Some oscillator notes

Thread Starter

KL7AJ

Joined Nov 4, 2008
2,229
or
"More of the stuff they don't teach you in school."

It's one thing to model an oscillator. It's quite another to make it work. Making oscillators work sometimes seems like an arcane art, conducted in dark basements by strange, bearded old men, surrounded by boiling cauldrons of vile smelling potions and such.

This is not really true. I don't have a beard. :)

There are several things that can make an oscillator work, or work in a strange manner. I'll address just one of the common issues here....moding.

Any resonant circuit can exibit two (and sometimes MORE) modes of operation. The two fundamental modes are series and parallel resonance. If you were to look at a seemingly simple solenoid inductor...let's say one of 1mH or so...on a network analyzer swept between a few hertz and a few GHz....you would be AMAZED at what a rat's nest you'd see! There are MULTIPLE resonances, both series and paralel, some harmonically related, some not. This is primarily due to PARASITIC capacitance...the capacitance that naturally exists between turns of an inductor. Capacitors also exhibit some parasitic inductance, which can also be a problem at UHF frequencies, but generally less of an issue for most basic oscillator designs.

The main trick for successful oscillator design is to be sure the oscillator only operates in ONE mode. This can be especially difficult to do with CRYSTAL oscillators, where the Series and Parallel resonant modes may be separated by a minuscule fraction of the nominal frequency. However, the clue to taming a crystal can be applied to any oscillator.....by assuring that your DESIRED mode has the Highest Q. For a SERIES resonant oscillator....using either a crystal or an LC circuit, you achieve this by using LOW load impedances. A crystal with a HIGH load impedance is almost guaranteed to operate in the PARALLEL mode, so you want to be sure both the source and the load impedances of your crystal are LOW.

If you WANT parallel resonance, then the opposite is true...you want your source and load impedances to be very high. This, again, applies to both crystals and LC's.

If you have an oscillator that tends to jump around in frequency...the FIRST thing to look for is possible MODE jumping. ALTER your load and source impedances and see what happens.


End of lesson one.

eric
 

flat5

Joined Nov 13, 2008
403
Thank you. Looking forward to lesson two :)

"It's one thing to model an oscillator"
Actually, I've had no success with simulating a phase shift audio osc. or a series Clapp 7mhz osc. using Proteus.
 

eblc1388

Joined Nov 28, 2008
1,542
How do one find out if a crystal is intended for parallel or series resonant?

I think the dumb answer is to try both and see which one gives a frequency closest to the marked frequency on the can.

Any other tests or observations that one can do?
 

Thread Starter

KL7AJ

Joined Nov 4, 2008
2,229
How do one find out if a crystal is intended for parallel or series resonant?

I think the dumb answer is to try both and see which one gives a frequency closest to the marked frequency on the can.

Any other tests or observations that one can do?
Howdy:

ANY crystal can operate in either the series or parallel mode. Which one it atually operates on is entirely dependent on the external impedances. Can markings USUALLY indicate the series mode, but don't count on it.


Eric
 

wr8y

Joined Sep 16, 2008
232
Thank you. Looking forward to lesson two :)
"Thank you" -sure, I can go along with that. But I am not ready for lesson two...

WHAT do you mean by "source" and "load" impedances? I suspect that "load" does not refer to the impedance of the stage following the oscillator.

And I have NO freaking idea what "source impedance" in an oscillator refers to.
 

Thread Starter

KL7AJ

Joined Nov 4, 2008
2,229
If you look at a Butler oscillator, it's a perfect example of an oscillator with low source and load impedances. The crystal is fed from the emitter of one transistor (an emitter follower stage, and feeds a common base amplifier (low input impedance). The output of this stage is fed back to the input of the first stage.

This low impedance all but guarantees it can't oscillate in the parallel mode.

eric
 

Thread Starter

KL7AJ

Joined Nov 4, 2008
2,229
If you look at a Butler oscillator, it's a perfect example of an oscillator with low source and load impedances. The crystal is fed from the emitter of one transistor (an emitter follower stage, and feeds a common base amplifier (low input impedance). The output of this stage is fed back to the input of the first stage.

This low impedance all but guarantees it can't oscillate in the parallel mode.

eric
http://www.icmfg.com/crystaloscillatordata.html
 
Top