Magnetic permeability coil tuning, pros and cons?

[neazoi]

Hi, I have made an oscillator shown here **broken link removed**
I have noticed that by bringing a magnet close to molded chokes used for the tuning coil, I am able to achieve a big tuning range, about 5-18MHz with a single coil.
It seems a small magnet can do the job of a big (costly and bulky) tuning capacitor.

What are the pros and cons of this tuning method anyone can think?
Why it is not commonly used instead of variable capacitors?

 

[BradtheRad]

I suppose you changed the Henry value of an inductor temporarily?
This concept could be useful. A fixed inductor becomes variable.

I may resemble the concept of the magnetic amplifier. By sending DC current through an auxiliary winding, you can alter how much current goes through the main winding.

However if the inductor contains a metal core, then there is the question whether the magnet leaves some permanent effect after it's removed.

Also whether it reduces the saturation point.

Also whether it affects other components on a board, such as inductors.

 

[neazoi]

I suppose you changed the Henry value of an inductor temporarily?
This concept could be useful. A fixed inductor becomes variable.

I may resemble the concept of the magnetic amplifier. By sending DC current through an auxiliary winding, you can alter how much current goes through the main winding.

However if the inductor contains a metal core, then there is the question whether the magnet leaves some permanent effect after it's removed.

Also whether it reduces the saturation point.
Also whether it affects other components on a board, such as inductors.

Yes you affect the magnetic permeability of the core, it is like taking out flux with the magnet and making the value of the indictor less.
In this specific circuit there are no other inductors so there is not an effect of the magnet in other components. In other circuits, containing other inductors as well, either you have to shield the tuning circuit from the rest of the circuit, or use toroidal inductors for the rest, because they are not affected, unless the magnet is very close to them.
In fact, if using toroidal inductors for the tuning coils, the tuning range is much much less because the flux is mainly concentrated inside the core.
Yes the core is saturated at some point, but that's not a problem in this specific circuit, it still produces a nice sinewave.
I have not noticed any serious residual permanent magnetization, at least on coils with a ferromagnetic core.

I am considered about the characteristics of the inductor before and after the magneric field. Thinking locically it's resistance stays the same, only it's inductance is changing?
I think I have seen somewhere that the Q of the coil also changes, which could lead to more instability/noise in higher frequencies?

 

[volker@muehlhaus]

Yes you affect the magnetic permeability of the core, it is like taking out flux with the magnet and making the value of the indictor less.

Is it really the static magnetic field that tunes your inductor (indirect through saturation) or just the permeability (µr) property of the magnet material that changes fields?

If it is the permeability, inductance should increase with the magnet near to it.
If it is the static magnetic field -> saturation, inductance should decrease with the magnet near to it.

 

[neazoi]

Is it really the static magnetic field that tunes your inductor (indirect through saturation) or just the permeability (µr) property of the magnet material that changes fields?

If it is the permeability, inductance should increase with the magnet near to it.
If it is the static magnetic field -> saturation, inductance should decrease with the magnet near to it.

No it is not the permeability of the magnet material, it is the effect of the magnetic field of the magnet onto the core of the coil that causes the total inductance of the coil to decrease. If you bring the magnet closer to the coil core, the frequency of the oscillator increases, which verifies that the inductange decreases.

As I said, I am not sure about the Q degradation of the coil as a result of the core flux alteration. However this has to do mostly with frequency stability and phase noise. Nothing beats a crystal at that point of course but it would be interesting to know what is really hapenning in the inductor.

 

[volker@muehlhaus]

No it is not the permeability of the magnet material, it is the effect of the magnetic field of the magnet onto the core of the coil that causes the total inductance of the coil to decrease. If you bring the magnet closer to the coil core, the frequency of the oscillator increases, which verifies that the inductange decreases.

Ok!

As I said, I am not sure about the Q degradation of the coil as a result of the core flux alteration. However this has to do mostly with frequency stability and phase noise.

