Meyer-Mace Isotopic NMR Generator

Author Topic: Meyer-Mace Isotopic NMR Generator (Read 141428 times)

Topic Options

Smudge

Re: Meyer-Mace Isotropic NMR Generator « Reply #50 on: 2015-07-01, 09:32:41 »

Group: Professor
Hero Member

Posts: 2306

Here is a paper showing that for transmission through iron it acts something like a high temperature Bose- Einstein Condensate. Not sure of this is of any interest but I'll send it anyway. And while on the subject, here are two other papers I have on Fe57 NMR.

Smudge

electrodynamics in iron and steel.pdf (596.44 kB - downloaded 738 times.)

Fe57 NMR in pure iron.pdf (322.54 kB - downloaded 467 times.)

Fe57 NMR and FMR equal.pdf (199.57 kB - downloaded 442 times.)

Verpies

Re: Meyer-Mace Isotropic NMR Generator « Reply #51 on: 2015-07-01, 15:58:33 »

Group: Administrator
Hero Member

Posts: 4440

Quote from: Smudge on 2015-07-01, 09:32:41

Here is a paper showing that for transmission through iron it acts something like a high temperature Bose- Einstein Condensate. Not sure of this is of any interest but I'll send it anyway. And while on the subject, here are two other papers I have on Fe57 NMR.

Those two NMR papers show that the NMR frequency of Iron decreases when the external field increases, which is the opposite to other materials. A significant difference.
Also, since when Iron saturates at 7.5kG (0.75T) flux density ?

Smudge	Re: Meyer-Mace Isotropic NMR Generator « Reply #52 on: 2015-07-01, 17:08:58 »
Group: Professor Hero Member Posts: 2306	Quote from: verpies on 2015-07-01, 15:58:33 Also, since when Iron saturates at 7.5kG (0.75T) flux density ? Perhaps because the sample was iron powder having spherical particles in the size range 1-4 um. Cast iron saturates at around 5kG. Smudge

Peterae

Re: Meyer-Mace Isotropic NMR Generator « Reply #53 on: 2015-07-01, 18:44:09 »

Group: Administrator
Hero Member

Posts: 4635

Errr surely this has to be significant, Meyer uses 23MHz which just so happens to be 46/2

and

Quote

Interestingly, the change in the entropy of the nuclear spin system in the region
where the NMR and ferromagnetic-resonance frequencies are equal can actually be
used to cause cooling (by analogy with the nuclear-demagnetization method).

Meyer-Mace Isotopic NMR Generator

meyer.jpg (18.9 kB, 447x326 - viewed 1916 times.)

« Last Edit: 2015-07-01, 19:12:07 by Peterae »

Verpies

Re: Meyer-Mace Isotropic NMR Generator « Reply #54 on: 2015-07-02, 07:23:50 »

Group: Administrator
Hero Member

Posts: 4440

Quote from: Peterae on 2015-07-01, 18:44:09

Errr surely this has to be significant, Meyer uses 23MHz which just so happens to be 46/2

Maybe in his times, 46Mhz excitation was too expensive to achieve technologically and he just compromised at a lower subharmonic.

Anyway, for the nuclei to resonate at 46MHz they have to be subjected to a huge magnetic flux density of 33 Tesla, which happens to be the magnitude of the internal Hyperfine field mentioned in that article.

This fact is very significant because 33T overshadows any field that we could generate in our device, making our DC bias field rather insignificant in comparison.
Consequently, iron's NMR frequency will stay largely independent from the external biasing DC field ( only 0.03% change up to 0.6T ), which is very different from the behavior of other substances experiencing NMR.

Peterae

Re: Meyer-Mace Isotropic NMR Generator « Reply #55 on: 2015-07-04, 08:49:36 »

Group: Administrator
Hero Member

Posts: 4635

Quote

Maybe in his times, 46MHz excitation was too expensive to achieve technologically and he just compromised at a lower subharmonic.

Exactly what i was thinking

I think we can assume he did this with the 146KHz or whatever the frequency was with the first device and copper rod.

