Corbino Effects and Coler

[Smudge]

There has been little attention paid to Hans Coler's (really Unruh's) inventions in this forum.  My work with Chava produced more background showing that his Stromerzeuger was based around the Norrby device as suggested by Fred Epps who posts here as Orthofield.  While there are those over at OU.com who think that it was all a fake I am still of the opinion that the S machine has some merit.  I have looked into variations of the Hall/Corbino effect taking it into a cylindrical form where the possibility exists that induced RF eddy currents flowing around the ferromagnetic cylinder flow only in a thin surface layer (skin effect) and if that surface layer has some static radial magnetization the Anomalous Hall Effect (AHE) will induce longitudinal voltage to drive surface current along the cylinder.

That there exists the possibility of some self oscillation is not new, D.A. Kleinman and A.L.Schawlow predicted just that possibility back in 1960, see Corbino Disk abstract below.  In fact they filed a patent assigned to Bell Telephone Labs, US patent 3089995, copy also attached here.  While that required cryogenics to get the low temperatures it demonstrates a solid-state self oscillating thingammyjig having no power source (although the patent didn't claim that, it just claimed negative resistance).  Since that date there has been more knowledge of the AHE in ferromagnetic material and I think it possible this could explain the working of the Stromerzeuger at room temperature.  More later.

Smudge

[Smudge]

Here is a paper I wrote a while back explaining the Corbino Effect where radial current in the disc creates additional circular current via the Hall Effect, the net current flow then being a spiral.  There are no practical machines using the effect so it is little known among engineers and experimenters.  The Inverse Corbino Effect (my name for the effect) induces circular RF current (eddy currents) so as to produce a radial electric field that can drive radial current into a load, which of course is RF.  I have never seen this described anywhere.  Then there is what I call the Cylindrical Corbino Effect where a thin walled cylinder is used instead of a thin disc.  A radial magnetic field through the cylinder wall then allows a longitudinal current to create an additional circular current via the Hall Effect, the net current flow then being spiral.  Finally the Inverse Cylindrical Corbino Effect uses that same thin walled cylinder, having circular eddy currents induced into it that then produce RF voltage across the ends of the cylinder.  This is all illustrated in the paper.  I have never seen these cylindrical possibilities described anywhere.  Of note is the fact that for a Fe rod RF currents will only flow in the outer skin (skin effect) so the thing will act like a thin walled cylinder.  And Coler's magnetized piano wire at one end of his rod would produce the necessary radial magnetic field, perhaps enhanced by his multiple coil arrangement.  If the coil driving the eddy currents is put in series with the rod and a capacitor then you have a tuned circuit that could possibly self oscillate if the induced Hall voltage along the rod is sufficiently high.  Something to think about there.

Smudge

[Peterae]

For experimental purposes it is not necessary to reproduce the complicated Strömerzeuger
circuit. A simple experiment using an iron rod with a magnet attached to one end will produce
leakage flux along its length that is radially unidirectional. If that is placed within a coil that
has been resonated with a capacitor to the claimed 180KHz frequency, with the iron rod
included in the circuit wiring, it should be possible to look for Q enhancement or degradation
when the magnet polarity is reversed. If that happens it is indicative of the Hall effect being
present, then it may be possible to embark on a process of adjustments to gradually improve
the enhancement to the point of self-oscillation. Such adjustments would almost certainly
involve the iron material but at least there would be a test-bed for exploring the effect of
different purities or treatments. Alternatively the RF Hall voltage along the rod could be
isolated and measured, then its phase noted with respect to the driving RF. A reversal of the
PM should create a 180 degree phase change.

OK i want to give this a go, i have the pure iron rod from my meyer setup and it has tapped ends for screw terminals, i should have a neo magnet that's similar diameter to place at one end 20mm iron rod and magnet.
Does the coil that's surrounding this magnet/iron rod matter if it's a larger diameter i think it's wound on a 40mm plastic pipe.

[Peterae]

OK i hooked up my pure iron rod and wound a coil down it's length and hooked it up to my switchable cap bank and adjusted the cap value for max amplitude, i used my scopes internal DDS to lock at 180KHz to drive the coil and parallel cap bank.

One end of the iron rod had the coil connected to it and the other end had a neo magnet with a wire sandwiched between the magnet and iron rod, i marked one side of the magnet so i knew which way round it was and then scoped the sine with the magnet each way round.
The DDS is only 2.5V Amplitude or 5V pk-pk.

There is a small measurable difference when the magnet is flipped, see both scope shots, each scope shot is with the magnet in different orientations.

One way the average pk-pk is 329mv
the other way the average pk-pk is 322mv

I tried a few times to make sure it was not the sandwiched wire disturbance that was causing this and the results were the same each time.
So i think we have a very small confirmation of the effect here 

any tips on how to progress would be helpful 

PS i did not know if the iron rod polarity matters, in the picture below, the black wire is sandwiched between the magnet and iron rod, this is one connection that goes to my cap, and the blue wire coming from the coil end near the magnet is the other connection to the cap, the left hand side of the rod and coil are connected together.

[Smudge]

Peter,

If I read you correctly you have the Fe rod in series with the coil and are looking for a change of Q when the magnet is reversed.  You could try putting another magnet on the other end of the rod.  Its polarity must be reversed to that of the first magnet so that both magnets push flux inwards (or outwards).  This then increases the amount of flux coming out radially.   The effect should then be slightly increased on the 7mV difference that you saw.  That should help confirm that what you measured really was this Hall/Corbino effect.

Another possibility is to not have the rod in series with the coil but drive the parallel coil/cap as you did.  Then scope across the Fe rod which should be almost a short circuit but may have a small measurable voltage.  If it is detectable look to see if its phase changes by 180 degrees when the magnet is flipped.

