How can I get an osciallator to work on its 3rd Overtone?

[hmic21]

Hello all,
I recently ordered a 40.7125MHz crystal to make an oscillator, but I can't get it to resonate at that frequency. I first tried a simple Colpitts configuration, and it was oscillating at 1/3 the frequency. I then tried a circuit with a cap between the base and the emitter, and a tank circuit at the emitter, which gave me an oscillation at about 95MHz. I've been on this for weeks now, but to no avail. Can anyone help me? Are there any good books or web sites to visit on this. I've looked all over the net, this is my last resort.
Thanks

 

[Warpspeed]

Any crystal will try to oscillate at it's fundamental, because that is it's most efficient and lowest loss operating point. Any oscillator will oscillate at a frequency where there is maximum overall loop gain, and suitable phase, (maximum feedback).

To force it into the third overtone, the oscillator circuit needs to have far more gain at the third harmonic than at the fundamental frequency. That usually always requires a tuned circuit tuned to the third harmonic, in addition to the crystal.

The tuned circuit does not need to be particularly sharp, or of high Q, but it needs to limit the low frequency gain sufficiently to prevent fundamental operation.

Feedback to the crystal must contain a lot more third harmonic energy than fundamental energy.

 

[VEBill]

Might worth checking the OEM website to see if they have any reference circuits. Probably a long shot these days...

 

[logbook]

I don't fully agree with Warpspeed here. The oscillator does not magically scan the frequency transfer function around the loop to find the maximum gain. The loop gain must only be greater than 1 at one frequency (given the loop phase shift is also zero at this frequency). If the loop gain is greater than 1 at more than one frequency (and if the loop phase shift is also zero at these frequencies) then the start-up oscillation frequency can be a bit indeterminate.

In short it is the external network around the crystal/amplifier which determines which over-tone is selected. The third over-tone is quite easy. I have made fifth overtone oscillators and there was a slight tendency to go off at the seventh over-tone, but only one in a thousand units though.

As long as the loop gain is slightly greater than one at the point of zero phase shift the oscillation will start. When it is oscillating stably the loop gain has to be exactly 1.

Make sure that the loop gain is less than one at all other frequencies than the required frequency and spurious oscillations will be gone.

 

[logbook]

Why not post your circuit with values and we will then be able to suggest which bits to tweak. If you spend more than a day on this sort of circuit to get an oscillation at the right frequency something is really wrong. You may spend several days getting the layout right or getting the phase noise down, but not getting the right frequency.

I assume that the crystal you bought is a 3rd over-tone crystal. Any decent crystal manufacturer has app note circuits for every frequency range they make. These may be CMOS gate oscilators for lower frequencies or transistors for higher frequencies.

40MHz could be done with a gate oscillator, but a transistor will be fine as well. We are only talking about adding an inductor or a capacitor or a resistor here and there to make it work.

 

[Warpspeed]

here are some crystal oscillator circuit ideas for you.

http://www.northcountryradio.com/PDFs/column007.pdf

 

[felixc]

If you are using logic gates to get a square wave, usually adding a series LC (tuned to 40MHz) to the crystal does it, provided that the loop has enough gain to run at 40MHz.

 

[hmic21]

http://img55.imageshack.us/img55/9171/oscweb4qa.png

This is a link to my current circuit. It should be noted that I don't have too many parts available and the inductors I used were too small to achieve that value, so I had to put 4 inductors in series, which I'm pretty sure will be a problem. They are all surface mounted components, on a homemade board. The output is going to into a multiplyinig circuit (X8) then into a mixer. What do you guys think I can do to change the frequency, I've played with the feedback cap and the tank circuit at the emitter but I can't seem to shift the frequency more than 10-15MHz either way. If I change those values too much, it just stops oscillating.
Thanks

 

[logbook]

I simulated the circuit on SPICE. Without the crystal a current source injected into the base shows a voltage peak at 10MHz, not 40MHz.

The circuit looks wrong. Try the circuit from

http://www.euroquartz.co.uk/

... their "technical" application note.

And please get a selection of inductors and capacitors. This is a very low cost compared to weeks of effort.

 

[kontiki99]

Here's a link to another outfit that specializes in crystals:

http://www.foxonline.com/

 

[hmic21]

So what components control the frequency, if I mess too much with any of those components, the circuit just stops oscillating completely. I got this circuit of a web site that claims this is circuit should oscillate at 40.7125MHz. That tank at the emitter is tuned to 10.7MHz, and I dont' know why, anyone have any idea?
Thanks everyone for helping me out.

 

[logbook]

There is good stuff on the web and duff stuff too. That is why I recommended a crystal manufacturer's website, because they know what they are talking about.

Which website did you get your dodgy circuit from?

 

[europus2002]

Hi Hmic21,
Just flip the crystal from current position to the feedback path between collector and base. Also change the inductor capacitor values to resonant frequency . The circuit should oscillate consistently. I had tried the suggested configuration at 5MHz and could get it to oscillate upto 15MHz.. Your overtone circuit will work at needed OT with high Q inductors ...

 

[hmic21]

Thanks everyone,
I finally got it to work. I had to tweak the tank circuit at the emitter to the frequency between the 2nd and third harmonic, the third being the one I wanted. I also changed the feedback cap from 150pf to 15pf. It took a while but I finally got it, thanks to everyone who helped me out.