Parametric Excitation

[Raui]

I remember my physics teacher earlier this year talking about a theremin when I was talking to him about electrostatic fields. I don't believe it's the inductance of the machine that changes, it's the capacitance that changes. It is an interesting thought though

Raui

[SuperCaviTationIstic]

sorry, you're right, it changes capacitance....

I've just been caught up in designing one that changes inductance with the antennas
(or perhaps reactance)
Use the Force!

[Raui]

Alls good, we all get mixed up in our thoughts at times I'm trying to get around to making a video of a setup I'm making which is much like how I'm picturing LtBolo's setup with a cut MOT transformer core, atleast until I can get my aluminum plates which might be a while off bit short on money. Can you go into detail about your antenna setup? Sounds interesting!

Raui

[SuperCaviTationIstic]

i was considering patenting it. but **** it. you heard it here first

Open source

Frequency is controlled by a bucking conical antenna,
Volume is with a bucking toroid.

as many turns of wire as you can fit on each
That simple.

Something tells me if I just make 1 single coil in the shape of a MUSHROOM i can do both at the same time!

[broli]

I want to share a more complete design of a presentation I made:

http://ziosproject.com/NJ/magPres/index.htm

I have attached the complete design. The motor looks very similar to the SG bedini motor. But this design is not a motor and should be driven BY a motor perhaps a bedini motor would do fine .

Anyway if you get the presentation you should get this design as well. Magnet attracts to core, it saturates the core, coil is energized with low inductance, magnet moves away without any bEMF because the already present current makes sure the core is kept saturated. Then when the magnet is far enough, ie it no longer affects the core, the coil is discharged.

In the design you might wonder why the use of a single magnet on the rotor. I did some quick calculations and found that to power an average load like a bulb you'll need quite some time for the coil to discharge completely, so it might be safer to start with one magnet, however multiple coils can be employed around the rotor.

And since it's cooler to name your inventions I called it the VIG, variable inductance generator. But don't be mistaken this should load the motor running it but imo far less than the huge increase in energy from the changed inductance. It's very easy to make a mechanical energy test between the core and magnet when current is off, and the same test when current is on. This test can be done per degree using a spring weight or so.

But I hope to be able to build this design.

[LtBolo]

Our parametric excitation experiments have continued. It is easy to make power, but so far everything we do loads the motor shaft. It appears that phase is critical, meaning, the parameter change needs to be the correct width to force the phase to where it can work properly.

I am also of the mind that the parameter change needs to be very quick and sharp in both directions, and should be as flat as possible when not changing. It may not generate useful power until the parm change is faster than the normal slew rate for that parameter. That becomes an interesting optimization problem though, since there is no single resonant freq, there are two, and the amount of each is a function of the parm change. I am starting to see that the shaft will only unload when you get it exactly right, but I am still pretty sure that the shaft will unload.

The fun continues...

[boguslaw]

What param are you changing ?

[broli]

What param are you changing ? You have to change it back and forth, in other words : add to it at correct phase and set to original when LC circuit is attempting to convert it back to other side.There would be then only one resonant frequency just pumping action would be going with different frequency but in phase with circuit oscillations.
All other ways are worthless, because once you change frequency the input energy is consumed doing this.
It's like a poking child swing at the peak of amplitude in correct direction, you cannot do it at any other time. Resonant frequency will not change , only voltage or current = power will accumulate.

This thought pattern is too simplistic. I tried to derive the current behavior from a continuous change in inductance and the results are far from trivial. You don't end up with nice sine waves.

Imo the best method is abrupt changes. When you see the current is maximum in the inductor you instantly increase inductance not giving the current a chance to react to it. The same the moment you see the current becomes 0 you quickly change to a low inductance. The real issue is increasing inductance without flux change. This is done by saturation of the core material. When current is at max the core is "double" saturated by the current and by an external magnet, at this point you also remove the magnet "quickly". The inductor senses no flux change because the domains that were held aligned by the magnet are now being held by the current. You now end up with a changed LC circuit, one with a higher L. Meaning longer discharge time.

