Transistor 100mhz oscillator coil info

I want to put together an xtal controlled oscillator to generate a 100mc buffered signal.
I have a 20mhz xtal, which I assembled a colpitts osc with using a couple of bsx20 trannies.
I'm trying to tune the oscillator to get the thing to oscillate at the 5th overtone, but I'm having issues with sustaining oscillations.
What kind of Q should I be looking for for this kind of circuit, typical fm bug style circuits spec a 5mm coil which has a Q of around 300, if I wound a 20mm coil I'd get 1000.

Consider that the crystal may not like operating on its 5th overtone.
Are you sure that the crystal is operating in a fundamental mode at 20MHz?
I once tried to use a 20MHz junk box crystal with a PIC, and it insisted on oscillating at 6.666MHz, it was obviously made as a 3rd overtone crystal for 20Mhz.
If your crystal is like mine, you will be trying to run it on its 15th overtone. That may be streaching things a bit!

What circuit are you using for your oscillator?

It may be easier to run the crystal at 20MHz, and use a X5 multiplier stage to get up to 100MHz.

JimB

JimB said:

Consider that the crystal may not like operating on its 5th overtone.
Are you sure that the crystal is operating in a fundamental mode at 20MHz?
I once tried to use a 20MHz junk box crystal with a PIC, and it insisted on oscillating at 6.666MHz, it was obviously made as a 3rd overtone crystal for 20Mhz.
If your crystal is like mine, you will be trying to run it on its 15th overtone. That may be streaching things a bit!

What circuit are you using for your oscillator?

It may be easier to run the crystal at 20MHz, and use a X5 multiplier stage to get up to 100MHz.

Click to expand...

I would agree, the classic 144MHz transmitters usually used 12MHz crystals, followed by 3x and 4x multiplier stages to get to 144MHz, this also had the added benefit of multiplying the NBFM deviation as well.

If his crystal is a radio crystal, as you say it may already be an overtone one - whereas if it's a micro clock crystal, then it will be a fundamental one.

Good point, not thought of that.
However yes I'm fairly sure its fundamental 20 mc.
Reason I say that if I remove the collector tank on the osc the thing oscillates at 20 mc, so either the thing is still oscillating 3rd harmonic untuned or it really is a 20 mc xtal.
I'll put it across my Vna and see what it says, the strongest peak will be its fundamental.
Its a processor type xtal so its probably series cut.
Thanks guys.

Turns out it was a 20 mc xtal.
There was a problem with my circuit, biasing, it used a 220k resistor to B+, I replaced that with a pair of 1n4148's as a reference biased by a 1k resistor.
I now get loads of amplitude, 0.5 volts, however I have not managed to multiply the frequency yet.
I have it working with a 25mc xtal, might try doubling that to 50 on the osc itself, then double again with another tranny to the 100mc, then buffer that with anpther trans.

I think i get it, its been a while since I played with Rf.
I basically took a untuned colpits xtal osc circuit and bunged a parralel L/C tank in the collector load, then took the o/p from the collector.
This wont work, the tank is high impedance at resonance and low impedance everywhere else, so the xtal is only going to ring at its fundamental or not at all.
What I should have done was put the L/C in parallel, across the supply rails, ie the L from collector to B+ which also allows dc bias, and the cap from the collector to ground, then all but the harmonic of interest will be damped (into the supply).
Funny how you 1/2 forget stuff if its been a while, I'll have another play tomoz & report back, time for bo bo's now, I'm an essential worker and still at it.

Have a look at oscillator 2, here - that should be an appropriate arrangement?

Interesting, first time I've seen a low value resistor to suppress the fundamental.
Presumably L1 and the series capacitance select the required overtone.
With only 1 inductor and 1 stage might just try that as a lash up later.

Further to my comment above, the LC tank in the collector would probably work if I made the L a transformer the secondary of which being the o/p,.

Sussed it in the end.
The circuit I'm using is a simple oscillator with a tuned circuit in the collector load, followed by a emitter folower buffer.
The problem I had was in the design of the buffer amp, seems the slightest load and the thing just oscillates at its fundamental.
A bit of working out and I now have nearly a volt of 100mc Rf, whuch now needs a screening can, its amazing how far a broadcast receiver picks it up, which isnt desired.