Pinpointer: Colpitts + MCU

[Teleno]

I'm playing around with a front end for a pinpointer (see LTSpice attached). It's based on measuring shifts in the bias point and frequency of a Colpitts oscillator at about 150 KHz. Measurement and post-processing is carried out by an MCU.

The Eddy currents of a target increase power consumption. This can be measured both as an increase in the base voltage of Q1 and an increase in frequency .

The contentions are:

1.- Would the quotient provide a means of discrimination?

2.- Would a magnetic ground cause an increase in power consumption (magnetic losses) comparable to that caused by a target?

3.- Ways of increasing the AC current in L1 without increasing the battery's voltage.


Circuit:

Oscillation freq: 160 KHz.
Coil current (AC) 160 mA p.p.
V(out) is proportional to the bias shift. V(freq) is where the MCU acquires frequency.
R13 adjusts the offset, trim to obtain V(out) approx. 200 mV without target.
R10 adjusts the gain.

[Chet]

Hi Teleno
Looks like a good plan. The delta bias and delta frequency should work. I have found some magnetic ground affects my Garret pinpointer. I adjust for it by placing it against the ground and doing a reset. I guess this is my poor man’s ground balance system. I wouldn’t be too concerned with battery voltage and current drain. I have one pinpointer that uses two 9v batteries in series. It is on so little time that the batteries last a long time.
I am interested in trying your design for the ability to discriminate ferrous verses nonferrous. If you can lead the way with some software code it would be greatly appreciated.
Have a great day,
Chet

[Teleno]

Hi Teleno
Looks like a good plan. The delta bias and delta frequency should work. I have found some magnetic ground affects my Garret pinpointer. I adjust for it by placing it against the ground and doing a reset. I guess this is my poor man’s ground balance system. I wouldn’t be too concerned with battery voltage and current drain. I have one pinpointer that uses two 9v batteries in series. It is on so little time that the batteries last a long time.
I am interested in trying your design for the ability to discriminate ferrous verses nonferrous. If you can lead the way with some software code it would be greatly appreciated.
Have a great day,
Chet

I need to rephrase my question regarding power consumption becaus to was misunderstood. I wanted to know whether magnetic ground would shift the bias of Q1 causing false signals. I suspect the magnetic losses would be reflected as a slight increase in power consumption at the Colpitts, thereby increasing th bias current of Q1.

[Chet]

Hi Teleno
I think you are correct on the change in bias current and possible false targets. A threshold level in the software might be a reasonable fix. I found the article below to be interesting. It is based on a Wheatstone bridge circuit.
Have a great day,
Chet


http://213.114.136.42/use/gdo.htm

[Teleno]

In the simulation, the bias point of Q1 does not change with variations of the inductance of +-10%.

In the field, this means the permeability of soil affects the frequency (it becomes lower) but not the bias point. This is already better than a BFO.

[Teleno]

I found a much better oscillator in a russian site.

It's class C and can drive large currents from very low voltages. Amplitude can be controlled by adjusting Re, by varying V1 or by limiting the current supply.

[Teleno]

This oscillator is stable and powerful, and it can work at low voltages.

You guys with LTspice got the .asc (see previous post) ... let's brainstorm ideas for a control loop (digital or analog) to keep the oscillatior's amplitude constant.

Anyone?

[green]

I found a much better oscillator in a russian site.

It's class C and can drive large currents from very low voltages. Amplitude can be controlled by adjusting Re, by varying V1 or by limiting the current supply.

Click on Re I get +- 500ma. Click on Q1, Q2 I get just over 10ma. What is the current path for Re?

[Teleno]

Click on Re I get +- 500ma. Click on Q1, Q2 I get just over 10ma. What is the current path for Re?

I can't duplicate that.

[green]

I can't duplicate that.

My error. Tried it 2 or 3 times yesterday and got the base lead by mistake. Not thinking.

[green]

[You guys with LTspice got the .asc (see previous post) ... let's brainstorm ideas for a control loop (digital or analog) to keep the oscillatior's amplitude constant.]

What is the advantage of keeping the amplitude constant vs measuring the change in amplitude?

[Teleno]

[You guys with LTspice got the .asc (see previous post) ... let's brainstorm ideas for a control loop (digital or analog) to keep the oscillatior's amplitude constant.]

What is the advantage of keeping the amplitude constant vs measuring the change in amplitude?

1. Temperature compensation. The zero point (no target) does not drift with temperature, preventing saturation of the amplifier with high gain.
2. A linear response. The additional power injected to maintain a constant amplitude is the power transferred to the target.