Dynamically Changing Inductor Value In RLC AC Circuit

Adding inductance lowers the resonant frequency. Are you sure this is what you want to do?

That's a good point. Was going to adjust the excitation frequency as the inductance changed. Am just interested in how it might be done. I'm sure you could use FETs but is that the easiest/best option?

Please describe the circuit in more detail. One of the options is to post an image of it in a following post.

There is no indication of what this circuit does yet, either.

How is the forum supposed to help given the scant detail available to date (rhetorical question - NNTR)?

Ok. Think metal detector, high current (1A +), low resistance in the RLC circuit. I want to add another inductor in series with the series RLC circuit using a digital signal (ideally). Increasing or decreasing the inductance increases or decreases the magnetic field generated by the coil and changes the search depth of the device.

I haven't created the circuit as yet - am just trying to get my head around the possible options - and to evaluate if it's worth the effort.

Thanks

I'm not sure I understand the difference between doubling the coil inductance and doubling the current. If you want to vary the sensitivity to go deep or go shallow, why not design for deep, and when you want less sensitivity for shallow, reduce the signal intensity?

Changing the current is a whole lot easier than changing the effective coil size. Or are these series coils of different size or geometry?

A metal detector has to be a holistic design. The physical configuration is probably more important than the electronics.

You could use an analog selector switch to connect to tap points of inductors in series using 4051B shown below. Analog selector switches are also available in other sizes, and for many taps, you can connect several channel-out-to-common in a tree structure.

http://www.ti.com/product/CD4051B-MIL/datasheet

Just a note: the 4051 has a fairly high 'on' resistance whose value can range from 125Ω to 1200Ω depending on supply voltage, part variation and mfr. For ST Microelectronics' HCF4051 (pdf), for example, at V_DD = 15 V the 'on' resistance can range from 125Ω (typ) to 280Ω (max), depending on the part.

Good point. No telling what the requirements are. Maybe just a manual rotary switch would be acceptable.

http://www.ckswitches.com/media/1349/arotary.pdf

Yes, either a rotary switch or, if he wants digital control, relays with transistor drivers are fairly straightforward, wired as a single-pole, multi-throw switch.

He needs the switch to have a low resistance so as not to load the circuit, killing the Q, as well as being able handle the current. If he wants high Q, those contacts will carry some current, OR, he could reduce the driving signal to something much less than 24 V_pp to reduce the current to something more manageable whilst maintaining a high Q.

On the flip-side, he mentioned something of the (aggregate) coil producing an "excitation" field. I'm not sure what that means exactly but a metal-detector comes to mind, in which case he would want that current as well.

Yip, Metal detector is along the right lines. The 'excitation field' is the magnetic field projected by the coil for 'scanning' metal objects, so I want high current and low resistance. Am aiming for a current around 1A or higher if possible.

Ideally I would like to use a digital solution (so no analog switches) so it can automatically 'switch' during the scan. I haven't done much with AC signals so I'm not sure how complicated the solution would be - just hope it's nothing like an audio amplifier.

The idea is to change the inductance as the required search depth changes - the benefit of this is the spatial resolution of the detector should improve over the range of depths.

You may want to check the specs on silicon carbide or gallium nitride mosfets. I believe either of these devices will handle your current and low on resistance as well.

What kind of arrangement would you suggest? I don't want to invert the +/- 12V sinusoidal signal.

He's just suggesting you use them as solid state relays.

solid state relay 2A ±12V

This search for mine detector excitation field frequency

might also help

You mentioned dynamically changing the inductance possibly FETs in post 2 and someone else mentioned using a selector switch arrangement to set inductor values a few posts later. Instead of inverting, my thoughts fell along using mosfets in a selector switch arrangement that allows configuration in real time (in case you need fast responses)

Good point (about the requirements). Sometimes I skip the general overview and start saying things like 'excitation field' without any context - friends point this out all the time. I've added a few more comments about the project - hope that's enough. Ideally the switching can be initiated by a digital signal, allowing the coil to be added without manual input during a 'scan' for the metal object Thanks for the feedback.

