? How to stabilize a cheap colpitts oscillator

First, what's 'wildly', and what are you expecting? You'll obviously lose some frequency stability in changing from a xtal to LC. And is the amplitude unstable as well, that's a key indicator of what the problem might be.

Also, IIRC a colpitts osc normally has one inductor. If you're using 2 toroids, is this 2 separate inductors? If so, should it be one inductor with a tap, providing the mutual coupling that's required?

Wildly: fluctuating from 1.0xxxxx MHz to 6.7xxxx Mhz by degrees of 500 KHz at a time or more. I was expecting the most significant digit to remain stable, and the fluctuations to involve the two or three least significant digits, instead of all digits changing over the span of a couple seconds. And, no, the amplitude and frequency on the scope are stable, they don't fluctuate at all. Every now and then it will settle down to roughly what I expected, but that doesn't last long. No obvious clues as to why.

The two inductors are switched into the circuit one at a time, to allow me to raise or lower the general frequency range to match that of the circuit I am trying to test.

Thanks for the reply. Please don't hesitate to offer anything you think of...

Dave

If it looks stable on a scope but it's jumping that much on a counter, it's the counter's fault, likely a trigger threshold issue. You could expect a decent LC colpitts to be short-term stable to 100 ppm sort of turf, wiggling in maybe the 4th or even 5th digit.

What frequency does the scope suggest? Does the scope waveform look stable while the counter is connected and jumping around?

John

I honestly wondered about the counter as a possible source of the problem. A trigger threshold problem would explain this perfectly in my mind. The scope shows a stable 5 to 10 MHz, (depending on what I dial up with the tuning cap in the Colpitts) while the counter says anything between 1 and

Thanks,

Dave

"Dave"

Have you got a spec for the thing ?

....... Phil

A spec for the counter? No, it's old and I got it used. Would a tank circuit produce a signal full of harmonics? I somehow thought it would be fairly clean... And 100 mV shouldn't be a problem, I don't think. The scope shows a 2v signal peak to peak.

Thanks,

Dave

"Dave"

Have you got a spec for the thing ?

Might be instructive.

....... Phil

Yes, the scope and counter are connected simultaneously. That's why I wondered about the counter...

Dave

"Dave" <

DESCRIBE YOUR TEST GEAR FULLY !!

....... Phil

I apologize, the freq counter is a Schlumberger 1252, and the scope is a Tek

Dave

"Dave"

What is the frequency range of the counter (Goolge supplied no hits on the model number) .

Die it have a trigger control or bandwidth filters ?

Why cannot YOU tell ME if the scope trace shows harmonics in the wave form ?

....... Phil

The counter goes up to at least 100 MHz, but that's all I know. The trace on the scope is clean. Again, I apologize, but I don't know what to tell you. I am still fairly new to RF, and am learning as I go along... If you have a question, please ask. This is why I am asking for help...

Dave

What's the purpose of the oscillaor? Perhaps a different circuit is called for. Was your frequency counter working with the old (xtal) version? Try:

Frank

Hello Frank,

The oscillator is an attempt at a very basic signal generator, for the purpose of determining the Q of the inductors in the tuning circuit of an active shortwave antenna. Now that you mention it, the counter did function properly when I had a xtal in the oscillator... Back to square one- how to stabilize the oscillator. I will try your suggestions of attenuating the signal, and decoupling the counter. Those are two things I would have never thought of...

Many thanks,

Dave

Well, a squirt of contact cleaner on the sensitivity adjustment knob (from the outside, I'll take it apart tomorrow and do it right) helped a little, but the decoupling capacitor had no effect, and I already had a 100K resistor between the stator of the active antenna's tuning capacitor and the frequency counter's input and as I recall that did help a little also. Now fluctuates less wildly, but still too much to tell what's actually going on. About to put it to bed for the night, and pick it up again in the morning. Thanks for the ideas though. If you think of anything else, please don't hesitate to suggest it. I apologize for being so ignorant about RF.

BTW, the freq counter is an autoranging model, with a 1 second, 10 millisecond and auto positions as well as a sensitivity adjustment. As I mentioned, it counts at least to 100 MHz, but more than that I don't know. It was a gift from my brother in law about 25 years ago, and it wasn't new then. I suspect it was his first frequency counter.

Thanks to all. Sorry for the frustration...

Dave

Does your frequency counter have a sensitivity control? If so, you may have it set too high or too low. Try various settings to see if you can get the reading to stabilize.

Good luck.

John

"Dave"

Was that resistor in line before or after the input cable ?????

How about a JPEG or two of your set up plus test gear on ABSE ?

.......... Phil

I replaced the crystal in my basic Colpitts oscillator with a couple hand-wound toroids and a variable capacitor, and the output now varies wildly. Any ideas on how to stabilize it a little? Or how to tell if the problem is my freq counter?

Thanks,

Dave

. I replaced the crystal in my basic Colpitts oscillator with a couple hand-wound toroids and a variable capacitor, and the output now varies wildly. Any ideas on how to stabilize it a little? Or how to tell if the problem is my freq counter?

Thanks,

Dave

----------------- Stabilize an oscillator??...This is messing with my feedback theorey. If it's too stable..wouldn't that be DC? Or how about... Making an oscillator by unstabilizing a DC source. :)

(I skipped reading the whole thread)

Practical I recall making unstable oscillators with the following methods:

A bad designer knows more about what doesn't work :) D

-------------

I may be ignorant, but I am sincere. The resistor feeds the input cable to the counter. IIRC it was necessary to prevent overloading the counter's input. Just thought of something: problem might be where I am taking the input from- the tank circuit of the active antenna. Damn. Should have thought of that sooner. I never claimed to be a genius, just persistant.

I'll see what I can do about a JPEG of the setup and test equipment. That will have to wait until tomorrow however. It's tired and I'm late.

Thanks all, for your patience.

Dave