Oscillating away in Colpitts country

According to Wikipedia, the Colpitts oscillator was developed by Edwin Colpitts in 1918.  I played around with this oscillator when I first started homebrewing around 2002 or so, and never could get it to oscillate for some reason!  Today, much different results with some breadboarded circuits.

First I recreated Farhan’s circuit on a breadboard, just to kind of calibrate the breadboard setup (and make sure it wasn’t dramatically different from the printed circuit kit board).  And it worked really great — very stable, very close to the 11 MHz crystal frequency, and the pot gave me about 3 kHz frequency adjustment.  For a BFO I’m thinking that makes sense due to a ~3 KHz bandwidth on a typical SSB signal.

bfo osc

Cutaway of the BFO oscillator circuit in the Bitx20


Farhan’s BFO oscillator rebuilt on a breadboard

Next, I decided to just try different crystals in the circuit and see if they’d still oscillate.  I got mixed results with this.  The ones that did oscillate on frequency were stable.  But others oscillated but on an integral fraction of the crystal frequency, which seemed a little weird.  Those had a noisier waveform.  And a few didn’t oscillate at all.  The range of crystals I chose was between about 3 and 18 MHz.

I then pulled the crystals entirely and just looked to see if the non-crystal Colpitts-style oscillator would work.  Sure enough, it did!  But it was VERY unstable.  In fact I could breathe on it and make the frequency go up and down depending on if I blew hot or cool breath!  It was sensitive to any touch of my hand or even the vicinity of my hand to the circuit.  In other words, not a practical circuit.

I don’t think this circuit was designed to be run without a crystal in it.  But the designs I’ve seen for Colpitts don’t look a lot different either.  The non-crystallized Colpitts circuits’ resonant frequency could be ballparked with the mathematics for Xc = Xl.  Components are never perfect and there are parasitics and so forth but I felt like I got pretty good agreement, as I had 12 years ago when I did this exercise.

The big thing for me now is to learn how the circuit works and how to design them for any arbitrary frequency.  One could always look up circuits and try to tweak them.  But surely there is a better way!

Looking through my texts I got a few answers percolating through the noise, but it has been hard slogging.  The vast majority of books just gloss over the circuit.  Sometimes all that’s shown is a particular case for a certain frequency.

Then you sometimes get a statement about phase, saying something to the effect that you need to phase shift the output of the transistor such that the input gets a fed-back version that’s again in-phase.  Often it’s not that clear.  They just say “feedback shifts phase by 180” or something.

My engineering book had a very nice and clear example for this topic.  It didn’t go into how one goes about designing a Colpitts, but it did analyze one very well in terms of the phase situation, running through the algebra.  It became clear that the network of two capacitors and one inductor shift the phase of the output of the transistor by 180 degrees.  Since the transistor does the same thing, the output of this process is a signal back in phase with the input again.

The network also selects the frequency of interest through its resonant frequency.  Signals that deviate from that frequency don’t get the necessary phase shift, so aren’t amplified.  There’s also an increase in impedance going through the feedback network for frequencies off resonance.  So this all makes sense to me.  And a lot of this is just basic info for all oscillators.

One might ask how the oscillation begins in the first place.  The explanation I read, I think in “EMRFD” but I’ve known from other learnings in the past, is just the broad spectrum noise that the transistor begins to amplify when you turn it on.  The frequencies of interest end up getting preferentially selected from the rest, and off it goes.

Still, I’d like to know the essentials, like how do you choose those two capacitors?  In the case of the Hartley, which has a tapped coil, the location of the tap is seemingly experimentally determined for strongest signal.  It seems like an impedance matching exercise with the transistor.  Maybe it’s the same with the Colpitts only using a “tapped” capacitor.  I think that’s true but I’m not certain yet.

A fairly in-depth analysis seems to lie in the book “Radio Frequency Design” by Wes Hayward.  This book was actually recommended reading by Wes’s EMRFD book, in the oscillator section.  I googled it and turns out I already own this book!  But never got around to reading much of it.

So I got to the section on the Colpitts and sure enough it looks like he digs into transconductances and small signal analyses, etc.  I’m vaguely familiar with it but definitely need to study up some more.

I guess the question we all have to ask ourselves too is, at what point is our level of understanding sufficient for our needs?  Do I really need to know all the details of how to design a Colpitts from scratch or not?  I think for me, I would like to know more than I do now.  I don’t think it’s that hard to learn the stuff, at least not in terms of math work.  It might be tedious and a little challenging but still very learnable.  The balancing act for me is time.  If it takes months of effort, I might be better off spreading that time out a bit, taking a break from it and working on some other sections and coming back to it.  Getting bogged down in details on one type of circuit can kind of be a drag.  But getting to that solid understanding is priceless, too.

At the end of the day, I’d like to be able to design things like oscillators and amplifiers more or less to meet my needs and not to have to rely on pre-published designs.  I think I’m slowly getting there but I feel like it might be another year or two before I hit my stride, depending on how much time I devote to it!  It should be fun though and that’s the main thing.  It’s great, to me, just to sit down with a breadboard and plug in different crystals and try different coils and see what happens.

To me, a wacky oscillator that changes frequency with my breath is almost more thrilling than one that is rock stable and never changes!  This is the sort of stuff that’s fun, not always just the pre-packaged, hermetically sealed stuff that we can buy.

This entry was posted in bitx20, books, Colpitts, oscillator, Wes Hayward. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s