PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2019-03-22, 16:38:45
News: A feature is available which provides a place all members can chat, either publicly or privately.
There is also a "Shout" feature on each page. Only available to members.

Pages: 1 [2]
Author Topic: Anomalous CMC Tests  (Read 994 times)
Group: Elite
Hero Member
******

Posts: 3426
It's turtles all the way down
I ran a few LTSpice sims and found that with perfect K=1 coupling and some generator impedance e.g. 0.1 ohm there is a near perfect null of the voltage at the tap, as would be expected with out of phase windings.
The generator is basically feeding into a short circuit (as F6FLT mentioned earlier)

The interesting thing is: with any value of leakage inductance between 0.9 and 0.99, you always get exactly half the applied voltage at the tap. I did not try lower values, but suspect it would not change.

As I suspected before I ran the tests, leakage inductance (any amount as long as it is equal which it probably must be) will always yield 1/2 applied voltage at the tap, the same as would be achieved as expected with normal auto-transformer (non-bucking) coils. So the uncoupled(leakage) inductance performs the function of auto transformer action or inductive divider.
Capacitive effects will certainly disturb the divider balance.

I suspect any large variations at the tap that yield voltage other than 1/2 is due to variations in capacitance coupling and may be vetted by low frequency testing on the bench or a build that carefully matches capacitance coupling and distribution, not only to the core, but between turns. Other factors may be the distribution of leakage inductance that may create an imbalance generating much higher voltage at the tap, but I'm still trying to wrap my head around how there could possibly be a difference in leakage inductance, because the windings share the leakage. CMC's would be prone to have flux leakage paths where the windings end and create a space, especially when in the severe bucking mode....the flux needs to exit the core at these spots.

I won't bother to post the sims since they are basic and you guys can run it live to see the results.

Naturally, I may be way off base, being new to this thread. I'll see what else I can run to add to the hypothesis.

Regards

edit: tested down to K=0.1, and V at tap =1/2 applied V

P.S exercise care with this type of testing on the bench because the out of phase windings present a "worst case scenario" to the driving generator or amplifier. Use a limiting device.
« Last Edit: 2019-01-07, 03:07:55 by ion »


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Group: Elite
Hero Member
******

Posts: 3426
It's turtles all the way down
With further testing, I removed the 0.1 ohms from the generator source and added 2 x 0.1 ohm  to the center tap. Setting the K =1, (eliminating leakage inductance) now I can get exactly half voltage at the tap instead of a complete null as in previous tests.

This proves that the DCR of the windings forms a voltage divider with the current that is forced into it by the bucking windings and generator source. Doesn't matter what the resistors are, as long as they are equal the voltage will always divide to one half whereas we would expect a null at the tap.

Conjecture: leakage inductance or DC resistance both create the same effect of 1/2 V at the tap. Since it is likely that the winding resistance would be very close to equal if the coil is carefully made, the effect we would think would be more likely due to large differences in leakage inductance, however this is counter intuitive and we would expect it also to be fairly equal. A deviation from V=1/2 at the tap would probably have to be attributed to large differences in capacitance of the windings.

We can try modelling this as two uncoupled inductors (leakage inductance with some stray capacitance and DCR), since in a well made CMC, the coupled inductance is antiphase, making the predominant inductance a null (invisible).

This will have to be bench tested to confirm the hypothesis, however maybe I can try a sim.

"When you have eliminated the impossible, whatever remains, however improbable, must be the truth". – Sir Arthur Conan Doyle, stated by Sherlock Holmes

edit: in CMC anom3 I added leakage inductance externally and set the K of the coupled inductors back to 1. Again we have vtap =1/2. I tested with 1k and lowered the frequency by 10x to 100Hz. Now to see the effects of a capacitive divider. Capacitance will be much more difficult to control in a hand wound CMC.
« Last Edit: 2019-01-07, 04:20:20 by ion »


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Full Member
***

Posts: 235
Hi Ion,

I came to the same conclusion. See below for the two equivalent simulations.

In my own tests, Vout dropped due to the impedance of the oscilloscope probe when the frequency is high. The higher the frequency, the lower the current, due to the higher impedance of the two leakage coils (the part of the coils that are not coupled: L3 and L4 in my model).

It should be noted that the current flowing through the coils is due solely to leakage inductances. If the two coils were perfectly coupled, then no current could flow, because any current in one coil would induce an equal but opposite current in the other.


---------------------------
"Chance favours only the prepared mind."  Louis Pasteur
   
Group: Elite
Hero Member
******

Posts: 3426
It's turtles all the way down
Hi Ion,

I came to the same conclusion. See below for the two equivalent simulations.

In my own tests, Vout dropped due to the impedance of the oscilloscope probe when the frequency is high. The higher the frequency, the lower the current, due to the higher impedance of the two leakage coils (the part of the coils that are not coupled: L3 and L4 in my model).

It should be noted that the current flowing through the coils is due solely to leakage inductances. If the two coils were perfectly coupled, then no current could flow, because any current in one coil would induce an equal but opposite current in the other.