I would like to bring everyones attention to this post
http://www.overunityresearch.com/index.php?topic=3107.msg49921#msg49921
verpies has answered a whole bunch of questions and it really paves the way for a prototype device.
Thanks verpies for your time

I have ordered the mild steel for the device frame, still need to get quotes for the pure iron rod.
We now know the NMR frequency of operation and will use this oscillator LTC6905
It's a 17MHZ to 170MHz resistor set oscillator, we then have the ability to have multiturn pots for finer control, and could even feed a current source to control it.
http://www.linear.com/product/LTC6905

I think i will get some PCBs made up with the oscillator, fet driver and Class E amplifier once i have the prototype working.

The first device will have to be a lower temperature version, i am first aiming for 200 Deg C, i have ordered some Kapton tape.

Also really good news verpies looked up the decay method for the various isotopes of iron and they all decay using the beta process, no neutrons are emitted by the looks of things, so we have a relatively safe device to work with as well, this is very important for a viable device.

« Last Edit: 2015-07-04, 09:16:26 by Peterae »

Smudge

Re: Meyer-Mace Isotropic NMR Generator « Reply #56 on: 2015-07-04, 10:23:09 »

Group: Professor
Hero Member

Posts: 2306

Quote

The size and proportions of the fuel rod determine its demagnetization field (HD). Maybe Smudge can help us calculate it with an elongated ellipsoid formula for that shape.

I can do better than that. It is possible to use Nagaoka's inductance formula for a close wound single layer coil to deduce the demagnetization factor for a cylindrical rod. I do this in one of my papers but I am not at home right now so I can't access it. I'll post it when I get home.

I'll also look into the energy required to flip an atomic dipole through 180 degrees but off the top of my head the angular force on a dipole of moment m amp-meters^2 in a field of B Tesla is m*B*sin(theta) where theta is the angle between the dipole axis and the field. Thus the energy to flip it is given by integrating m*B*sin(theta) from 0 to 180 degrees. That gives a value of 2*m*B. And the value for m will be the Bohr magneton.

Smudge

Verpies

Re: Meyer-Mace Isotropic NMR Generator « Reply #57 on: 2015-07-04, 12:53:48 »

Group: Administrator
Hero Member

Posts: 4440

Quote from: Smudge on 2015-07-04, 10:23:09

Thus the energy to flip it is given by integrating m*B*sin(theta) from 0 to 180 degrees. That gives a value of 2*m*B. And the value for m will be the Bohr magneton.

Why the Bohr magneton and not the Nuclear magneton, which is 1836x smaller ...or better yet the Fe⁵⁷ magnetic moment which is 20300x smaller ?

Using the last number yields 18mJ of energy to flip 1 mole of Fe⁵⁷ nuclei at ~33T.
But how quickly is this energy dissipated and what is the needed RF power level in bulk Iron rod?

Peterae	Re: Meyer-Mace Isotropic NMR Generator « Reply #58 on: 2015-07-04, 22:39:10 »
Group: Administrator Hero Member Posts: 4635	Just starting to plan the Oscillator & Class E Amp + PCB, any advice on the fet driver stage for 45.6 MHz fet stage.

Verpies

Re: Meyer-Mace Isotropic NMR Generator « Reply #59 on: 2015-07-04, 23:08:27 »

Group: Administrator
Hero Member

Posts: 4440

Quote from: Peterae on 2015-07-04, 22:39:10

Just starting to plan the Oscillator & Class E Amp + PCB, any advice on the fet driver stage for 45.6 MHz fet stage.

Carefully. Millimeters between the driver and the MOSFET with minimal trace loop area there - just like Itsu.
Litz wire inductors. Air coils on plastic formers.

It is very hard to charge and discharge a huge gate capacitance quickly. High voltage gate drive helps a lot but is dangerous to the MOSFET.
At 45.5MHz a 2nF gate capacitance acts as a 1.7Ω resistor to ground to the gate driver. That's over 11A of average current for a gate drive voltage of 20V...and the instantaneous gate currents are even higher. With special RF transistors, e.g. this one and this one or this one, it is possible to cut that in half or ¼.