"The inductor is a multi-turn coil which creates an excitation field."

An 'excitation field' for what? Are you saying this coil serves another purpose in addition to its role in setting the resonant frequency?

As 10 kHz is fairly low, what sort of coil construction are you using? Are these ferrite-core toroids or something?

Also, what is your R value? If it is very low (a few ohms) you'll have be careful in your choice of switch.

Hi Andrew, The coil's main purpose is to project a magnetic field for finding metal objects (like you guessed :) ) The resonant circuit is simply to reduce the resistance (and increase the current) in the circuit. Coil construction would be ferrite core -with several 'independent' coils on the same core, adding the coils (to the circuit) as required via the 'switching' - that's the idea anyway.

What sort of current are you talking about?

I think I'd err on the safe side and use relays.

Texas Instruments TS5A3166 is a 100mA, 0.9Ω, 1.65 to 5.5V supply analogue switch.

Yep. At resonance an RLC looks like this. Depending on the value of R that current could be an amp or more.

Thank you - that's a perfect representation of the circuit

Ideally 1A plus. Thanks for the reply.

Instead of adding or subtracting turns to your search coil to decrease/increase the magnetic field (varying the ampere-turns basically), used a fixed coil and vary the current (ie, your 'excitation level') to vary the field strength. A lot less complicated.

What is your objection to using an audio amplifier? You can buy fairly powerful audio amplifier chips/modules off-the-shelf for cheap.

Also curious about your choice of frequency: 10 kHz. Why so low? Metal detectors typically operate in the 100s of kHz.

Yip was going to do that as well (vary the current).

By adding a coil I can change the inductance (& magnetic field) and more importantly I can change the diameter of the coil. As you've suggested large diameter coils are best for deep penetration but they are not ideal for shallower objects & visa versa. The larger the diameter the lower the lateral resolution, so using a single large coil can result in lower spatial resolution.

As for the audio amp idea - no objections, I just remember trying to build a D-Class and it was tricky to generate a distortion free output.

I agree it's starting to sound complicated - just trying to understand how all the potential variants (in my head) could be implemented. Changing the coil diameter seemed like a good way to use the optimal coil diameter and maximise spatial resolution.

Thanks for hanging in there & giving heaps of feedback!

Couldn't you change the inductance with a moving slug core like in the old AM radios?

Hey cool....old school....

http://www.xtalman.com/kits.html

Both the slug-tuned ferrite loopstick and tapped coil are great for radios; not so good for metal detectors. For these you want a large-area, ideally-planar loop.

Maybe a combination VLF and pulse induction detector would work more efficiently....?

https://www.lammertbies.nl/electronics/PI_metal_detector.html

http://www2.gi.alaska.edu/~jesse/treasure/misc/howdetector.html

Wouldn't it be easier to switch capacitors and leave the inductive loop as a constant?

Some reed switches can carry 1 Amp (eg. Digikey 374-1127-ND) and can easily be controlled by digital circuitry.

Hi Guys,

Options that I can see to switch the AC signal.

1. I found this post which uses a single MOSFET and a bridge rectifier as an AC switch. The main disadvantage I can see is the voltage drop across the diode.

2. Bidirectional switch using two back-to-back P-MOSFETs

3. TRIACs ?

Any thoughts on which would be better for this circuit?

Well, why don't you try them all out and report back on your findings?

Hi Win, I'm a bit confused by the above diagram.

According to the data sheets I've found TLP191B is an opto coupler; the PVI5033R is a dual Photovoltaic Isolator (I can't find any data sheet for the non R version).

So, should the U1 and U2 be labelled the other way round?

In any case the relevant part of the circuit as far as the OPs concerned is just the Q3 Q4 part, as long as he can work out a suitable way to drive it?

I think it may be more difficult to make an continuously adjustable metal detector than you think. It isn't enough to change the inductance (and frequency) of one coil.

Unless you are using a pulse system, you need need to be adjusting two coils to have what would be the same inductance in the lack of metal case. Any difference in adjustment between the two will result in 'beats' and give false indication.