Meyer-Mace Isotopic NMR Generator

[Verpies]

I am having very little luck plating an iron rod, it is working but i will be old by the time it is long enough.

How are you plating it?

So if we run at 0.5 Tesla as per the iron then we get an operating frequency of 5.66 MHz

Yes, copper at 0.5T will resonate at 5.66MHz.
However individual separate iron nuclei resonate at 690kHz at 0.5 Tesla but metallic iron at room temperature resonates at 45.5MHz.  Lower if it is hot.  To get it down to 21MHz you'd need to heat it up to 1000°K (that is lower than the Curie temperature, though).

Trouble is we still have the skin depth problem at this frequency...

Yes but the evil skin effect can be mitigated by minimizing the differential permeability and/or conductivity.
This can be accomplished by magnetic saturation and/or heating ...or both since even approaching the Curie temperature also decreases the saturation magnetization (see Fig.4)
It is also possible to completely work around the evil skin effect by using acoustic waves.

Heating iron seems more and more promising.

[Peterae]

How are you plating it?

Using the method Robert Murray smith found see link, it's a 3 chemical mix for any metal.
http://www.overunityresearch.com/index.php?topic=2518.0

I can see the advantages of heating the iron but is it practical.

Sometime ago i started a thread about a hot coil electromagnet, the question was if a tungsten coiled wire had a current flowing to heat it to a really high temperature as in a bulb filament then would a magnetic field be produced as if it was an electromagnet.

Now if we wound a solenoid coil using iron wire and fed a High current DC bias to heat the iron coil to 1050 Deg C and superimposed the required MHz frequency, what would be the minimun wire thickness we could use, i would think that if the fuel wire was responsible for producing the bias magnetic field then it should be permeating from every atom because the bias magnetic field is being produced from all atoms of the heated wire, no core would be need and no overwindings to suffer a fate of hot temperature, so the complete device is just a white hot iron filament.

[Verpies]

Sometime ago i started a thread about a hot coil electromagnet, the question was if a tungsten coiled wire had a current flowing to heat it to a really high temperature as in a bulb filament then would a magnetic field be produced as if it was an electromagnet.

Simply, Yes

Now if we wound a solenoid coil using iron wire and fed a High current DC bias to heat the iron coil to 1050 Deg C and superimposed the required MHz frequency, what would be the minimum wire thickness we could use, i would think that if the fuel wire was responsible for producing the bias magnetic field then it should be permeating from every atom because the bias magnetic field is being produced from all atoms of the heated wire, no core would be need and no overwindings to suffer a fate of hot temperature, so the complete device is just a white hot iron filament.

There are two problems with this approach. The radii of the Lorentz orbits and the distribution of magnetic flux in a helical winding, which is zero in the center of the wire.



The DC coil can be made out of the Nichrome wire and be a heater of the core at the same time...but the core cannot be deleted.
Nichrome melts at 1673ºK while the Curie point of iron is 1043ºK so it will work for this purpose. The heat resistance of insulation is the only problem.

[Peterae]

There are two problem with this approach. The radii of the Lorentz orbits and the distribution of magnetic flux in a helical winding which is zero inside the wire.

OK.
So we need a hot iron rod, i suppose we could sleeve this with quartz tube then sleeve again with an air gap with something like a ceramic tube and wind coils over this.

The radii of the Lorentz orbits

Is it this that is limiting our iron rod diameter?

[Verpies]

So we need a hot iron rod, i suppose we could sleeve this with quartz tube then sleeve again with an air gap with something like a ceramic tube and wind coils over this.

Fused quartz would work as well as white silica. S2 glass fiber mat would also work and would be the easiest IMO.
Yes, the other coils can be wound over that.  Copper will not melt until 1357ºK but it will oxidize without insulation. For heat-resistant windings of low turns, such as the MHz coil, a copper capillary (e.g. from an old refrigerator) can be used if water is run through it and it is not touching the DC heater winding.

Is it this that is limiting our iron rod diameter?

Yes, according to that formula in my first post in this thread.  The further you get away from this Lorentz radius the lower the efficiency because a lot fast electrons escape the fuel rod then.
A millimeter diameter wire is not only way below this orbit radius but also the magnetic field in the center of the wire is zero!

[Peterae]

Thanks verpies for the explanation.

