crystal wavefrom issue

[cool.man]

Hi
I have a question regarding crystal frequency waveforms.i am using pic18f252.
i use two different crystal's (4Mhz and 25Mhz). when i checked the waveforms of the crystal's the following results as the images shows.

i think waveform of 4Mhz crystal is not correct.25Mhz crystal shows the fine waveform.
My question is that, Is my crystal is damage or it is correct?

I change 4 to 5 crystal's of value 4Mhz but got the same result.

 

[betwixt]

I can't see the scale on the second one to check it's frequency but both waveforms look fine to me. What do you think is wrong with them?

Brian.

 

[cool.man]

betwixt when i scale down the oscilloscope, the waveform of the 4Mhz looks like in the picture attached, i think the waveform in blue is not correct as compare to yellow one. Does this behavior not effect the microcontroller process. the waveform in blue is connected to the OSC2 pin of the microcontroller and yellow is connected to the OSC1 pin of the microcontroller.

 

[Golden Electronics]

1. Use CLKOUT pin for frequency.
2. Check Required Load Capacitance value according to datasheet.

 

[betwixt]

The waveforms are fine. In theory, and if real life physics didn't get in the way, they would be square waves but the limited current supply in the oscillator output, 'Q' of the crystal and the loading capacitors limits the rise and fall times and gives you the rounded appearance. The oscillator must be working properly or you wouldn't get any waveform at all. Inside the PIC there are circuits to clean the waveform and then divide it to produce the lower frequencies needed for it's peripherals. So don't worry, there isn't a problem, what you see is quite normal.

Brian.

 

[FvM]

Crystal oscillator waveforms on driver side vary between almost pure sine and square wave. There's nothing about right or wrong, unless the datasheet specifies specific waveforms.

Determining parameters are oscillator inverter gain, output and crystal series impedance. In terms of emc behaviour, low level, low distortion waveforms are preferred, a overdriven crystal oscillator can easily generate spectral components up to GHz range. On the other hand, a gain reserve resulting in some clipping guarantees reliable start-up with different crystal and load capacitances.

If you want low distortion waveforms or minimal interferences, you can try with a series resistor and possibly a low-pass capacitor on the driver side, as specified by some manufacturers.

 

[jamespond]

The waveform not must be sine on xtals. Your scope cable cause more load on the signal and increase the signal malforming. If the contoller works good, I think this is not a problem.