Verpies	Re: Meyer-Mace Isotropic NMR Generator « Reply #60 on: 2015-07-05, 07:23:49 »
Group: Administrator Hero Member Posts: 4440	Of course it is also possible to buy a ready 1kW LDMOS amplifier covering 2MHz - 54MHz on eBay for $280 (without a heatsink) Meyer-Mace Isotopic NMR Generator Amp 1kW.jpg (247.84 kB, 1600x931 - viewed 2095 times.)

Grumage	Re: Meyer-Mace Isotropic NMR Generator « Reply #61 on: 2015-07-05, 12:51:37 »
Group: Renaissance Man Hero Member Posts: 3095	Dear Peterae. With regard to our telephone conversation yesterday here is the link to the suggested material for thermal insulation. http://www.ebay.co.uk/itm/Ceramic-Fibre-Paper-1-mm-x-A4-Sheet-/321798556477?pt=LH_DefaultDomain_3&hash=item4aecb04b3d Cheers Grum. --------------------------- Nanny state ? Left at the gate !!

Peterae

Re: Meyer-Mace Isotropic NMR Generator « Reply #62 on: 2015-07-05, 16:35:38 »

Group: Administrator
Hero Member

Posts: 4635

Hi verpies

Thats a nice amp $380 with heatsink though, not funny if it blows.

The fet drivers that i have seen the data for do not seem to be rated for 45.5MHz ?

Found a mosfet that looks good DE375-102N12A
http://uk.farnell.com/ixys-rf/de375-102n12a/mosfet-n-rf-de375/dp/1347740?MER=baynote-1347740-pr

and one even cheaper DE275-IXZ210N50L
http://uk.farnell.com/ixys-rf/ixz210n50l/mosfet-n-rf-de275/dp/1347732?MER=baynote-1347732-pr

Grumage
Thanks for the link to that insulation, it will make a high temperature device easily workable.

« Last Edit: 2015-07-05, 17:21:29 by Peterae »

Verpies

Re: Meyer-Mace Isotropic NMR Generator « Reply #63 on: 2015-07-05, 17:51:07 »

Group: Administrator
Hero Member

Posts: 4440

Quote from: Peterae on 2015-07-05, 16:35:38

Thats a nice amp $380 with heatsink though, not funny if it blows.

No typo - if you read the description you will find out that it is $100 cheaper if you make the heatsink yourself.

Quote from: Peterae on 2015-07-05, 16:35:38

The fet drivers that i have seen the data for do not seem to be rated for 45.5MHz ?

But they are! For example the BLF528 transistor that the $280 amp board is based on, is rated up to 500MHz.

Quote from: Peterae on 2015-07-05, 16:35:38

Found a mosfet that looks good DE375-102N12A

...but C_ISS=2000pF

...and C_RS=30pF

...and V^GS(th)=5V ...and R_DS(ON)=0.9Ω

Smudge

Re: Meyer-Mace Isotropic NMR Generator « Reply #64 on: 2015-07-05, 19:41:38 »

Group: Professor
Hero Member

Posts: 2306

Quote from: verpies on 2015-07-04, 12:53:48

Why the Bohr magneton and not the Nuclear magneton, which is 1836x smaller ...or better yet the Fe⁵⁷ magnetic moment which is 20300x smaller ?

Using the last number yields 18mJ of energy to flip 1 mole of Fe⁵⁷ nuclei at ~33T.
But how quickly is this energy dissipated and what is the needed RF power level in bulk Iron rod?

Ooops, yes, my mind is fixated on flipping ferromagnetic dipoles. Of course you must use the moment of the dipole you want to flip. Do you want to flip it 180 degrees? In NMR it is usual to flip 90 degrees and that gives you the maximum free induction decay. If you want spin echoes then you use additional 180 degree flips to develop the echoes. Not sure what you want for this application.

While on the subject of NMR I am not an expert but my guess is that the large hyperfine field that determines the Fe57 frequency is aligned with the ferromagnetic dipole, like being inside the current loop that is the ferromagnetic dipole. If so then you need the DC bias to get aligned dipoles in order to get the right geometry of the RF coils. If the DC bias coils are around the rod to magnetize it along its length, then the nuclear dipoles lie along that axis and RF driving coils must provide a crossed field. You can use pairs of crossed coils to create a rotating RF field to get better drive into the nuclear dipoles.