ION found the solution to our problem, he used large bore ceramic resistors to reach 1000 Deg C, the iron rod can be placed down the center hole which can be quite large, will do some hunting for data sheets tomorrow when i get time to get data on sizes.
http://www.overunityresearch.com/index.php?topic=2888.msg47411#msg47411

With the 150 Ohm resistor, a bit over 900 C is possible, I was able to get to 1000 C

[Verpies]

ION found the solution to our problem, he used large bore ceramic resistors to reach 1000 Deg C, the iron rod can be placed down the center hole which can be quiet large,

Yes it can be, but remember that the intent is to heat the iron rod to 1000°K which is close to its Curie point at 770ºC (1043ºK) at 0T, so a bore much larger than the diameter of the rod is not advantageous here.
It is the other windings that need protection from the heat - not the iron rod.

[Grumpy]

Try this instead:

[Grumage]

Dear Peterae and Verpies.

I have been doing some work with the RMC induction oscillator recently, what is great is that Richard found out by accident that his particular design can run without the need for a centre tap coil.

A thought?

Could this circuit be used to provide both the heating effect and modulation frequency in one? All that would then be needed is the correct DC bias for the magnetic field.

I also looked into the subject of pure Iron rod, it's readily available in short lengths and many diameters. A quick Google was all it took! http://www.pureiron.co.uk/index.htm

Cheers Grum.

Edit.

http://www.rmcybernetics.com/shop/induction-heater-circuit

[Peterae]

Hi Grumage

Thanks for the links indeed i did see the pureiron previously, and maybe i will need to go their, just seemed really expensive, minimun order of £50 and then £20 P&P so to get a 10cm piece of iron will cost £70, but i think it is the only source i have seen so far, so i might need to go there, shame there's not more builders, we could bulk order.

Whats the frequency of the RMC induction oscillator, not seen it before.

EDIT
OK it goes up to 100KHz 

Cheers
Peter

[Peterae]

verpies

but remember that the intent is to heat the iron rod up to its Curie point, which is 770ºC (1043ºK) at 0T, so a bore much larger than the diameter of the rod is not advantageous here.

ION uses the resistor packed with nickle, it is the nickle inside the resistor that is heated but yes the heat would dissipate from the outside surface as well unless insulated with something, as for size we can find one that fits our iron rod diameter or pick a rod size that fits the resistor.

However for iron the nuclear resonance at 0.5 Tesla occurs at 690kHz and for saturated cold iron (@ 2.14 Tesla) it occurs at almost 3MHz.

Is there a way to work out how the NMR frequency alters versus temperature.

At 2.14T the Lorentz orbits of 2.7Mev electrons will have the diameter of slightly under 10mm and NMR frequency is almost 3MHz.

So it looks like we have some base line parameters for a device
Fuel rod 20mm pure iron
Fuel rod Temperature approximately 1000 Deg K
Bias Field 2.14 Tesla.
Main oscillator 3MHz 400 Watts 20 turn coil

[Chet K]

Peter
Quote
shame there's not more builders, we could bulk order.
-------------------------------------------

we could share "your" purchase and shipping costs ?

its only fair ??

perhaps a Little Kitty for this should be maintained here ?

[TinselKoala]

http://www.youtube.com/watch?v=azMbJWr9Nqg

The circuit for the non-centertapped version is only slightly different, it uses exactly the same components plus one more RF choke.

[Verpies]

Is there a way to work out how the NMR frequency alters versus temperature.

It varies a lot, like this:



Also, the saturation flux density varies with temperature.

So it looks like we have some base line parameters for a device
Fuel rod 20mm pure iron
Fuel rod Temperature approximately 1000 Deg K
Bias Field 2.14 Tesla.

Basically yes except the Bias Field necessary to reach the magnetic saturation, will be lowered at 1000ºK.
Also a choke in series with the DC bias and heater Nichrome winding is an important element, as it prevents the necessary loading of the MHz coil.

Main oscillator 3MHz 400 Watts 20 turn coil

Maybe.
The power level requirement might be relaxed at high temperature as the evil skin effect is minimized and RF penetration maximized. The choke helps, too.
Personally, I would make the frequency tunable from 500kHz to 25MHz.