Hi F6FLT

Agreed

I think where it will get interesting is when one or both of the leakage inductors become resonant with the stray capacities involved.

We can expect some high voltage to appear at the tap under this condition, especially during transition when different resonant points appear on each.

You also stated:
Quote
The temporary conclusion is therefore that the cables and/or the ground of the measuring equipment drastically reduce the voltage measured at the mid-point of the coils, for a reason that is not yet clear but remains very interesting. Indeed, LTspice simulations indicate that it is neither the capacitance of the measurement probe nor its internal resistance that can cause the voltage to drop so much.

This may possibly be explained by the probes shifting resonant points by adding their capacitance, in light of the leakage inductance/stray capacitance resonance hypothesis.

I will add the fourth sim test that includes stray capacitance later and possibly along with some bench tests.

Regards


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Group: Elite
Hero Member
******

Posts: 3426
It's turtles all the way down
Here is the circuit with added stray capacitance.

Resonant frequencies for the two are approximately 1.59 MHz, and 2.25 MHz. With the FG set for 1.83 MHz, we see  at the tap a peaking of over 36 Volts.

This is an extreme case of 2:1 difference in stray capacitance, in practice it may be much less.

edit: corrected lower resonant frequency

Conjecture: This does lead to some interesting thoughts e.g. what happens to the atoms of a ferrite material (or for that matter the ether itself) if two or more precise LC oscillators that are coupled in some way (and properly juxtaposed?) and while stressing the medium hit upon some "magic" frequencies. Would this create another "imploding TV of SM effect?" The CRT TV yoke and driving circuits satisfy some of the "possibly required" ingredients.

P.S. Bench testing with a CMC,  I found that the addition of some capacitance to one side changes the resonant frequency and can give 2x to 4x output at the tap. Doesn't matter which side the capacitance is added to. This tells me the choke windings may share the  leakage inductance, across which the resonant voltage can rise.
« Last Edit: 2019-01-07, 18:46:09 by ion »


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   

Hero Member
*****

Posts: 2081
Quote:-

Conjecture: This does lead to some interesting thoughts e.g. what happens to the atoms of a ferrite material (or for that matter the ether itself) if two or more precise LC oscillators that are coupled in some way (and properly juxtaposed?) and while stressing the medium hit upon some "magic" frequencies. Would this create another "imploding TV of SM effect?" The CRT TV yoke and driving circuits satisfy some of the "possibly required" ingredients.

So right I think Ion, I will have some spectrum analyzer shots soon on that very subject O0

Regards

Mike 8)


---------------------------
"All truth passes through three stages. First, it is ridiculed, second it is violently opposed, and third, it is accepted as self-evident."
Arthur Schopenhauer, Philosopher, 1788-1860

As a general rule, the most successful person in life is the person that has the best information.
   
Full Member
***

Posts: 235
...
   
Quote from:  F6FLT
    ...Indeed, LTspice simulations indicate that it is neither the capacitance of the measurement probe nor its internal resistance that can cause the voltage to drop so much.
   
This may possibly be explained by the probes shifting resonant points by adding their capacitance, in light of the leakage inductance/stray capacitance resonance hypothesis.
...

I should have removed that mention. In fact, LTspice was right. I didn't choose the right parameters, the frequency was not high enough according to the inductance of the coils. After correction, LTspice works and shows what I observe in reality.  I don't see any more anomalies.  >:(



---------------------------
"Chance favours only the prepared mind."  Louis Pasteur
   
Full Member
***

Posts: 235
...
Conjecture: This does lead to some interesting thoughts e.g. what happens to the atoms of a ferrite material (or for that matter the ether itself) if two or more precise LC oscillators that are coupled in some way (and properly juxtaposed?) and while stressing the medium hit upon some "magic" frequencies. Would this create another "imploding TV of SM effect?" The CRT TV yoke and driving circuits satisfy some of the "possibly required" ingredients.

P.S. Bench testing with a CMC,  I found that the addition of some capacitance to one side changes the resonant frequency and can give 2x to 4x output at the tap. Doesn't matter which side the capacitance is added to. This tells me the choke windings may share the  leakage inductance, across which the resonant voltage can rise.

Until the material is non-linear, the fields simply add up, so we only have a superposition (no "modulation", no "multiplication"). This means that you will never see signals at other frequencies than that of the generators, evidence of a linear operation.

If the material is non-linear, which you implicitly assume when you talk about "stressing the medium", the above statement still applies, but only at each instant of time. This means that you have to use the non constant values L(t), C(t), µ(t), ε(t) and so on in the equations and to integrate to know the real effect. Then new frequencies appear in the system, that are not provided by the generators. But these parametric changes have an energy cost equal to the possible gain, and we also have generally higher losses than in a linear mode.

Nevertheless the idea remains interesting if the energy for the parametric change can be taken from the environment, like heat (see the thread "non linear capacitor"). Heat is a problem because we don't see how to heat up/cold down a material at a sufficient repetition rate to get useful energy. Now there are perhaps ambient energy sources other than heat that we could tap. If we make this supposition, then the most highly non-linear material will be the best choice for our purpose and the stronger stress the best method.