The effective demagnetization factor for rods can be found in my paper "Energy around Coils and Magnets" attached. The reluctance of the external field through air acts like an mmf drop (like voltage drop across a resistor) to the flux in the magnetic closed circuit, and this gives the effect of demagnetization. Nagaoka's k factor allows you to calculate that mmf drop.

As regards the total energy required for flipping I think you need to first establish the RF requirement for your flip (either 90 degrees or 180 degrees). From what I remember of my work on NQR (which is similar to NMR) you need a certain pulse length of RF, in other words a certain number of RF cycles to pump the dipole into its new position. If you know that, and the magnitude of the crossed RF magnetic field, then it should be possible to estimate the losses. But it might be quicker to just measure them.

Smudge

Energy around Coils and Magnets.pdf (216.61 kB - downloaded 389 times.)

Peterae

Re: Meyer-Mace Isotropic NMR Generator « Reply #65 on: 2015-07-05, 20:15:35 »

Group: Administrator
Hero Member

Posts: 4635

Quote

But they are! For example the BLF528 transistor that the $280 amp board is based on, is rated up to 500MHz.

I thought the BLF528 is a power fet not driver, how do you drive this board, directly from the cmos output of the LTC6905.

Quote

No typo - if you read the description you will find out that it is $100 cheaper if you make the heatsink yourself.

I need to explore other cheaper ways to do this, with vat & import duty it's going to be approaching £240

I cannot use the BLF528 as that is £248 from farnell alone.

EDIT
Maybe we could soup this up a bit.
http://www.qsl.net/va3iul/Homebrew_RF_Circuit_Design_Ideas/50MHz_40W_MOSFET_PA_DJ9FG.gif

« Last Edit: 2015-07-05, 22:49:44 by Peterae »

Verpies

Re: Meyer-Mace Isotropic NMR Generator « Reply #66 on: 2015-07-06, 00:30:33 »

Group: Administrator
Hero Member

Posts: 4440

Quote from: Peterae on 2015-07-05, 20:15:35

I thought the BLF528 is a power fet, not driver

It is, but it has a sensitive gate (V_GS(th)=1.7V) with high transconductance, that makes it drivable directly.

Quote from: Peterae on 2015-07-05, 20:15:35

how do you drive this board, directly from the cmos output of the LTC6905.

You apply 46MHz 2Watts sine wave to the input (which is normalized at 50Ω) and get 800W output (also normalized at 50Ω).

Can the LTC6905 output 2Watts ? If "not", then a preamp is needed, e.g. like this or a plain vanilla 2W MOSFET driver followed by an LC filter, to convert its square wave to a sine wave.

Quote from: Peterae on 2015-07-05, 20:15:35

I need to explore other cheaper ways to do this, with vat & import duty it's going to be approaching £240

Of course, whatever works and costs the least ...but I am afraid that achieving 500W @ 45.5MHz might be more expensive any other way.
Let's observe how Itsu fares with non-RF components at 13.6MHz, which is a frequency much easier to rein-in.

Quote from: Peterae on 2015-07-05, 20:15:35

I cannot use the BLF528 as that is £248 from Farnell alone.

Yup, apparently Farnell is ripping us off, if that Russian seller can offer the BLF578 and the pretuned PCB with associated components for $280.

Quote from: Peterae on 2015-07-05, 20:15:35

Maybe we could soup this up a bit.
http://www.qsl.net/va3iul/Homebrew_RF_Circuit_Design_Ideas/50MHz_40W_MOSFET_PA_DJ9FG.gif

Hmm, for 500W you'd need to have 13 of these amps and add phase coherent power combiners to combine their output. The input power requirement would go up 13x too - I don't think the LTC could handle thirteen of them without a preamp.
How expensive would that be?

...and don't forget the 50V power supply capable of delivering at least 20A of ripple free DC that is needed to power any of these RF amplifiers

Verpies

Re: Meyer-Mace Isotropic NMR Generator « Reply #67 on: 2015-07-06, 05:44:36 »

Group: Administrator
Hero Member

Posts: 4440

Quote from: Smudge on 2015-07-05, 19:41:38

...my guess is that the large hyperfine field that determines the Fe57 frequency is aligned with the ferromagnetic dipole, like being inside the current loop that is the ferromagnetic dipole.