[Verpies]

http://www.youtube.com/watch?v=azMbJWr9Nqg
The circuit for the non-centertapped version is only slightly different, it uses exactly the same components plus one more RF choke.

Neat circuit!
I wish your video better showed the LC frequency as you inserted the the thicker workpiece while its Curie temperature was reached.
The LC frequency should increase upon reaching it, which would be educational.

P.S.
I always wonder with these inductive heaters, what percentage of the heating is caused by hysteresis losses vs. eddy currents.
May I suggest a second fixed local oscillator, a 4-diode mixer and a speaker for bringing the LC frequency into audible range for easy observation "by ear".

[TinselKoala]

Neat circuit!
I wish your video better showed the LC frequency as you inserted the the thicker workpiece while its Curie temperature was reached.
The LC frequency should increase upon reaching it, which would be educational.

P.S.
I always wonder what percentage of the heating is caused by hysteresis losses vs. eddy currents.
May I suggest a second fixed local oscillator, a 4-diode mixer and a speaker for bringing the LC frequency into audible range for observation "by ear".

Hmm... that's an interesting experiment to try. I'll see if I can set something up that would show the "step" in frequency as the Curie point is reached/exceeded.

I might be able to kludge together some kind of BFO system as you suggest, or capture it on the scope. Or, if I can figure out the code, it might be possible to implement that function on the Arduino-- the basic metal detector that I made a few days ago detects frequency change digitally and displays the "error" or change from a reference, as clicking sounds, like a Geiger counter. In fact now that I think of it I may be able to use that entire Arduino metal-detection system practically as-is.

[Verpies]

@Grum

What temperature does the "work piece" reach in Tinsel's video above?
You are the perfect person to ask since you melted a lot of metals in your life and you probably can estimate their temperatures just by looking at their colors.

[Verpies]

I always wonder with these inductive heaters, what percentage of the heating is caused by hysteresis losses vs. eddy currents.

...because if the hysteresis losses are the major heating cause, then the workpiece will tend to sit at the Curie temperature (due to negative heating feedback) and the LC frequency shift will not be as pronounced as with exceeding the Curie point.

[Grumage]

@Grum

What temperature does the "work piece" reach in Tinsel's video above?
You are the perfect person to ask since you melted a lot of metals in your life and you probably can estimate their temperatures just by looking at their colors.

Dear Verpies.

Aye, the induction furnaces were using Kilowatts though!! 

Without knowing the background light in the video, a rough estimate is in the 700°C area. What I found interesting was that the sample didn't carry on increasing in temperature !! My Foundry days are long past and for the life of me I can't remember how the furnaceman controlled things.

But TK's video clearly shows that a sample can be held at a specific temperature, within reason, and at a fixed frequency.

I also remember when Richard was developing his oscillator he wound the Litz wire coils onto a pair of Ferrite "C" cores. Placing a screwdriver blade between the gap he was able to heat the blade rapidly. What was interesting is that when the Curie point was reached the input current dropped. A little observation to ponder on?

Cheers Grum.

[ion]

Great work fellows, glad to see the interest in this.

So many possibilities for creative work in this area.

I will follow with interest and hope to contribute as my time permits.

Kindest regards,
ION

[Verpies]

What was interesting is that when the Curie point was reached the input current dropped. A little observation to ponder on?

When the Curie point is reached the iron stops being "magnetic".  In nerdspeak: it stops being a ferromagnetic and starts being a paramagnetic.
Paramagnetics do not have a magnetization hysteresis curve, so their heating cannot be caused by hysteresis loss anymore and the temperature drops.
When that happens, the iron can become a ferromagnetic again and can become heated by the hysteresis loss.  Then its temperature rises...and so on, effectively hovering around the Curie point.

The material can also be heated by eddy currents.  That is how eg. aluminum can be heated in induction ovens despite aluminum never being a ferromagnetic.
The eddy current heating depends on resistance of the material being heated, so it has its limit since most materials increase their resistance when heated.
However this eddy current heating limit can be above the melting point, when enough power is used.

It is important to remember that the workpiece is losing heat into the environment all the time by radiation, conduction and convection (the higher the temperature the faster it loses the heat energy) and the inductive heater must make up for these losses if it wants to keep increasing the temperature.  Eventually a thermal equilibrium is reached.


