crystal oscillator circuit

[mdanh2002]

I have 4MHz, 8Mhz, 20MHz and 40MHz crystals (the 2-pin type, not the 4-pin crystal oscillator that doesn't need any external parts to create square wave). What is the quickest and easiest circuit to create square wave from these crystals? I just want to test my oscilloscope at various input frequencies.

 

[ericgibbs]

I have 4MHz, 8Mhz, 20MHz and 40MHz crystals (the 2-pin type, not the 4-pin crystal oscillator that doesn't need any external parts to create square wave). What is the quickest and easiest circuit to create square wave from these crystals? I just want to test my oscilloscope at various input frequencies.

hi,
Look at this pdf.

 

[JimB]

One thing to bear in mind with the 40Mhz crystal is that it will probably be a third overtone type and as such will not run at 40Mhz in the circuits proposed by Eric.

If it is a third overtone crystal, it will actually oscillate at 40/3 = 13.33 Mhz.

Also, what are you expecting to see on your oscilloscope?
Depending on the scope bandwidth, the 4 and 8Mhz square waves will look like reasonable square waves.
The 20Mhz will probably look a bit rounded at the corners, and the 40Mhz (if you get 40Mhz) will look quite rounded, almost a distorted sinewave unless the scope has a bandwidth up above 200Mhz.

JimB

 

[mdanh2002]

I have 2 analog oscilloscopes, a 20MHz and a 100Mhz, both by INSTEK. Was thinking to buy a low-end RIGOL 50Mhz digital oscilloscope for my microcontroller work. I have heard for a long time that oscilloscope bandwidth are rated at -3db point (the frequency at which the amplitude is already distorted by around 71%) and want to experiment this. With my current function generator generating up to 2 MHz square wave, both oscilloscopes display just fine.

Also, some of the circuits in Eric's PDF file show a 10M resistor (Rf). The highest I have is around 2M. Do you know where I can find one, or can a lower value be used?

 

[ericgibbs]

One thing to bear in mind with the 40Mhz crystal is that it will probably be a third overtone type and as such will not run at 40Mhz in the circuits proposed by Eric.
JimB

hi Jim,
There is an overtone oscillator circuit on page #3 of the pdf.

hi mdanh,
You could try a 2meg resistor, or say two or more 2meg in series, as its only a test fixture.

 

[Diver300]

You can use any resistor over about 100 kohm as the feedback resistor in any CMOS oscillator circuit with AT-cut crystals. The 4 - 40 MHz crystals will be AT-cut.

In fact, you may be able to get the 40 MHz to run at its correct 3rd overtone frequency by simply having a feedback resistor that it about 3 kohm.

The only other type of crystal is the watch crystal, at 32.768 kHz, and for those the larger resistors are needed.

 

[JimB]

hi Jim,
There is an overtone oscillator circuit on page #3 of the pdf.

Damn!
Eric Gibbs is right!

That is what you get by:
1 looking at something
2 going away and thinking about it
3 having a (good?) idea
4 writing up the good idea without checking the basic facts!

JimB

 

[MrAl]

I have 4MHz, 8Mhz, 20MHz and 40MHz crystals (the 2-pin type, not the 4-pin crystal oscillator that doesn't need any external parts to create square wave). What is the quickest and easiest circuit to create square wave from these crystals? I just want to test my oscilloscope at various input frequencies.

Hello there,

Another way to do this is to build just one oscillator, or simpler BUY one oscillator, say 40MHz, and use logic devices such as binary counters to generate the other frequencies. With one counter you can get frequencies (in MHz) of 20,10,5, 2.5, and with an 8 bit counter even more than that. With a little rewiring you can generate other frequencies too, for example dividing by 5 will get you to 8MHz, and that divides down to 4,2,1,0.5 MHz. So you can see you can get a wide variety of outputs, and all will be available at the same time.
Back in the 1970's i built a board that divided a 2MHz oscillator down to 0.5 Hz. That took a lot of counters but it also put out a lot of frequencies which could be used for testing various things. All of the output frequencies have the same accuracy as the oscillator used to generate the highest frequency. Putting the oscillator inside a temperature controlled oven keeps the frequency very constant, although to test a scope you dont really need that.