Tube oscillator, multivibrator?

[loogie]

I want to build a variable, audio frequency oscillator or a multivibrator (I think) using tubes.  It would be nice if it could do sine and square waves.  I've hunted around a bit and I've found a couple of things.  I've also considered basing it on an Eico signal generator that I have. I've scanned RDH4, but I still need to do some more digging through it.  The ultimate goal would be to make it voltage controlled, so basically I'm talking about a tube based VCO.

Ever tried it?  Any general direction you could point me in?

[sluckey]

Just copy V5A in this schematic. It doesn't get much simpler than this...
http://www.schematicheaven.com/bargainbin/gretsch_6160.pdf

V5A is a phase shift oscillator (sine wave) just like most of the tube tremolo circuits, except the timing resistors and caps are much smaller. It's used as a guitar tuner so the freq is audible. Don't have a clue how wide the range is but that would only be a minor challenge.

You could run the output through another gain stage and then use back to back diodes to clip the sine into a reasonable square wave.
 

[loogie]

Thanks, that's good.  I figured there was some kind of an LFO in my future, but it never occurred to me that I could use a phase shift oscillator for anything but.  So far I have neon relaxation, thyratron and multivibrator.  I'll add that to my list.

[PRR]

> Don't have a clue how wide the range is

Roughly third-root of 350K/100K. 1.51:1. Just enuff to cover 20% tolerances.

Say 360Hz-540Hz.

Taking the upper resistance far higher than the others (270K) won't drop the frequency much (the third-root approximation is failing).

For wider range you may use a 3-gang pot. Now freq varies with resistance. 1Meg to 100K will approach 10:1 freq range.

You can switch caps for other ranges. A 3P3T switch will manage say 20-200, 200-2K, 2K-20K. Maybe. The caps get very small above 10KHz and will interact with the triode's grid capacitance.

Multivibrators are dandy for square-waves.

There are ramp-waves on the grids of a symmetrical multivibrator but extracting them loads the timing action and spoils it.

There are tons of ramp-wave tube circuits for TV sweep, much harder to get a symmetrical triangle.

For a wide-range Sine: don't be foolish. eBay yerself an H-P 200AB. Don't worry about condition, age, or known-working. They are easier to fix than the sick Champ in your closet. The only sticky-spot is that the massive tuning-knob's shaft grease tends to turn to gum in only 100,000 hours of always-on duty. And once gummed-stuck it is hard to get at the shaft-screws to dismantle it. Eyedropper some nasty solvent into the bushings and be patient. When it will come undone, scrub the shafts and lube with good engine oil (3 drops from the car dipstick, right after an oil-change, is ample). Do it again before the year 2029 rolls around.

For square/triangle/"sine", you just can NOT beat the little chips made for the purpose.

Voltage-controlled tube oscillator..... they exist. Any oscillator will drift a bit when supply voltage changes. Some GHz tubes can only be tuned with an electrode voltage. FM radio AFC uses a second triode as a vari-gain capacitor to drift the LO. Conceptually a multivibrator can be VCOed, and this is easy with transistors, maybe not so easy with tubes.