A CURIOSITY:
It is possible to program an induction heater to turn off when its LC frequency rises above the "empty" frequency.  This way the workpiece will always be heated by hysteresis losses and never by eddy currents only, since the heater will turn off as soon at the workpiece reaches its Curie temperature and the LC frequency increases....effectively keeping the temperature there for exact measurement.

[Peterae]

Been trying to design our bias coil, but having trouble.

As the temperature of the iron reaches and sits at our curie temperature then the iron relative permeability approaches zero, when i plug 0 relative permeability into the magnetic field calculator i don't get any magnetic field,

http://www.calctool.org/CALC/phys/electromagnetism/solenoid

Do you think we can accept using mild steel as our fuel rod, it would make life a lot easier for sourcing the iron rod, surely the bulk will still be iron.

Tuning this may be problematic, i understand that it can be 15 minuets for the device to start up and then 15 minuets to stop.

also some questions.

How do we detect the neutrons, and what shielding do we need.


EDIT
Oh just found this

The saturation magnetization goes to zero at the Curie temperature.

So it is not possible to magnetically saturate iron when it's at or above the curie point  

[Verpies]

As the temperature of the iron reaches and sits at our curie temperature then the iron relative permeability approaches zero,

Nope, it approaches 1.

Do you think we can accept using mild steel as our fuel rod, it would make life a lot easier for sourcing the iron rod, surely the bulk will still be iron.

Yes, the bulk would be Fe but the additives will significantly change the crystalline structure.  Who knows how this will affect the viability of this experiment.

How do we detect the neutrons, and what shielding do we need.

There is a special gel that's easy to use.  
It's pretty hard to shield neutrons.  Light elements (e.g. hydrogen in ordinary water) interact with them most easily but distance is the most practical countermeasure.

Oh just found this So it is not possible to magnetically saturate iron when it's at or above the curie point  

Yup, it already behaves as if it was saturated when above the Curie temperature. Not only the differential permeability falls to one then, but the absolute permeability falls to one, too.

[Peterae]

Nope, it approaches 1.

OK makes sense, so we are really operating as if we have no core at all with regards to field calculations.

I found a rough quote of $700 for this gel, need to look at it a bit closer.

So at the curie point we have fuel that is saturated and no longer is effected by skin depth, as you stated earlier.

Now the problem with the NMR bias field being part of the heater is that it will get cycled on and off, or pwm controlled if we went complicated.

I am thinking it maybe better to run an ON/OFF heating control cycle and just run the MHz drive signal when the heater is off and use a separate coil for the bias field, my reasoning on this is we don't need a strong bias field now (or do we still for Lorentz orbit ) and that will lower the NMR frequency, but because we are using an E class amplifier which is more complicated to build tunable then if we stick to one center park frequency, all we need to do to tune this beast is vary the DC voltage/current on the bias coil to shift our NMR up and Down to hit out Class E frequency, so much easier to vary a voltage in mv steps than a frequency and associated amplifer to 0.001Hz

Now we are in a position to just choose a frequency say 1MHz and we can work out the bias field needed and the current and voltage variation to give us a 500KHz to 2 MHz swing for the NMR frequency, then with all the unknowns we should be able to find in this case the magic bias value for isotropic mutation at 1MHz or any other frequency we choose.

So the next thing then is to find our static NMR chosen frequency, are the benefits for a higher frequency say 10MHz as opposed to a 900KHz frequency, apart from the different bias field strength.

How do we control the speed of conversion? maybe by varying the MHz power, or is it done by offtuning as in the McFreey Device, or maybe we can control the amount of atoms being converted by controlling the skin depth.

I think we need to thrash this Lorentz orbit out a bit, we need an orbit small enough to keep the fast electrons inside the core? Why we already have Neutrons flying out, are fast electrons going to matte, and surely anyway if the atom that they come from is on the outside perimeter of the rod then they will surely be outside rod containment anyway.

Presumably it is these fast electrons that are our power out source, if so i can see we want as many contained as possible.
Already answered, my bad memory sorry.

The further you get away from this Lorentz radius the lower the efficiency because a lot fast electrons escape the fuel rod then.

Basically yes except the Bias Field necessary to reach the magnetic saturation, will be lowered at 770ºC.

Yup, it already behaves as if it was saturated when above the Curie temperature. Not only the differential permeability falls to one then, but the absolute permeability falls to one, too.

