10
\$\begingroup\$

I am currently using a variable inductor, like this one. I'm trying to find a way to make the adjustment programmatically controlled, kind of like a digital potentiometer. Does such a device exist, or are there other good methods for achieving this? It's being used for matching resonance to an imperfectly manufactured device, which is why it cannot be fixed value.

Edit #1, added schematicenter image description here

\$\endgroup\$
  • \$\begingroup\$ Frequency range is 500kHz to 1MHz. \$\endgroup\$ – jgaro Dec 4 '17 at 20:51
  • \$\begingroup\$ What is imperfect? the Tempco , the initial tolerance or both?? and how much in ppm/'C and %L \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Dec 4 '17 at 20:54
  • \$\begingroup\$ Both. I'm seeing 3x signal amplitude change with temperature when feeding into an ADC. \$\endgroup\$ – jgaro Dec 4 '17 at 21:01
  • \$\begingroup\$ Then you are operating at high Q with poor temp compensation. What are Rs, L, C and f? and also ppm/'C shift in peak? Can you make f self-tuning with positive feedback? That would be the sensible solution. Normally L has NTC and ceramics also but some are PTC. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Dec 4 '17 at 21:15
  • \$\begingroup\$ Thanks for the help! Just added a snapshot of this section of the scheme. \$\endgroup\$ – jgaro Dec 4 '17 at 21:29
7
\$\begingroup\$

You should mention in what frequency range you want to use this.

Since the link points to tunable inductors mainly used in IF filters and oscillators in the 10 - 200 MHz range I'm going to assume that's your intended frequency range as well.

Maybe a digitally controlled inductor could be made but it would be complex, expensive and large. Either you would need many different inductors and switch between them or have an inductor with many taps and switch between those.

That's why almost everyone uses variable capacitors instead. Either voltage controlled (varicap) or digital.

These inductors are almost exclusively used in LC tanks in oscillators and filters. That's LC as in Inductor and Capacitor. The solutions for making a variable capacitor are much simpler and cheaper. Any diode is already a variable capacitor, if you bias it in reverse mode the capacitance changes with the reverse voltage.

The tunable inductors are often tuned in manufacturing to get a coarse tuning, if needed additional fine tuning can be done electronically using varicaps or using digital control which switches on/off banks of small capacitors.

\$\endgroup\$
  • \$\begingroup\$ Thanks- just added the comment. Does the same comment apply for this range? \$\endgroup\$ – jgaro Dec 4 '17 at 20:52
  • \$\begingroup\$ Even more so because at 500 kHz - 1 MHz you will need a large inductor to be of any use. \$\endgroup\$ – Bimpelrekkie Dec 4 '17 at 20:53
7
\$\begingroup\$

Instead of using an actual inductor for this. Use a combination of 3 parts.

So first you start with some small capacitance and multiply it with a digital potentiometer. Then you turn this capacitance into inductance. And now you're done.

Alternatively, solve it in software, use a µC(microcontroller), measure voltage with ADC, calculate what the voltage should be with X µH. Put that output on a DAC. Can easily be implemented in software with a digital filter. Hmm, this wouldn't work in this setup now when I'm thinking about it. The input is the same as the output of this inductor. Also an inductor doesn't have several MΩ impedance, as an ADC would. However replacing the entire hardware filter with software (DSP) would make sense. Then it's just a matter of fiddling with a register if you want to tune something.

But in my opinion, solving it in hardware means there won't be any aliasing issues, no need to put some lowpass filter on the input and etc. Also, if you choose to solve it with digital filters, then you should go to DSP.stackexchange.

Here's a schematic that will solve it in hardware:

enter image description here

  • The left graph = input (CLK)
  • The middle graph = The output at the real inductor
  • The right graph = The output of a capacitor + multiplier + gyrator

Replace the potentiometers with digital potentiometers and you're set. You will need some good op-amps that has high bandwidth (probably around 10-100 MHz).

Here's the link in case you want to simulate it in your web browser.


I just realized that you don't need a capacitance multiplier, there is already a multiplier in the gyrator.

enter image description here

  • The left graph = input (CLK)
  • The middle graph = The output at the real inductor
  • The right graph = The output of a capacitor + gyrator

Replace the potentiometers with digital potentiometers and you're set. You will need some good op-amps that has high bandwidth (probably around 10-100 MHz).

And here's the link for this one.

\$\endgroup\$
  • 1
    \$\begingroup\$ This is the proper answer for the question as asked. (Although for the concrete application that the OP wants this for, a small variable capacitor is probably more practical.) \$\endgroup\$ – leftaroundabout Dec 5 '17 at 10:45
  • \$\begingroup\$ @leftaroundabout It's either a gyrator + a digital potentiometer (turn capacitor into inductor). Or a capacitor multiplier + a digital potentiometer. (turn capacitor into larger capacitor). Or, a variable capacitor where you physically trim it. - There's so many viable solutions. \$\endgroup\$ – Harry Svensson Dec 7 '17 at 6:08
6
\$\begingroup\$

You can use a nonlinear magnetic material with a DC bias current to make variable inductance, in a small range. These have been called 'increductors'. In the vacuum-tube era (1950?) this effect was in common use. There aren't any current suppliers of off-the-shelf increductors as prebuilt components, but descriptions are available: current controlled inductor

\$\endgroup\$
6
\$\begingroup\$

The main way to make a variable inductor is to have a coil and insert a core only partially int it. A programmable way to change it would be to have a stepper motor control the slide/core. This would be bulky and basic, but would do what you're after.

If it were me though, I'd definitely try and figure out if a variable capacitor could work in the circuit since it's likely smaller and easier to design around.

\$\endgroup\$
  • \$\begingroup\$ +1 I did the same thing when I was turning a variac into a digitally-controlled one :) \$\endgroup\$ – Rohat Kılıç Dec 5 '17 at 4:54

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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