oscillator

[dsavitsk]

I am not sure where to put this, but I thought a tube circuit might exist. I would like to build a circuit to generate a very clean sine wave -- cleaner than what I can get from the wall. I need 60Hz, but being able to adjust it would also be good. Any suggestions?

[Eli Duttman]

3 candidate circuits come to mind: the Armstrong, Hartley, and Colpitts oscillators. Usually found in RF equipment, value changes (large caps.) will produce the 60 Hz. you want. Adjustable is not a problem, as that's exactly the case in superhet radio receivers.

A regulated B+ rail is an important step in controlling frequency drift.

Good luck!

[EC8010]

Sadly, none of the above RF oscillators can be easily scaled to become a 60Hz oscillator. In fact, 60Hz is quite hard to do with low distortion. Wien bridge is the obvious choice, but even 0.1% is hard to do. After that, the solutions become far more complex.

[Eli Duttman]

The memories are FUZZY at best, but I do recall a code practice oscillator that used Colpitts topology. High impedance dynamic headphones did double duty, providing both the audible O/P and the inductance needed by the Colpitts circuitry.

Moving the generated tone down to 60 Hz. could easily be difficult as EC8010 indicated. Also, I can't begin to guess what the distortion pattern of that CPO is.

[timpert]

Hi,

I have on my desk the Hewlett Packard 200CD oscillator. Mine uses 2 pcs. 6AU6 and 2 pcs. EL86. It is basically a Wien-bridge oscillator that is tuned using a variable air capacitor. It does anything between 5 Hz and 600 kHz, so your needs should be covered by it. It is a tube circuit to boot, Googling "HP 200CD schematic" yielded a schematic within the first few links. If you are a little dexterious, you should be able to bodge up something based on it, that suits your requirements. The fact that you don't need wide range tunability means that you can eliminate a lot of components.

A word about signal quality: when you hook up the secondary of a low voltage transformer to the input of an oscilloscope and look at the waveform, you'll notice that its cleanliness isn't particularly hard to beat...

Have fun experimenting!
Jurgen

[dsavitsk]

Originally posted by timpert
I have on my desk the Hewlett Packard 200CD oscillator.

Thanks, that's an interesting suggestion. These actually appear to be widely available, so I might just pick one up to tinker with.

-d

[hagtech]

Easiest way to generate a 60Hz sine, is to take the secondary outptu from a power transformer and low pass filter it. The more filtering you add, the less distortion.

jh

[dsavitsk]

Originally posted by hagtech
Easiest way to generate a 60Hz sine, is to take the secondary outptu from a power transformer and low pass filter it. The more filtering you add, the less distortion.

Thanks Jim, I thought about this, but I am trying to build a power supply for a LP12, and it would be nice to vary the frequency both to get right at 33rpm, and to try to hit 45. For a short term solution it might not be a bad way to start, however.

[EC8010]

Somehow, I thought the word "turntable" would turn up soon. Do a search in this forum and you will find that there are various turntable oscillator threads .

[hagtech]

Aha! A turntable speed control is an entirely different matter. Then you will probably want crystal controlled accuracy. Or at least an oscillator with frequency stability.

I would then suggest a high frequency oscillator that drives a counter that feeds a ROM or lookup table that outputs to a DAC. An ARB, as they call it - arbitrary waveform generator. The ROM just has the PCM codes for a sinewave as you clock through it. Or some similar circuit.

Check out some of the new resistor controlled oscillators at www.linear.com. Interesting, but pretty high in frequency. Maybe too much adjust range. No xtal required. With an xtal, you will have discrete frequency steps.

Or perhaps, the poor man's version uses a square wave oscillator (LM555?) that feeds a string of RLC low pass filters. You can get a pretty good sinewave doing such. Doesn't have to be perfect to drive a motor, and the small frequency variations are no problem.

jh