Effects of Magnetic Bias on Common Mode Chokes

[ion]

I'm reposting this here as I thought it was interesting (formerly I posted it on builders site) Anyone have any ideas what might be going on here? It is one of the stranger effects I've personally seen, but can't quite explain.

Common Mode Chokes: Biasing and Noise Effects

Here is an interesting passive circuit I've been fooling with. It is a pair of identical common mode chokes 45 mH inductance (one side). The windings are wired so that each choke has its dual windings in series additive mode, making a composite inductor.The first choke is the input choke, the second choke has two capacitors to ground thus forming a pi network.

This circuit behaves interestingly with no magnet on the pi filter. As you sweep up towards 1kHz, the inductors begin to saturate at resonance and you get a large square wave when driven with sine input. Go a little over 1 kHz and the output snaps off and the signal on the output is now very low. The circuit will not recover untill you start very low in frequency and sweep up towards 1 kHz again.

With a small biasing magnet on the pi filter, a random hash type noise is seen on the drive side at one point as you sweep towards 1 kHz. On the output a very large random ringing signal is noted. Once again, go too high and the circuit snaps off and can only be reset by  slowly sweeping up from a low frequency.

When you listen to the output signal through a headphone, it sounds like super -regenerative noise superimposed on the 1 kHz. If you've ever played with the squelch on a super-regenerative receiver you will know what I mean. The headphone should be isolated with a few thousand ohms, so as not to load the circuit too severely.

So the two things that are interesting are the snap off effect, which is very sharp when you go a few cycles over the peak point, and the random hash noise which occurs only when the biasing magnet is applied to the pi-filter cm choke.

This is not an effect of the amp directly driving a reactive load, as I isolated the amp with some resistance and the amp side is clean while the noise still shows up on the input choke and network output. There are all kinds of noise and frequencies running around in this circuit. Could this be what SM saw in his tube power supply? Some supplies are known to have a choke input filter followed by an additional pi-network filter.

So is this some form of Barkhousen noise or other domain noise as the core goes in and out of saturation...and why with the magnet applied?

If you want to see the interesting effect of a biasing magnet on a cm choke and the effects of saturation near resonance, try duplicating this circuit.

Note:  I'm using a small 10 Watt tube amplifier driven from an HP function generator for these tests, but some effects can be seen from the generator alone.  The scope was synced to the generator for these shots.

All waveforms presented were with a biasing magnet applied to the pi-network choke..

The first two waveforms represent a clean sine wave driving the network (the noise on the driving amp waveform is reflected from the network).

The second two represent gentle overdrive of the amp into clipping.

Also I used CM chokes with EE core configuration, and toroidal cores also work but seem to have less influence from permanent magnet. The effect is still there, however.

My camera doesn't like to focus well at close range....apologies.

[Grumpy]

is an amplifier required?

the snap-off effect might be some sorft of domain latching effect

the noise might by domains modulating between the crossed mag fields, one of which is static (the magnet)

Could be that the 2-field interaction is interacting with something in the background.

Robert Shannon posted a Scalar Detector that used the Barkhausen Effect (might be related):

http://www.amasci.com/freenrg/bark.html

[ANODE]

oops!

[MileHigh]

Ion:

I think that you are correct that it is Barkhousen induced noise, although I didn't know that it had a name!  When a magnetic domain snaps one way or the other this induces a spike in the signal.  I am assuming that the chokes may be air core.  If that's the case then the domains in the external magnet are snapping back and forth as the current through the chokes cycles up and down.  Even if the chokes have some sort of ferrite core, I will assume the magnet's domains are responsible for the induced noise.

You have to assume that the effect gets less and less noticeable as the physical chokes/magnet setup gets larger.

This setup is a variation on a pair of series LC filters.  A series LC filter acts as a short circuit at its resonance frequency.  Once you pass resonance and are on the 'far' side of the resonant frequency, the impedance shoots rapidly up and your output is very small.  It all depends on the 'Q' factor and all that stuff.

MileHigh

[ion]

I think it best to actually build the pi filter and try it out yourself.

When played with for a while, you will get a good feel for the weirdness factors.

Not sure you can even use a simulator for Barkhausen noise.

I've built up several different versions and apologize for the bad scope shots, you need to see this on your own scope.

The most fun little passive circuit I've played with in a while.

As I said, I used a small P-P tube amplifier 6BQ5 output tubes, but the effect can also be seen driving from a signal generator. More pronounced with the tube amp.

P.S.   I agree with MileHigh,  the higher frequency after the shift is shorted by the capacitors hence the low output, it's a low pass filter after all.

I'm familiar with how PI filters work, this is not the same.