The method I proposed doesn't need an LC tank. It's a very discrete solution to the problem. There is some time period when the magnet is leaving the core the current is flowing in the coil waiting for the magnet to leave far enough to discharge. This time portion is lost as joule heating in the resistance. But if the parameters are chosen right it can be a very small time period and thus low energy loss compared to the massive energy gain through the coil discharge.

[LtBolo]

What param are you changing ?

Inductance, and like broli mentioned, you do not get clean sine waves and most predictions about frequency end up wrong.

[boguslaw]

broli - your way is correct
LTBolo - use electromagnet instead of permanent
no need to move it...

[boguslaw]

Inductance, and like broli mentioned, you do not get clean sine waves and most predictions about frequency end up wrong.

why not ? when you pushing child on swing your predictions end up wrong ? sure - if you do it wrongly ...

[broli]

why not ? when you pushing child on swing your predictions end up wrong ? sure - if you do it wrongly ...

Only experiments can shed light on the truth. The problem with changing induction if the core is not saturated is that the induction change is accompanied by a a flux change. This flux change will cause a back emf in the coil trying to lower the current.

I don't like to deal with that hence core saturation. The above concept wouldn't work if you couldn't saturate cores. Of course cores have coercive losses that's why it's wise to make a good pick. But I'm confident that the energy gain will be plenty if the inductance difference is high enough.

[LtBolo]

why not ? when you pushing child on swing your predictions end up wrong ? sure - if you do it wrongly ...

If this is truly like the swing example, then we are wasting our time. In the swing example, you are decreasing...not increasing...the parameter (pendulum arm) at peak current (speed). The child is fighting centrifugal forces to raise their feet...the system is loaded.

In truth, I think that is what we are seeing...an electronic swing.

[broli]

Here's some crude calculations. First let's consider the complete charge time to be 4*L/R to play it safe.

When the magnet is at TDC the inductance of the coil is 0.1mH. Assuming we have a 30V power supply that can supply 5A then we choose R = 6 Ohm so the charge time is 66µs. Assuming we use neo magnets and our C core gets perhaps saturated 10° before and after TDC. So we give ourselves 10° of charge time. Using our time constant this leads to a constant rpm of 2500 RPM. After this 10° the influence of the magnet on the core starts to diminish, but also meanwhile current is used as joule heating. Let's assume that 45° later the magnet has no more effect on the core and the coil can be discharged. First let's calculate the joule heating energy.

given period = 45°, since rpm = 2500 time is then 300µs. Joule heat enegy = V*I*t = 0.045J.

Notice that I ignored inductive energy input and just put it as a "loss" in joule heating, I'm that confident .

So finally now we can discharge our inductor without any external influence. I took the final inductance to be 0.1H and the load 333.33 ohM. First of all lets calculate the energy of the inductor.
E=0.5*I^2*L=0.5*25*0.1 = 1.25J.

Electric wise the energy out / energy in = 27 or 2700%.

Since this needs again time to discharge. 4*L/R = 1.2ms. From our 2500 rpm rotation that is equal to 180° of rotation. So it will take 180° of rotation to discharge our high inductive coil through the 333.33 Ohm load.

And to finish it all let's look at the power generation. We know it rotates at 2500 rpm and that every rotation steals 0.045J and gains 1.25J. In average watt values this results in 1.87 Watt of loss and 52.08 Watt gain.

And that's just one example of high energy gain from this system. Many parameters can be adjusted to gain any desired cop. I might make a javascript calculator that gives these results automatic given a certain parameters.

[boguslaw]

If this is truly like the swing example, then we are wasting our time. In the swing example, you are decreasing...not increasing...the parameter (pendulum arm) at peak current (speed). The child is fighting centrifugal forces to raise their feet...the system is loaded.

In truth, I think that is what we are seeing...an electronic swing.

I'm not expert but in case of child swing the change is done twice in period and always in the moment of converting kinetic energy into potential. The net result of both changes should be zero (because changes are opposite) but obviously energy is gained in system and it's clear that it is sucked from gravitation field.
The same is purely possible with magnetic field as I described in one old post. Magnetic fields always interacts so when magnetic field collapse in the bubble inside large magnetic field this collapse has the energy gains from two collapsing fields no matter how weak is the large field ;-) Normally we don't see this maybe because much larger looses in circuit which consume energy in a form of voltage spike ?