0
\$\begingroup\$

I've built the circuit shown in the picture on a breadboard. The simulation works as expected but when I build it I don't get any oscillations. I've measured all the DC voltages and they are as expected. The measure hfe of the transistor seems to be 216 while I assumed 100 in my calculations.

I've done a continuity check using my multimeter and all seems to be in order. Do you guys have any idea why this might be happening?

The BJT is biased at 5mA. The collector voltage is 6V.

Edit1: Added a 10uF capacitor. Also added a 0.1uF emitter bypass capacitor to increase gain. Still no go.

enter image description here

enter image description here

enter image description here

\$\endgroup\$
6
  • \$\begingroup\$ I'm assuming that you're using an oscilloscope to watch the oscillations (or lack thereof)? If you are, can you get a screen shot of that? \$\endgroup\$
    – user103380
    Commented Mar 13, 2020 at 2:56
  • \$\begingroup\$ @KingDuken Just did that. Thanks! \$\endgroup\$
    – Pratik Kunkolienkar
    Commented Mar 13, 2020 at 3:04
  • \$\begingroup\$ Did you compute Q? \$\endgroup\$
    – D.A.S.
    Commented Mar 13, 2020 at 5:32
  • \$\begingroup\$ Using a trim potmeter to adjust the bias might help. If the transconductance of the transistor is too low, the circuit will not oscillate. Also, the emitter degeneration may be too high. Bypass the emitter resistor. \$\endgroup\$
    – Bart
    Commented Mar 13, 2020 at 7:22
  • \$\begingroup\$ @TonyStewartSunnyskyguyEE75 If i'm doing this right, my inductor resistance was measured to be 0.2 Ohms, so \$0.2*\sqrt{\frac{820}{7.5}} = 2.0913\$ \$\endgroup\$
    – Pratik Kunkolienkar
    Commented Mar 13, 2020 at 13:32

4 Answers 4

Reset to default
1
\$\begingroup\$

Where is your VDD Cbypass cap?

You need a 0.1uf from the top of R1 to ground.

\$\endgroup\$
2
  • \$\begingroup\$ if you mean the 47u, its the blue thing above the transistor \$\endgroup\$
    – Pratik Kunkolienkar
    Commented Mar 13, 2020 at 4:12
  • \$\begingroup\$ @PratikKunkolienkar, no that is the coupling capacitor. Analogs' answer is valid. Omitting the power supply decoupling capacitor can complete void the functioning of a circuit like this. \$\endgroup\$
    – Bart
    Commented Mar 13, 2020 at 7:20
1
\$\begingroup\$

BJT is biased at 5mA and Collector voltage at 6V.

Just by looking at the 16.6k/7.32k divider for the transistor base bias voltage, that is already not the case. At least a 100nF decoupling capacitor for the power supply should be added. The simulator doesn't care for that because the voltage source used as a supply is ideal and does the perfect "decoupling" for you. Reality is very different. The L/C ratio (820nH and 15nF) is off and the values are not practical for a 2MHz oscillator. The 47uF base coupling capacitor is excessive, 10nF would suffice. Try building the oscillator dead bug style on a (single sided) copper clad board to get a nice and clean RF ground. Those plastic breadboards are barely adequate for DC circuits let alone RF.

Below is an updated schematic. To preserve linearity in the gain stage, I added an external soft limiter to the loop to get a clean 4Vpp output voltage with low harmonics. Beware that the output is DC coupled. Inductor Q is important and if too low the oscillator will not work. The chosen min. Q is 50 @ 2MHz. Higher Q will give a higher output voltage and you would have to maybe decrease the value of resistor R5 in the soft limiter and/or the output coupling capacitor C4.

enter image description here

enter image description here

\$\endgroup\$
2
  • \$\begingroup\$ Thanks for the answer! Could you explain why the L/C ratio is off and how the soft limiter works? \$\endgroup\$
    – Pratik Kunkolienkar
    Commented May 6 at 10:48
  • \$\begingroup\$ @Pratik * Reactance of tank circuit components is only about 10 Ohm. An inductor value in the higher nH range together with capacitor values in the mid nF range is unusual for a 2MHz oscillator. The diodes in the limiter represent a nonlinear load and clip hard. In order to limit the voltage and have as low (soft) as possible impact on the distortion performance of the output voltage, the series resistor isolates the diodes from the tank circuit. \$\endgroup\$
    – Raonoke
    Commented May 8 at 1:15
0
\$\begingroup\$

The problems that I see with this circuit:

  • No power supply decoupling.
  • The gain at the collector is 12 at most (1200/100), which may be too low. Loading at the collector by the feedback pi-filter, (and the oscilloscope!) will further reduce the gain.
  • No adjustable bias voltage. The emitter bias current controls the gm (transconductance) of the transistor. This value is important for the correct functioning of the circuit.

I suggest replacing the 1.2k resistor by a choke which resonates at the intended frequency of the oscillator, maybe bypassing the emitter resistor to increase the gain, and making the base voltage adjustable with a potmeter.

I recently built a 16MHz Colpitts VCO using this approach, albeit in common base configuration. Getting it to oscillate took a minute of adjusting.

\$\endgroup\$
3
  • \$\begingroup\$ Added a 10uF capacitor this morning. Also added a 0.1uF emitter bypass capacitor to increase gain. Still no go. How do i spec the choke? Isn't a choke just an inductor? how do i make sure it resonate at the right frequency will that be its self resonating frequency? \$\endgroup\$
    – Pratik Kunkolienkar
    Commented Mar 13, 2020 at 12:39
  • \$\begingroup\$ It is just a coil that has parallel resonance at the working frequency so the collector will see it as a very high impedance. electronics-tutorials.ws/oscillator/colpitts.html \$\endgroup\$
    – Bart
    Commented Mar 13, 2020 at 12:45
  • \$\begingroup\$ Just realised that i might need a lower ESR bypass capacitor so it works correctly at higher frequency. I was using an electrolytic \$\endgroup\$
    – Pratik Kunkolienkar
    Commented Mar 13, 2020 at 13:34
0
\$\begingroup\$

In 1975 when I graduated working for Bristol Aerospace, a gold medalist EE Bill Whitehead was my mentor and he showed me these RLC Impedance graphs. Since we used a slide rule before the LED calculator came out in the early 70's I understood how to do log math on a slide rule.

Since I assume you have some Impedance math skills, you ought to be able to do the same.

This is a test. Does this filter attenuate? What is the loop gain?

enter image description here

enter image description here

Final question: What is the phase shift at resonance and is it correct?

\$\endgroup\$
4
  • \$\begingroup\$ Hint: -90 before and +90 after \$\endgroup\$
    – D.A.S.
    Commented Mar 15, 2020 at 6:14
  • \$\begingroup\$ Well it will attenuate with a 180° (should be 90°?) phase shift won't it? Isn't that the point? and then the transistor amplifies it to satisfy the Barkhausen criterion. right? \$\endgroup\$
    – Pratik Kunkolienkar
    Commented Mar 16, 2020 at 0:18
  • \$\begingroup\$ keep in mind jumpers are about 0.5~1nH/mm and where is Vcc cap? \$\endgroup\$
    – D.A.S.
    Commented Mar 16, 2020 at 1:19
  • \$\begingroup\$ I've added a 0.1uF after someone on here reminded me about it. Someone also said that i didn't have enough gain. So I've ordered some 33uF inductors and 15pF caps. \$\endgroup\$
    – Pratik Kunkolienkar
    Commented Mar 16, 2020 at 2:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.