If you do a pspice simulation of two resonant systems at the same frequency, you will see the two peaks instead of one, even though you try very hard to make them equal.

I'm familiar with this effect from when I used to tune bass reflex enclosures, beating speaker resonance against Helmholtz cabinet resonance. You get two peaks of much lower amplitude than the original large peak of the free air resonance of the loudspeaker alone. Add a little damping material over the port and you had a well tamed bass reflex enclosure, nice tight well defined bass, not grossly overdamped as in some of the early acoustic suspension systems.

I do tend to prefer transmission line enclosures tho

[Chet K]

Sir,
I know you have studied Newman
can you comment? [mostly the water Pump]

NOT NOW I just wanted to put this here since I couldn't find a Newman thread!!
Chet
From user bolt at OU
AT giantkillers thread
Here    http://www.overunity.com/index.php?topic=8227.msg261927#new
---------------------------------

Think about the Newman motor the "Giant Hotdog"   He has converted a low voltage stack PP3 battery source into an ELF pulse through large inductor. The inductor the coil of the motor prevents current flowing so the power factor has been trashed it now like a PF almost 0

http://www.josephnewman.com/  yes i know he is as Loopy as a fruit cake but he worked this out decades ago!

"The energy machine is proof that the strength of the motor's magnetic field
is dependent upon the VOLTAGE --- NOT the CURRENT!

For the first time, Joseph Newman demonstrates the
energy machine that turns a 1,650-lb rotary and powers a
375-lb positive displacement pump using LESS THAN 20 WATTS!

In the video demonstration, the energy machine produces phenomenal
speed, torque, and continuously pumps water using LESS THAN 20 WATTS!"

In so doing now you have a machine which is basically KVARS due to current < very small on the batteries yet in circulation is > massive together with a circulating voltage of hundreds or thousands of volt  from a few PP3 batteries. The central motor contact ARCS violent on switching.

WHERE does the machine gets its power from?  The answer is a PF 0 trashed power factor conditions create large magnet forces from the ambient.  So powerful in fact a magnet 12ft away hanging on a thread spins violently! This magnetic energy is converted to motor torque which provides more HP out then the power IN.  This energy came from the ambient due to scaler PF 0 electrostatic field causing disruption to local electron spin. The misalignment is over corrected in the windings and iron cores where internal spin domain currents create a magnetic field as a result.

Anytime you have a moving charge you get a magnetic field.

So as Newman motor works so slow its the best visual example to see how to convert VARS to magnetic force then use that to create shaft power far in excess of the PP3 batteries.

As also per Don Smith teachings a HV source can become electrostatic (VARS) and THEN modulated to create a powerful alternating magnetic field. Use this FREE magnetic field for the source of your energy and NOT the EM fields. Alike his 1 to 4 copy Tesla tower device.

As TPU its the free magnetic rotational field which does all the work.  The conditions to create a powerful magnetic field in rotation WITHOUT power banging fets is the key!  Class C modulated electrostatic fields is the answer.

If the coils have a rotational 3D 2 layer magnetic force spinning it only takes another standard coil placed close by to make a generator. There is no EMP, no banging not in uS anyway only gentle switching. No CPU's and no 30 volt 20 amp lab PSU's required. OH shit that rules out just about EVERYTHING in the last 5+ years:)

[ion]

Sir,
I know you have studied Newman
can you comment? [mostly the water Pump]

NOT NOW I just wanted to put this here since I couldn't find a Newman thread!!
Chet
From user bolt at OU
AT giantkillers thread
Here    http://www.overunity.com/index.php?topic=8227.msg261927#new
---------------------------------

Think about the Newman motor the "Giant Hotdog"   He has converted a low voltage stack PP3 battery source into an ELF pulse through large inductor. The inductor the coil of the motor prevents current flowing so the power factor has been trashed it now like a PF almost 0

http://www.josephnewman.com/  yes i know he is as Loopy as a fruit cake but he worked this out decades ago!

"The energy machine is proof that the strength of the motor's magnetic field
is dependent upon the VOLTAGE --- NOT the CURRENT!

For the first time, Joseph Newman demonstrates the
energy machine that turns a 1,650-lb rotary and powers a
375-lb positive displacement pump using LESS THAN 20 WATTS!

In the video demonstration, the energy machine produces phenomenal
speed, torque, and continuously pumps water using LESS THAN 20 WATTS!"

In so doing now you have a machine which is basically KVARS due to current < very small on the batteries yet in circulation is > massive together with a circulating voltage of hundreds or thousands of volt  from a few PP3 batteries. The central motor contact ARCS violent on switching.