If you can tune the RF inductance with a DC magnetic field, then the core is operating outside the linear range (saturated). This indeed causes extra loss (lower Q), as you have noticed. That's bad for use in oscillators etc. because lower Q results in less frequency stability and higher phase noise.

 

[neazoi]

Ok!
If you can tune the RF inductance with a DC magnetic field, then the core is operating outside the linear range (saturated). This indeed causes extra loss (lower Q), as you have noticed. That's bad for use in oscillators etc. because lower Q results in less frequency stability and higher phase noise.

I think that frequency stability is not expected to be so good with an LC anyway. However there are solutions about this if an electromagnet is used instead of a magnet http://www.hanssummers.com/huffpuff/fast.html but the circuit becomes more complex.
And one could use another winding onto the core fed with DC, to create a core saturation instead of a permanent magnet, just like magnetic amplifiers. But in situations where multiple plugin coils are exchanged in the same circuit, this is more difficult.

I do not think the core is always saturated. It depends on the strength of the external magnetic field applied to it. As this field increases, the H point increases in the BH inductor curve and the core is still operating in linear mode. However when this field increases even more up to a threshold, the H reaches into the non-linear region of the inductor indeed.

In my circuit, if I bring the magnet very close to the inductor, I notice that some harmonics start to appear in the frequency domain. I believe these are due to the inductor operating in it's non-linear mode, but I cannot verify this. Logically thinking, it could be caused by the inductor, starting to operate in it's non-linear region. In fact, there were systems at the past that used non-linear inductors (already saturated) to generate harmonics of a fundamental.

What do you think?

 

[vfone]

The only advantage using a magnetically biased inductor is to get higher inductance for the same inductor volume (almost double of inductance, as using standard cores).
Thermal aging for magnetically biased inductors is a big challenge, because is hard to get a constant magnetic field over time (even for small periods of time).

 

[neazoi]

The only advantage using a magnetically biased inductor is to get higher inductance for the same inductor volume (almost double of inductance, as using standard cores).
Thermal aging for magnetically biased inductors is a big challenge, because is hard to get a constant magnetic field over time (even for small periods of time).

Very interesting, that was the sort of info I was looking for.
I am a bit confused, I thought that if I apply greated magnetic field, the inductance of the coil decreases, not increases. Does it have to do with linear and non-linear modes?

The information I have found about thermal stability, relate more to the inductors used in power electronics, where thermal stability is of concern. Do you think that in the very low currents involved in this oscillator it would be of much concern?

I am also not sure about my previous assumption if the inductor operates in linear or non-linear mode.

 

[volker@muehlhaus]

I think that frequency stability is not expected to be so good with an LC anyway.

True, but I think this would be worse than "normal" inductors. Also, if you use a magnet to tune the inductor by changing the distance, it would be very sensitve to any movements (vibration, microphone effect).

I do not think the core is always saturated. It depends on the strength of the external magnetic field applied to it. As this field increases, the H point increases in the BH inductor curve and the core is still operating in linear mode.

My view: If the inductor is linear, there should be zero effect from fields or currents at other frequencies. If your static field tunes the RF inductance, it is no longer linear. But if the DC field (bias) is much larger than the RF field, it might be "almost linear" for the small RF signal.

 

[chuckey]

Marconi Instruments used this system to get FM on their signal generators.
Frank

 

[neazoi]

Marconi Instruments used this system to get FM on their signal generators.
Frank

Did they achieve FM by some magnetic means? Or you mean to generate the carrier frequency in an FM modulator?

 

[chuckey]

The oscillator coil was wound on a ferrite core which sat in the gap where the centre limb of E I transformer core would be, with the audio windings on the outside limbs. Its along time ago (1960s valve circuit), but thinking of it now, you could also use the DC through the windings to fine tune the frequency or apply AFC.
Frank

 

[mtwieg]

I would assume that using a powdered iron material with soft saturation characteristics would give a smoother response and better Q. Depends on the frequency though.