Smudge

[Peterae]

Yes thats correct, the rod is in series with the coil.

Sounds good Smudge, i will give that a try when i next venture to the workshop.

[Peterae]

OK i just did the magnet part of the test, first i want to check with you, i am connecting my signal generator across the LC i am not loose coupling is that OK, if not i will re do the experiments.

So the tests below.
One magnet scope shot 1
2 magnets attracting, 1 each end scope shot 2
2 magnets opposing 1 each end scope shot 3
1 magnet flipped so opposite face is on iron rod scope shot 4
2 magnets attracting scope shot 5
2 magnets opposing scope shot 6

then i switched the end of the coil that was in series with the iron rod.
1 magnet scope shot 7
2 magnets attracting scope shot 8
2 magnets opposing scope shot 9

Strange results?
1=320mv 1 magnet
2=324mv 2 magnets Attracting
3=316mv 2 magnets opposing
4=314mv 1 magnet other way around
5=306mv 2 magnets attracting
6=318mv 2 magnets opposing
7=334mv 1 magnet with other end of solenoid in series with iron so as test 4 but with the different solenoid end in series with the iron
8=316mv 2 magnets attracting with solenoid wire the same as 7
9=320mv 2 magnets opposing with solenoid wire the same as 7

I have to say even though the magnets are opposing they still stick to the iron rod with force at both ends

[ion]

It would be interesting to try a  setup by having two identical coils and iron rods, both initially without magnets and subtracted using Ch1 minus Ch2 of your scope. Then as you apply the magnets to only one channel, the difference signal will be displayed with much higher sensitivity.

Or you can make a simple passive bridge circuit to do the initial nulling and use just one channel of your scope to see the resultant difference.

In this way you have a control and a DUT that are dynamically compared.

[Peterae]

after thinking about the results somewhat, to connect the iron rod i sandwiched a stripped wire between the magnet and the iron rod end relying on the attraction to jam the wire, thats ok when i have 2 magnets one at each end, but when i had 1 magnet i had to connect the wire to the other end, i did this by poking it in a hole which is at the end of the iron rod, maybe this allows the wire to enter the iron rod by 5mm and thus we have a slightly shorter iron rod and hence a slightly larger value in test 7.

Maybe the flux is not leaking out at all, i mean with 2 opposing magnets at each end they still both get attracted to the iron rod really strongly and are not really opposing because of the length of the iron rod maybe.

The other thing not accounted for in my test is whether i need to retune my capacitor for maximum resonance, i wonder if the inductance can change when the iron has a magnetic attached to it, i guess i could try this while connected to my inductance meter.

I think next i need to try scoping the iron rod for a phase change and see if anything is detectable next.

Hi ION
Yeah good idea, although at the moment i don't have 2 identical pieces of iron, although it should not be hard to find something, the other thing is to try and design for maximum Q at 180Khz to increase the effect.

[Smudge]

Maybe the flux is not leaking out at all, i mean with 2 opposing magnets at each end they still both get attracted to the iron rod really strongly and are not really opposing because of the length of the iron rod maybe.

That attraction is to be expected. See FEMM plot of flux in rod with opposing magnets at each end.  All the flux lines entering the rod have to leave the rod.  But the concentration of flux at the contact point between magnet and rod indicates the magnet is attracted to the rod.  When you have a very short rod then you can get the magnet being repulsed at each end.

The other thing not accounted for in my test is whether i need to retune my capacitor for maximum resonance, i wonder if the inductance can change when the iron has a magnetic attached to it, i guess i could try this while connected to my inductance meter.

and from your previous post

OK i just did the magnet part of the test, first i want to check with you, i am connecting my signal generator across the LC i am not loose coupling is that OK, if not i will re do the experiments.

Since signal generators generally have low output impedance connecting them directly across the LC will result in very low Q and you will not see the change in Q you are looking for.  So I would redo the experiment with the sig gen loosely coupled, e.g. by a single turn.  Then you may well find the need to re-tune as in each case it is important that the tuning is spot on else the Q enhancement effect effect will be overshadowed.  Another way of loosely coupling a sig gen is to put a high value capacitor in series with the tuning one at its earthy end.  By high I mean much greater than that of the tuning one, like between 10 and 100 times.  Then you connect the sig gen across that high value cap.

P.S.  The little things on the rod are single turn coils as this plot is from some work I am doing looking into various anomalous effects using flux gradients. 

Smudge

[ion]

Smudge said:

Another way of loosely coupling a sig gen is to put a high value capacitor in series with the tuning one at its earthy end.  By high I mean much greater than that of the tuning one, like between 10 and 100 times.  Then you connect the sig gen across that high value cap.

Normally good advice and in the days of valve signal generators you might easily get away with this, but I suspect the newer solid state ones would not be so forgiving. I lost a nice pulse generator by accidentally driving a low Z load when a couple of transistors in the output stage gave up. The single turn loop could also be dangerous.

For best safety, sometimes I just use a high value resistor for coupling as the load is then sufficiently decoupled so that you can see changes in Q, but it is at the expense of signal level. Alternately a loose coupler that has enough inductance such that the sig gen is not driving too hard or saturating could be used. Hopefully the 50 ohm output impedance of the sig gen is an actual resistor, not current feedback derived.

Ultimately you should check the manual of your generator to see what it says about driving capacitive or other loads much less than 50 ohms.

[Peterae]

good advice Thanks Ion & Smudge, i will switch to my old sig gen which has a higher output anyway and re try the tests.

I did try the 2nd test looking for a phase change across the iron rod but the phase 90 degree leading did not change enough to see it change when the magnet was flipped.