WHERE does the machine gets its power from?  The answer is a PF 0 trashed power factor conditions create large magnet forces from the ambient.  So powerful in fact a magnet 12ft away hanging on a thread spins violently! This magnetic energy is converted to motor torque which provides more HP out then the power IN.  This energy came from the ambient due to scaler PF 0 electrostatic field causing disruption to local electron spin. The misalignment is over corrected in the windings and iron cores where internal spin domain currents create a magnetic field as a result.

Anytime you have a moving charge you get a magnetic field.

So as Newman motor works so slow its the best visual example to see how to convert VARS to magnetic force then use that to create shaft power far in excess of the PP3 batteries.

As also per Don Smith teachings a HV source can become electrostatic (VARS) and THEN modulated to create a powerful alternating magnetic field. Use this FREE magnetic field for the source of your energy and NOT the EM fields. Alike his 1 to 4 copy Tesla tower device.

As TPU its the free magnetic rotational field which does all the work.  The conditions to create a powerful magnetic field in rotation WITHOUT power banging fets is the key!  Class C modulated electrostatic fields is the answer.

If the coils have a rotational 3D 2 layer magnetic force spinning it only takes another standard coil placed close by to make a generator. There is no EMP, no banging not in uS anyway only gentle switching. No CPU's and no 30 volt 20 amp lab PSU's required. OH shit that rules out just about EVERYTHING in the last 5+ years:)

Lets ask the moderator to move this to an appropriate "Newman" thread, then we can have some fun looking at it. Meanwhile I'll check out the video......Cheers

[Chet K]

Sooo...........................
You're like a "Supreme Being" Now?

Kunta Kintee?

A"BLUE GUY"?

"Global Moderator Extraordninaire" Even?

Couldn't of happened to a better guy!!

Chet

[giantkiller]

@RamSet a.k.a. Buana Jim!
Thanks for posting that.

He also posted the sine wave amplifier I built. Now I build a better one with digital control.

The Tesla coil is hit with fast sharp pulses. Massive chatter is produced. A sinlge note is played from a guitar into an amp for feed back. It sounds really, really good until that moment when things break down. It all started with tubes which hold a small environment for ringing. So I have been plagued with this smooth coersion of atomic dipoles. It holds to the key to everything. Other types of processes have their place and effect. But the sine wave approach is still the most far reaching.
As far as Bolt goes. He has this moniker of lightning. I build a circuit he shows me. And strange things appear in the vids. The house rocks, trembles, vibrates because of simple heterodyning sinewaves. I stand back in awe. 3 sinewaves and I did this? Wow. I see how the sine waves can coax the stuff around us to move things whether that means break, vibrate, or shake at the matter level or subatomic level.

Look at Howard Johnson motor. The motor field is in the generator field and vice versa. They are wired together only to get them started. Then the parallel positioning does the rest. The 2 devices ring in unison just like a guitar with feedback.

Same as it ever was. Pick the target and apply the correct sinewave. All the big heads do it. Don Smith devices have a little trouble because he uses 2khz. Ah!, but SM uses 5khz.   Merry Christmas!
Here is a new one:
After the TPU was shown SM never mentioned changing the frequencies only the length of the wires. Why? The effect is a coordination of resonance with other ambiance.

Back to the bench.

[ion]

Here is a circuit I have been experimenting with in a desire to exploit the pulse sharpening and high current pulses available with a saturating common mode choke.

I was looking for a way to develop  pulses using a CMC to drive a bifilar zipcord at high current levels.

While this schematic shows a signal generator, an audio amplifier or simple transistor circuit can be used to drive the core into resonance and saturation.

Replacing the 0.01 ohm resistor with a long length of zipcord might produce some vibration due to repulsion effects.

Do not use electrolytics, best results with large yellow caps as seen in SM17.

The capacitor allows resonance,  high current, and saturation to develop in the secondary while the resistor enables you to view the current pulses using a scope across the 0.01 ohm without killing resonance. If you have a current probe, that would be a better choice.

I have tried numerous toroidal and EI cored inductors but the effect works best with toroidal CMC's around 2" OD stripped out of old monitors or power supplies, such as the type used for line filtering in switchmode supplies.

Sweep your generator up slowly from one to 10 kHz, go too high and the effect snaps off and you need to start low again.

Magnetic bias not needed with this arrangement.

[poynt99]

Good stuff ION.

So, is your working hypothesis that the CMC was used strictly as a pulse generating/sharpening component, or would it have a secondary use as well?

.99

[ion]

I was asked this question in a PM

I was wondering about your schematic. I have a controller that can produce 5ns pulses and with this have produced the dual pulse protocol into bifilar zip. I mention this because I am curious what would the advantage to using your circuit?

I was looking for a way to generate high current pulses with relatively short duration using CMC's as seen in the TPU designs.

