I have this Colpitts oscillator circuit. (Original had 2N3694 and 10uH.) I'd want to use it with a microcontroller running at 5V. I've run it in LTSpice and it works, but do I need to change any of the bias resistors for proper 5V operation in the real world?

Schematic and LTSpice file attached.

 

Try what you have at 5 volts it will probably work.

 

So the only inductor I found was 820uH. That's fine, I can test different L for different output F. I also used 2N3904, as I couldn't find any 2N2222 on hand. The circuit works from about 2V to about 4.5V for the power supply. After that it stops oscillating. I'd like it to work for different values of L, from very low to about 1mH.

EDIT: Also, the frequency output is dependent on the power supply. 2V: 479kHz, 4V: 561kHz. So I must have something wrong.

 

I swapped the emitter resistor, R3, for 2K and got better output. Vpk is better and works to much higher voltages. But frequency is still dependent on Vin.

 

This sounds more or less typical. If you want the frequency to be independent of power supply voltage replace C3 and L1 with a crystal.

 

More experimenting: Using 2N3904, the frequency output is very voltage dependent. Using BC547, the frequency is fairly immune to power supply voltage. I must not have the right design for use with the 2N3904.

 

Looking at a BC547 datasheet it appears to have higher gain than a 2N3940 at low currents. You could try the 2N3904 at something like 5 to 10 mA Ic.

 

What version of the BC547 are you using. The BC547 are binned by hFE.

 

They say:

BC547
BK912

...So "B"?

 

Your BC547 is not a selected A, B or C so its current gain is anywhere from 110 to 800.
The pins of a 2N3904 are EBC but the pins of a BC547 are CBE, the opposite. Maybe that is why your BC547 does not change the frequency with supply voltage changes like it should.

The capacitance of a transistor changes when the supply voltage changes that changes the frequency when the pins of a transistor are connected properly.

You said the oscillation stopped when the supply voltage was 4.5V and higher, maybe because the pins are connected backwards causing avalanche breakdown.

 

No, pins are connected correctly. BC547 has a *little* change in frequency with supply voltage, not none. I didn't know transistor capacitance changes with supply voltage, though.

 

BC547 has a *little* change in frequency with supply voltage, not none. I didn't know transistor capacitance changes with supply voltage, though.

The datasheet of a transistor shows the capacitance changing when the supply voltage changes. That is how a simple FM transmitter circuit produces FM.
Many radios are tuned with the voltage changing the capacitance of a diode.

 

...the diode Audioguru again referred to is the collector-to-base junction.

 

Is the typical thermal voltage of 26mV applicable for *all* BJT's?

Also: Does an op-amp make a more stable (better?) oscillator than a BJT? I was thinking because of the "infinite" gain thing, it might be.

 

The datasheet of a transistor shows the capacitance changing when the supply voltage changes. That is how a simple FM transmitter circuit produces FM.

Not to get sidetracked, but I thought the transistor in a simple FM transmitter is acting like a mixer?

 

I swapped the emitter resistor, R3, for 2K and got better output. Vpk is better and works to much higher voltages. But frequency is still dependent on Vin.

Sorry, what is Vin? Are you referring to Vcc by chance?

 

Sorry, what is Vin? Are you referring to Vcc by chance?

Vin = power supply voltage

 

Is the typical thermal voltage of 26mV applicable for *all* BJT's?

The silicon base-emitter diode forward voltage changes with temperature changes the same as silicon diodes and the same as other transistors as shown on datasheets. The saturation voltage change with temperature change is also shown.

Does an op-amp make a more stable (better?) oscillator than a BJT? I was thinking because of the "infinite" gain thing, it might be.

A transistor can amplify or oscillate at a very high frequency. An opamp cannot do anything at a very high frequency. Its gain is almost infinite at DC and very low frequencies.

Not to get sidetracked, but I thought the transistor in a simple FM transmitter is acting like a mixer?

An FM transmitter does not mix audio frequencies with the VHF carrier frequency. Instead the varying low frequency and amplitude of audio varies the capacitance of a tuned circuit which varies the carrier frequency.

 

Vin = power supply voltage

Ah, OK.
Using a properly regulated voltage is not enough?

 

Temperature changes affect a few things of a transistor which causes frequency changes.