Sensing small inductances

The amplitude of the current pulse from the emitter sets the tank voltage. This dumps energy into the tank. The two capacitors across the tank are in series. The tank voltage is monitored by the base of the transistor to tell when to deliver the current pulse. The power is delivered by the emitter to the bottom capacitor. The capacitors act as a voltage multiplier the same as in an autotransformer.

In a Pierce, the resistor from the CMOS output to the tank sets the power level. You are feeding a constant amplitude square wave into a constant impedance tank. The power increases until the power delivered to the tank equals the power lost in the tank. You set the power level by changing the value of the resistor.

In a cc Colpitts, the power is set by the emitter current. This is adjusted by changing the emitter resistor. As the resistance decreases, the current pulse delivered to the tank increases and the voltage increases.

You set the value of the emitter current to reach the desired amplitude. This is reached when the power delivered to the tank equals the power lost in the tank.

Meanwhile, you watch the tank voltage to ensure it doesn't approach VCC, which would forward bias the base-collector junction and cause clipping, which increases the noise in the oscillator.

You also monitor the base-emitter voltage to ensure the voltage does not approach the reverse breakdown junction voltage. This would also generate noise and could harm the transistor.

For crystal oscillators, you monitor the current through the crystal. Since you know the ESR of the crystal, you also know the current through the ESR.

You calculate the power with the formula P = I^2 * R and compare this to the recommended power level from the crystal manufacturer. You adjust the emitter resistor until the power meets the manufacturer's value.

Hey John,

I am pleased you are asking these questions about the Colpitts oscillator. There are a lot more questions to ask. The harmless looking Colpitts is one of the most complex circuits in electronics.

To completely understand the operation, try thinking of how the oscillations start and how they build up then level off. This should raise some more questions to think about.

Some of these questions are hard to answer. You have to keep so many things in your head simultaneously. It is very good mental exercise.

Thanks

that's used as an LCD backlight inverter circuit in many products from the 1980s, except the collector winding in that drawing is instead coupled to the base via the base capacitor, and the collector goes to VCC thru an RF choke

e.g.

You're making it too complicated. A triangle-wave current drive makes a squarewave voltage across your inductor, of amplitude proportional to the unknown inductance. In-phase demodulating removes the e.s.r. artifact.

An ideal inductor >> Lx makes a decent triangle-wave current approximator. You could also consider it as an inductive voltage divider.

Plop this into LTSpice with a 500kHz 5-ohm 3.3V CMOS squarewave drive, and look at the voltages at Lx and Vout.

-. 47uH | .-.-.-. |----||---' ' ' '-----+-----> Vout | C1 L1 |

-' 470nF .-. | | R.esr '-' | ) ) Lx ) 10-500nH | ===

All you need is some gain, and a few CMOS switches for a demodulator.

Cheers, James Arthur

a monitor,

I know how to calibrate my monitors, and I ignore anything that is so slopp y that it can't be viewed on a properly setup display. To me, it indicates that anything related to crap images was only posted to make the owner thin k he is smarter than he really is.

A Video professional would be quite embarrassed to post that crap. I suppos e you maladjusted every monitor you touched at the TV studio rather than co rrect a video problem? I would have fired anyone who did that, and I was th e chief engineer at two of the tree TV stations I worked at. I turned down the job of chief at a forth station, that was looking for their third Chief in under a year for going on air.

That's a heap of "except for."

I did that oscillator somewhere in the late '60s. The intent was to get a very stable amplitude and frequency sine wave, to excite a magnetic-pendulum inclinometer, to measure the level-ness of the aircraft.

There's a paper somewhere showing all tuned indutively-coupled one-transistor feedback oscillator topologies are essentially equivalent in function, where you decide to inject the feedback prolly depends on your desired frequency range and output power requirements/active device characteristics