I found that resonating a CMC into saturation produces such pulses and as SM said "No mass electronics..very simple"

Good stuff ION.

So, is your working hypothesis that the CMC was used strictly as a pulse generating/sharpening component, or would it have a secondary use as well?

.99

That is just one hypothesis that I am trying to prove on the bench. Note that to drive a long length of zipcord to produce vibration, high currents of short duration are required. This seems to provide them without mass electronics. A single transistor can drive the CMC using a little feedback.

[poynt99]

I found that resonating a CMC into saturation produces such pulses and as SM said "No mass electronics..very simple"

That is just one hypothesis that I am trying to prove on the bench. Note that to drive a long length of zipcord to produce vibration, high currents of short duration are required. This seems to provide them without mass electronics. A single transistor can drive the CMC using a little feedback.

And these pulses; would they need to start out as high current, or could a series of low-current pulses be built up over time to produce higher and higher currents? 

.99

[ion]

And these pulses; would they need to start out as high current, or could a series of low-current pulses be built up over time to produce higher and higher currents?  

.99

I cannot answer that question at this time, however if even a tiny COP > 1.0 e.g 1.001 were possible, the system could be easily looped to provide increasingly higher current pulses. (a system that feeds itself).

There are at first cut, only two mechanisms that could produce vibrational effects in the wires of the TPU. The first would be electrostatic in nature acting on the dielectrics of the wires. To cause heating effects however, the frequency would have to be in the diathermy range for organic insulators such as rubber.

The second would be high current short duration pulses in the wire causing repulsive or attractive effects. I'm leaning this way because of the heating effects at relatively low frequencies.

Of course there is a third possibility and that is some type of interaction with the earth magnetic field or a local magnetic field, static or alternating.

I'm interested in testing the first two effects, then move on to other more exotic interactions.

I would like to try some of this out in a sim, but lack enough info on hysteresis curves for CMC's. Maybe you could give this a shot since you have more experience with sims and magnetics.

[poynt99]

Hi ION,

Yes, a sim. It would be interesting to see if you can produce the same output as what you're seeing on the bench.

Using the "core.lib" is not too difficult, and I think we may have gone through one example a while ago?

You simply place the number of inductors you want coupled together, connect them in a CMC configuration, then select a core type and size, and couple it to those inductors with a coupling factor K. It will of course require some experimentation.

I won't be able to run sims for a few weeks time, but I might be able to help out with questions.

Cheers,
.99

[ion]

Hi ION,

Yes, a sim. It would be interesting to see if you can produce the same output as what you're seeing on the bench.

Using the "core.lib" is not too difficult, and I think we may have gone through one example a while ago?

You simply place the number of inductors you want coupled together, connect them in a CMC configuration, then select a core type and size, and couple it to those inductors with a coupling factor K. It will of course require some experimentation.

I won't be able to run sims for a few weeks time, but I might be able to help out with questions.

Cheers,
.99

Yes, thanks, I sort of remember the process, but may require some handholding e.g. which core type is closest to a CMC core? Which has similar saturation characteristics. I have very little info on CMC's, most of their specs are given in db attenuation and max amperes.

[poynt99]

ION,

Download this handbook from ferroxcube. This will allow you to select a core material and core size. The core sizes you will be interested in are toward the end of the handbook (TN etc), and core types are discussed near the beginning.

The "core.lib" does not contain every single core material and size in the handbook, but most. You will have to do some research I suppose to see what type of core material (B-H characteristics) is commonly used in CMC's then try to find that material (or equivalent) in the library. The size you need should be relatively straight-forward, and will already minimize your selections.

.99

[Peterae]

Hi ION
I have the most respect for you and your research, and would love to join in, lately time has restricted me from carrying out experiments though  , the noise on the CMC sounds so familiar when you listen to what SM talks about with regard to out of phase transformers.

The noise you have super imposed on those sine waveforms surely requires energy from somewhere to create the fast dv/dt transients, this would be energy that can be extracted if it could be coupled with, but where has it come from, i mean normally a fast transient requires a motive trigger, so a sine wave must at one point not created this transient, but then the transient is there,i think you are on to something big  depending on how much energy the sine wave needs to be sustained.

[ion]

POYNT thanks for steering me to the Ferroxcube catolog. I used them many years ago and probably have a paper copy somewhere but good to have an updated version. I may also have a bunch of sample cores somewhere.

Peterae...thanks for the kind words but, I may be on the wrong track, as the first cut efficiency looks pretty bad, as you have already found out and posted that info earlier. Of course my meter may not be able to handle the crest factor.

I don't want to lead anyone into the forest here, I just thought the current spikes were rather interesting, especially the snap off phenomenon.

I have yet to figure out what is going on with the earlier circuit posted.