---------------------------
"Chance favours only the prepared mind."  Louis Pasteur
   
Full Member
***

Posts: 235
Another strange but explainable effect is when two opposite coils are well coupled on a ferrite, and you gradually remove the ferrite (for example a ferrite rod): the resonance frequency may decrease instead of increase.

This is due to the fact that the two opposite coils in series have an inductance M=2L*(1-K) where K is the mutual induction coefficient.
When you remove the ferrite, even if each individual inductance decreases because the core is removed, M increases because the coils are less coupled, and therefore the series inductance increases and tends towards 2L.
I observed it again yesterday. This may have possible applications, for example it allows to reverse a parametric variation (decreasing the permeability µ will increase the inductance), I started to think about it.
« Last Edit: 2019-01-09, 08:49:33 by F6FLT »


---------------------------
"Chance favours only the prepared mind."  Louis Pasteur
   
Group: Elite
Hero Member
******

Posts: 3426
It's turtles all the way down
Until the material is non-linear, the fields simply add up, so we only have a superposition (no "modulation", no "multiplication"). This means that you will never see signals at other frequencies than that of the generators, evidence of a linear operation.

If the material is non-linear, which you implicitly assume when you talk about "stressing the medium", the above statement still applies, but only at each instant of time. This means that you have to use the non constant values L(t), C(t), µ(t), ε(t) and so on in the equations and to integrate to know the real effect. Then new frequencies appear in the system, that are not provided by the generators. But these parametric changes have an energy cost equal to the possible gain, and we also have generally higher losses than in a linear mode.

Nevertheless the idea remains interesting if the energy for the parametric change can be taken from the environment, like heat (see the thread "non linear capacitor"). Heat is a problem because we don't see how to heat up/cold down a material at a sufficient repetition rate to get useful energy. Now there are perhaps ambient energy sources other than heat that we could tap. If we make this supposition, then the most highly non-linear material will be the best choice for our purpose and the stronger stress the best method.

To clarify, with my conjecture, I was not implying that extra energy could be taken from the environment (ferrite atoms or ether itself) although this may be possible. I was rather stating the possibility of a coupling that could force an accumulation of energy in a vortex structure that could produce effects not seen before.

As an example, in the TV CRT we have an electron beam in a vacuum being accelerated at close to light speed. This beam passes through the yoke coils and normally ignites the phosphors of the screen in a well controlled manner by the yoke producing normal images.
What happens when a failure of the yoke's driving electronics or the coil itself whips the electron beam into a tight circle of high speed? I know this is far out conjecture and would need to be proven by experiment. Thus I have saved some CRT devices for future experiments along these lines. There were a couple of people that were experimenting with using a TV yoke to provide vortex structures in liquids if memory serves correct. This was a magnetic effect. I can no longer find these experiments on the web. Others have suggested the conjecture I proposed so it is not new, but I don't know of any other experiments along these lines.

Regards


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Group: Elite
Hero Member
******

Posts: 3426
It's turtles all the way down
Another strange but explainable effect is when two opposite coils are well coupled on a ferrite, and you gradually remove the ferrite (for example a ferrite rod): the resonance frequency may increase instead of decrease.

This is due to the fact that the two opposite coils in series have an inductance M=2L*(K-1) where K is the mutual induction coefficient.
When you remove the ferrite, even if each individual inductance decreases because the core is removed, M increases because the coils are less coupled, and therefore the series inductance increases and tends towards 2L.
I observed it again yesterday. This may have possible applications, for example it allows to reverse a parametric variation (decreasing the permeability µ will increase the inductance), I started to think about it.

A good observation. I wonder what energy is required to move the ferrite rod since the windings oppose?

At any rate I think I have derailed the thread with the imploding TV hypothesis, so will save that for the appropriate thread.

Apologies to partzman and all.


---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Group: Elite
Hero Member
******

Posts: 3426
It's turtles all the way down
In LTSpice, the opposed windings are a null.

In these tests I have removed the opposed windings leaving only two inductors (what would constitute the leakage inductors) and left in the stray capacitance at 2:1, 100pF and 50pF

Around 30 volts is at the tap for 1 volt input.

You can try this with much smaller ratio, e.g 100pf to 95 pf proving a small difference in stray capacitance produces a increase in voltage at the tap from the expected value.



---------------------------
"Secrecy, secret societies and secret groups have always been repugnant to a free and open society"......John F Kennedy
   
Full Member
***

Posts: 235
A good observation. I wonder what energy is required to move the ferrite rod since the windings oppose?

At any rate I think I have derailed the thread with the imploding TV hypothesis, so will save that for the appropriate thread.

Apologies to partzman and all.
Hi Ion,
You understood what I meant, but I wrote the opposite. I corrected my post that you quoted:
[when you gradually remove the ferrite] "the resonance frequency may decrease instead of increase"
I also corrected a minor error of sign in the formula, it is  M=2L*(1-K)  (not K-1 as written the first time).
Sorry




---------------------------
"Chance favours only the prepared mind."  Louis Pasteur
   
Pages: 1 [2]
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2019-03-22, 16:38:45