from the scope in order to see the waveform!

I will check the fuses on will attach them to here..

Thanks a lot
-G

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
I wanted to compare the UART outputs together with the clock rising/falling edge.

If you are doing this in order to base a design on the observed phase relationship - don't. Atmel specifically say that there is no guaranteed timing relationship between the oscillator and the edges of peripheral transitions.

Roger