Fixing Colpitts Oscillator Not Producing Sine Wave

[Jarhead42]

hello!
I have built a colpitts oscillator, but it is not producing a sine wave.
the wiring diagram is attached.
I used:
L1= 1.066 millihenries
L2= .142 millihenries
R1= 10000Ω
R2= 10000Ω
C1= 470 microfarads
C2= 680 microfarads

for the transistor i just used a small npn.
Tahnks!

 

[skeptic2]

What does the output waveform look like?
What is the voltage of your source?
What is the peak to peak amplitude of the output waveform?

 

[Jarhead42]

My source of power is a small 3-volt. There is no output waveform at all.

 

[skeptic2]

Try eliminating R1 and R2 altogether. You may want to put a DC blocking capacitor on the output.

 

[Jarhead42]

Thanks! I'll try that.

 

[NascentOxygen]

Try eliminating R1 and R2 altogether.

That does not sound like good advice.

 

[NascentOxygen]

My source of power is a small 3-volt. There is no output waveform at all.

A tank circuit containing variable capacitors does not usually comprise huge electrolytics!

What values of caps does the original circuit use, the circuit you copied?

What frequency are you aiming for? Is this just a lab class, or do you have a purpose in mind for a variable-frequency sinewave oscillator?

Could you measure the resistance of your two coils, and let us know these values? Where did the coils come from?

 

[davenn]

Hi NascentOxygen

Yes I agree with you about the resistors, haven't seen a Colpitts or similar osc without them

Also about the capacitor values, something ~ 100pF and 10pF as suggested in that tutorial would be more suitable

I found the www page that the OP's circuit came from ...

http://www.electronics-tutorials.ws/oscillator/colpitts.htmlMaybe the OP didn't get the circuit directly from that page and maybe that's why the suggested component values weren't used ?

OP go to that page and used all the values of resistors, capacitors and inductors as suggested, then see how you get on

I would also suggest that the supply should be a bit more than 3V, 5V to 9V would be much better

cheers
Dave

 

[willem2]

I found the www page that the OP's circuit came from ...

http://www.electronics-tutorials.ws/oscillator/colpitts.html


Maybe the OP didn't get the circuit directly from that page and maybe that's why the suggested component values weren't used ?

C3 is missing. There's a DC path from the collector to the base through L, so the transistor will be saturated always.

 

[davenn]

C3 is missing. There's a DC path from the collector to the base through L, so the transistor will be saturated always.

indeed, I didn't notice that in the OP's rendition of the circuit

D

 

[skeptic2]

Friends,

Before I answered, I found the colpitts circuit online and saw that this circuit was missing the two capacitors. I recognized that the base was biased at 1.5 volts which would drive the transistor firmly into saturation and decided that that was probably the reason it wasn't oscillating. Then I entered the OP's circuit into LTSpice and discovered it didn't oscillate either. So I adjusted values of the two resistors until the transistor was biased at about 0.65 volts. The circuit oscillated quite well. Then I added the two capacitors of the internet circuit to his circuit and it would no longer oscillate.

I've had quite a bit of experience with PSPICE and one advantage PSPICE has over LTSpice is that you can set initial conditions for capacitors. SPICE tends to calculate steady state conditions for all components before it starts running and many times oscillators will not oscillate in SPICE when the circuit starts with steady state conditions. Oscillators depend on an imbalance of charge to start oscillating. With PSPICE you do that by setting the initial conditions of the capacitors to 0 V. So when the internet circuit didn't oscillate I looked for a way of setting initial conditions but couldn't find any. Obviously if a circuit doesn't oscillate, though it may be correct, it can't be analyzed. Thus I went back to the OP's original circuit but discovered it would oscillate quite well without the resistors. Note that since with the resistors biasing the transistor at 1.5 V, the circuit with the resistors will always draw more current than one without them. If he hadn't already blown the transistor from his original circuit, removing the resistors wouldn't blow the transistor. If he had already blown the transistor, by eliminating the resistors it will become obvious that the transistor is already blown.

Either way, the circuit I would like to propose to him, still works in LTSpice and yet will reduce the current through the transistor. See attached.

 

[NascentOxygen]

I recognized that the base was biased at 1.5 volts which would drive the transistor firmly into saturation and decided that that was probably the reason it wasn't oscillating. Then I entered the OP's circuit into LTSpice and discovered it didn't oscillate either.

There's nothing like runnng a parallel simulation to discover new things about a dodgy or unconventional circuit!

What do you mean by the base being biased at 1.5V??

How did you know the resistance of OP's tuning coil? Much does hinge on it.

You didn't know OP's particular β, so you established its operational status/non-status over the range of typical β's?

BTW, I can't see the transistor being saturated here. Certainly, a sub-optimal operating point, but I think perfectly feasible as a demonstration, given the 3V power supply goal.