So our bias field now is only needed purely for the Lorentz orbit diameter.

Then we need to know what needs to be prominent, our DC bias, lets say at the moment we went with 0.5T then how prominent do we want our MHz signal, is there a way of knowing how much magnetic field we need to swing +- 0.25T for instance when coupled with our bias would give a total swing in the rod of +0.25 to +0.75T is that enough, can we calculate how much energy or at what field strength the atom flips 180 Degrees.

So things we need to nail down & choose
Bias Field strength Tesla's
Static NMR Frequency
Static NMR Frequency field strength +/- ?Tesla's or in terms of watts power would do.
50Hz field strength +/- ? Tesla's or power in watts.
Fuel rod length and or diameter i could settle on 20mm diameter, but length is open for debate, but if the Lorentz orbit is larger because of a low mag field strength then we would go to 40mm or greater, but this will make it harder to heat due to thermal mass, maybe this does not matter, the device could be heated and started and as the fuel core heat permeates the rod then things get better as time passes, a sort of slow low start.

Is there any other way of making the orbit much smaller mm or so other than field strength.?

Once these items can be guesstimated i think a device can be assembled.

things i have found to help built it are Car Exhaust putty good for up to 1000 Deg C
Kanthanal wire heating element or ceramic tube resistor.
Foam fire bricks used in gas fires as fake coals as insulation.

[Verpies]

OK makes sense, so we are really operating as if we have no core at all with regards to field calculations.

Above Curie temperature - yes.
Below Curie temperature (e.g. 500C) some H field will still be needed to reach saturation, but less than for a cold iron (the H field is made by the DC coil and is expressed in ampturns/meter where the "meter" refers to the length of the magnetic circuit (rod + U).

So at the curie point we have fuel that is saturated and no longer is effected by skin depth, as you stated earlier.

Oh, it is still affected but much much less.  Compare, the 50% attenuation skin depth of:
a) cold unsaturated iron at 1MHz is 1.56µm
b) cold saturated iron at 1MHz is 110µm
c) hot 770C iron at 1MHz is 340µm (at 100kHz is 1080µm)

But that is for iron conducting electric current, not magnetic flux, so it is not directly applicable.
Of course alternating magnetic flux in any conductor will induce electric eddy currents, but the attenuation of AC magnetic flux is very different from the attenuation of AC electric current.

Now the problem with the NMR bias field being part of the heater is that it will get cycled on and off, or pwm controlled if we went complicated.

Yes, that's why pure DC should be used.

I am thinking it maybe better to run an ON/OFF heating control cycle and just run the MHz drive signal when the heater is off and use a separate coil for the bias field,

A separate heating and bias coil can be used if the heating coils is wound in bucking bifilar mode so it does not contribute to the DC bias.

my reasoning on this is we don't need a strong bias field now (or do we still for Lorentz orbit ) and that will lower the NMR frequency,

Yes, the Lorentz confinement is still an issue but I'd suggest keeping the DC bias at 0.5T because that's what the patent recommended and 0.5T is square in the middle of the domain magnetization rotation zone mentioned in that NMR paper from Cyril.

but because we are using an E class amplifier which is more complicated to build tunable then if we stick to one center park frequency, all we need to do to tune this beast is vary the DC voltage/current on the bias coil to shift our NMR up and Down to hit out Class E frequency, so much easier to vary a voltage in mv steps than a frequency and associated amplifer to 0.001Hz

With other substances that would be very true, but as you can glean from that NMR paper, iron has its own internal magnetic hyperfine field of great density (~33T) that overshadows our weak 0.5T field  and the resonance frequency will stay at 45.5MHz regardless whether we apply that 0.5T biasing field or not.  At 1000°K that frequency falls to 21MHz.

Now we are in a position to just choose a frequency say 1MHz and we can work out the bias field needed and the current and voltage variation to give us a 500KHz to 2 MHz swing for the NMR frequency, then with all the unknowns we should be able to find in this case the magic bias value for isotropic mutation at 1MHz or any other frequency we choose.

Again, Cyril's paper changes everything because we learn from it that the line width for non annealed iron is ±80kHz and for annealed iron it is ±25kHz.  (impure Iron could be much less sharp though).
That small frequency deviation is well within the range of class E amplifiers.