It would seem so but don't forget about the contribution of the total domain wall volume described in that Fe NMR paper.

Quote from: Smudge on 2015-07-05, 19:41:38

If you know that, and the magnitude of the crossed RF magnetic field, then it should be possible to estimate the losses. But it might be quicker to just measure them.

The problem is the the RF amplifiers need to be sized before the experiment.
I think Peterae calculated some ~400W/mole but that needs to be verified because these amplifiers are expensive and it is not easy to upgrade their power once the wrong one is chosen.

Quote from: Smudge on 2015-07-05, 19:41:38

If so then you need the DC bias to get aligned dipoles in order to get the right geometry of the RF coils. If the DC bias coils are around the rod to magnetize it along its length, then the nuclear dipoles lie along that axis and RF driving coils must provide a crossed field.

I am aware of all that, but I am intrigued by the value of the polarizing field chosen in the Mayer patent (0.5T), which is right in the middle of the "domain rotation zone".
In that "zone" many of the 33T hyperfine field vectors will not be parallel to the external polarizing field along the length of the rod.

Also, all segmented coaxial solenoids operating in the bucking mode will produce between them a component of the magnetic flux that is perpendicular to their axis as depicted in Diag.4a and Diag.4b below. The Mayer patent suggests such layout of its windings.

Meyer-Mace Isotopic NMR Generator

CtrTapCoaxialWindings.png (27.24 kB, 948x580 - viewed 1643 times.)

Verpies

Re: Meyer-Mace Isotropic NMR Generator « Reply #68 on: 2015-07-06, 06:09:20 »

Group: Administrator
Hero Member

Posts: 4440

Quote from: Smudge on 2015-07-05, 19:41:38

You can use pairs of crossed coils to create a rotating RF field to get better drive into the nuclear dipoles.

Of course that there are more efficient coil arrangements for causing NMR.
I think, that they should be attempted if the coil arrangement stated in the patent does not yield the expected results.

Note, that NMR is not the ultimate goal here - stimulated beta decay is (along the lines of this and this and this phenomenon) ...and Lorentz confinement of the resulting beta decay products by the DC bias field.
The coil arrangement depicted in the Mayer patent might be more optimal for this goal than for the NMR alone.

Meyer-Mace Isotopic NMR Generator

Beta Decay Asymmetry.png (25.5 kB, 387x220 - viewed 1617 times.)

Verpies

Re: Meyer-Mace Isotropic NMR Generator « Reply #69 on: 2015-07-06, 15:22:27 »

Group: Administrator
Hero Member

Posts: 4440

Quote from: Peterae on 2015-06-29, 18:53:43

So things we need to nail down & choose:
...
3) Static NMR Frequency field strength +/- ?Tesla's or in terms of watts power would do.

I just noticed that e2matrix's translation of the French Mayer patent contains an approximate answer to this question:

Quote

The second coil (4) is traversed by a sine wave of 21 MHz frequency of 10^-4 Tesla which is an activator of Nuclear Magnetic Resonance observer permitting rotation of 180° of the nuclear spins of the iron nuclei.

Peterae	Re: Meyer-Mace Isotropic NMR Generator « Reply #70 on: 2015-07-06, 17:28:07 »
Group: Administrator Hero Member Posts: 4635	Hi Verpies Quote The second coil (4) is traversed by a sine wave of 21 MHz frequency of 10-4 Tesla which is an activator of Nuclear Magnetic Resonance observer permitting rotation of 180° of the nuclear spins of the iron nuclei. I thought i had seen that somewhere but when i went back could not find it, must be the first patent then i think? Seems small for pumping 400 watts in at 23MHz

Peterae

Re: Meyer-Mace Isotropic NMR Generator « Reply #71 on: 2015-07-06, 18:37:14 »

Group: Administrator
Hero Member

Posts: 4635

How about this amp £37 MRF151
http://uk.mouser.com/ProductDetail/MACOM/MRF151/?qs=3Wmz%2FrCSAaHwKuWRW8NS6A%3D%3D from mouser

CISS=350pF and CRS=15pF and VGS(th)=3V and RDS(ON)=???

Circuit
http://www.qsl.net/va3iul/Homebrew_RF_Circuit_Design_Ideas/VHF_300W_MRF151G_PA.gif

Just says VHF maybe wideband?

and another with VRF151 same price, sounds a bit like the same device
Circuit
http://www.qsl.net/va3iul/Homebrew_RF_Circuit_Design_Ideas/HF_300W_VRF151_MOSFET_PA.gif

« Last Edit: 2015-07-06, 19:08:33 by Peterae »

Smudge

Re: Meyer-Mace Isotropic NMR Generator « Reply #72 on: 2015-07-06, 19:42:00 »

Group: Professor
Hero Member

Posts: 2306

Quote from: verpies on 2015-07-06, 05:44:36

I am aware of all that, but I am intrigued by the value of the polarizing field chosen in the Mayer patent (0.5T), which is right in the middle of the "domain rotation zone".
In that "zone" many of the 33T hyperfine field vectors will not be parallel to the external polarizing field along the length of the rod.

The chart showing that zone shows the magnetization to be between 90% and 100% of saturation value so most of the ferromagnetic dipoles hence also nuclear dipoles are aligned.

Quote

Also, all segmented coaxial solenoids operating in the bucking mode will produce between them a component of the magnetic flux that is perpendicular to their axis as depicted in Diag.4a and Diag.4b below. The Mayer patent suggests such layout of its windings.

But the three coils shown in the patents have different requirements and are not connected as depicted in your diagrams. One is for DC bias, one is for RF drive and the third is for AC output.

I am mystified by the axial RF drive coil. If my reasoning is correct and the nuclear dipoles are aligned along the magnetization axis then you need cross coupling to drive them. Note that all the nuclei are already precessing but at random phases so the net effect is zero. To pump them with RF you need to access their off-axis component that is rotating at their resonant frequency, not their on-axis component which is constant. They can then absorb energy over a number of cycles that tries to speed up the precession process, creating the torque to ultimately flip them through the desired angle. The on-axis RF magnetic field does not do this. Maybe it is the radial component of the RF field at the ends of the drive coil that do the work, which means nuclei flip only over limited regions of the rod. Also I do wonder whether the patents reflect someone's aspirations rather than an accomplished and verified experiment.

Smudge

Peterae

Re: Meyer-Mace Isotropic NMR Generator « Reply #73 on: 2015-07-06, 19:54:48 »

Group: Administrator
Hero Member

Posts: 4635

Coils 8 & 9 are indeed connected in bucking mode, I believe the whole device is supplied with 50Hz AC into point 12 , since the ac current flows through the iron rod feeding coil 9 with 50Hz AC current in antiphase from the exiting AC current in Coil 8, the centre tap seems to feed the DC bias coil, i don't understand how a DC current flows through the Bias coil unless a DC potential is the result o beta decay and the series path of the rod itself.

It looks to me as if the load which has the generated current flowing which governs the 50HZ AC current in the coils 8&9, so if more current is needed the device automatically adjusts the AC current in coils 8&9 and thus drives harder to supply the extra current needed.

Meyer-Mace Isotopic NMR Generator

meyer.jpg (36.32 kB, 723x428 - viewed 1567 times.)

Verpies

Re: Meyer-Mace Isotropic NMR Generator « Reply #74 on: 2015-07-06, 23:50:47 »

Group: Administrator
Hero Member

Posts: 4440

Quote from: Peterae on 2015-07-06, 18:37:14

How about this amp £37 MRF151
http://uk.mouser.com/ProductDetail/MACOM/MRF151/?qs=3Wmz%2FrCSAaHwKuWRW8NS6A%3D%3D from mouser

This one will perform exactly as stated in the datasheet.

Note, the power gain @ 50MHz is 19dB, which means that 1.8W of input power will be needed to achieve 150W output power.
For 500W you'd need 4 of these amplifiers + 3 phase coherent power combiners. The combined input power will be around 7W.

Quote from: Peterae on 2015-07-06, 18:37:14

Just says VHF maybe wideband?

The 150MHz qualifies as VHF. 150MHz is merely the upper frequency limit of this transistor.
I am sure it can perform as advertised in a wideband circuit between 1MHz and 150MHz.