So the next thing then is to find our static NMR chosen frequency, are the benefits for a higher frequency say 10MHz as opposed to a 900KHz frequency, apart from the different bias field strength.

Looks like nature made that decision for us and fixed the resonance frequency at 45.5MHz at room temperature.

How do we control the speed of conversion? maybe by varying the MHz power, or is it done by off tuning as in the McFreey Device, or maybe we can control the amount of atoms being converted by controlling the skin depth.

I think just like in McFreey's device because each pulse is self quenching since the circulating fast electrons generate electric current that generates magnetic field that opposes the 0.5T confinement field.

I think we need to thrash this Lorentz orbit out a bit, we need an orbit small enough to keep the fast electrons inside the core? Why we already have Neutrons flying out, are fast electrons going to matter,

Yes because only electrons and positrons will generate our output electric current and neutrons will not.  Neutrons are just deadly, if they are emitted at all.

and surely anyway if the atom that they come from is on the outside perimeter of the rod then they will surely be outside rod containment anyway.

Yes, but if the 0.5T containment field exists outside of the fuel rod (e.g. because the 0.5T DC solenoid is larger than the rod) the the path of such electron will be curved back into the rod.

Presumably it is these fast electrons that are our power out source, if so i can see we want as many contained as possible.

Fortunately their motion destroys their own confining magnetic field.

So our bias field now is only needed purely for the Lorentz orbit diameter.

Yes and also maybe to polarize iron in the middle of that domain rotation zone described in Cyril's paper

Then we need to know what needs to be prominent, our DC bias, lets say at the moment we went with 0.5T then how prominent do we want our MHz signal,

The frequency shift below 0.6T is proportional to (H_BIAS / 33.02T)² so it is very small

can we calculate how much energy or at what field strength the atom flips 180 Degrees.

Usually a burst of specific length is needed for that.  Maybe Smudge has already calculated it.

So things we need to nail down & choose:
1) Bias Field strength Tesla's
2) Static NMR Frequency
3) Static NMR Frequency field strength +/- ?Tesla's or in terms of watts power would do.
4) 50Hz field strength +/- ? Tesla's or power in watts.
5) Fuel rod length and or diameter i could settle on 20mm diameter, but length is open for debate,

1) 0.5T just like in the patent AFAIR
2) 45.5MHz at room temperature and 21MHz at 1000°K  (a Litz wire winding might be advantageous here).
3) f_RESONANCE =1381564(H_BIAS - H_D - 33.02T) at toom temperature.  I don't know about the RF power and about the demagnetization field (H_D) for the particular dimensions of the fuel rod.
4) I don't know but I guesstimate not more that the DC bias field.
5) 20mm diameter feels fine. I cannot be certain because I do not know the energy of the fast electrons.  If they are 1MeV then 20mm should be enough.
The size and proportions of the fuel rod determine its demagnetization field (H_D). Maybe Smudge can help us calculate it with an elongated ellipsoid formula for that shape.

Iron-57 is stable so it normally does not emit any fast electrons/positrons and thus their energies are not listed in literature, but the whole idea of this contraption is to destabilize iron and make it emit these charged particles.
Its neighbors, e.g. Iron-59 spontaneously emit 1.565MeV electrons, see: http://www.periodictable.com/Isotopes/026.59/index.full.dm.html

The destabilization of nuclei through electrical means has been documented before, so it is not that far out.  For example read up here on Dysprosium ¹⁶³Dy that is also normally stable but can be made unstable using non-nuclear (electronic) methods.

Is there any other way of making the orbit much smaller mm or so other than field strength.?

For given speed/energy of the electrons it is the only way.
We could be overestimating their speed/energy, though.  For example Iron-55 decays only with 231keV and such small speed/energy of these fast electrons would result in smaller Lorentz orbits.


Isotope  Energy        Lorentz orbit diameter at 0.5T
------------------------------------------------------------------------
Fe⁵⁵         231keV     7.2mm
Fe⁶⁰         237keV     7.3mm
              1000keV    19.0mm
Fe⁵⁴       1364keV    24.0mm
Fe⁵⁹       1565keV    26.8mm
Fe⁵³       2721keV    42.6mm

All of the above isotopes decay in Beta modes without any neutron emissions (see here)