PopularFX
Home Help Search Login Register
Welcome,Guest. Please login or register.
2024-11-23, 07:07:46
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 3 4 5 6 [7] 8 9
Author Topic: Meyer-Mace Isotopic NMR Generator  (Read 115609 times)
Group: Tech Wizard
Hero Member
*****

Posts: 1193
Hi Peter,

You could terminate the output of the DDS with a carbon or metal film resistor where you see the missing one on the PCB, the output of the multiple stage filter should be terminated anyway.
OR for this measurement, temporarily use a 51 Ohm or 56 Ohm or close to this value a resistor across the end of the coax where it meets with the first 3.3 pF capacitor leg.
The strange behaviour you find may come from the unterminated DDS output and the piece of coax cable you use to drive the coil via the 3 series capacitors.   
Will be back later.

Gyula
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
OK fitting a 51Ohm resistor across the coax output just before the 3.3pf caps has helped, although that strong peak is still there.
Are we expecting a bigger amplitude across the Inductor as we don't it's smaller.

Yellow trace is our LC and cyan trace is across the 51Ohm termination just before the 3off 3.3pf caps

So we are getting 100mV pk-pk across LC & 528mv pk-pk across the 51Ohm termination

These scope probes will have just over 12pf each i think
   
Group: Tech Wizard
Hero Member
*****

Posts: 1193
I assume the trimmer capacitor was still set at its minimum value possible, right?

You could reduce the effect (if any) of the self capacitance of the scope probe which is across the coil by placing a few pF capacitor in series with the hot input of the probe, even a gimmick style capacitor would do. This series capacitor will divide down the voltage across the coil but the strong peak may get shifted away in frequency.

Yes, a bigger amplitude across the coil versus across the terminating resistor would indicate the LC tank impedance would be higher than 51 Ohm, this would be desirable and this would mean the Q of the coil would be higher than 1.

By the way, if your trimmer cap is at its minimum, it may have say 5 pF capacitance and adding to this the 12 pF probe capacitance and adding some further stray capacitance of say 3 pF, these total as 20 pF. And assuming the measured 0.58 uH coil inductance more or less remains the same (maybe unlikely but may be close to that), then the resonant frequency is 45 MHz from the formula if the tuning capacitor is 21.6 pF. Here is an online calculator I used:
http://www.daycounter.com/Calculators/LC-Resonance-Calculator.phtml

From the video sweep it seems this core performs miserably in the MHz range for sure. The coil has a quality factor under 1 it seems.

Gyula
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
Quote
I assume the trimmer capacitor was still set at its minimum value possible, right?
Each time i try to adjust the trimmer for max amplitude at 45MHz, so the answer is no, i have not bothered to measure the trimmed value though because i would need to unsolder it.

I just tried swapping the 3off 3.3pf in series for a 6.8pf cap and that large top end peak has now gone.
See Video shortly, just uploading
https://youtu.be/ZHM_48rD3yo

Also i just took a snap with the trimmer adjusted for 45Mhz with the 6.8pf and we are just under the driving amplitude now.
LC=468mv pk-pk and 51Ohm termination is 516mv pk-pk

   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
I just disconnected the series C and left just the trimmer cap across the inductor with a scope probe, using my dip meter near the coil i set to 45Mhz and adjusted the trimmer cap for max amplitude using the scope chan, i then unsoldered the trimmer from the Coil and reconnected the scope probe and measured the capacitance using the LCR at 46pf, but i had to disconnect the scope probe from the BNC on the oscilloscope end as it seemed to make the measured cap value unstable.

The LCR measured Inductance was probably wrong as it was at the bottom end of it's ability to measure.

That now gives us a calculated value of 266nH which is much closer.
   

Group: Professor
Hero Member
*****

Posts: 3494
See Video shortly, just uploading
https://youtu.be/ZHM_48rD3yo
What do the yellow and blue trace represent?
Could you show probe positions on a schematic?
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
Sure
I've been using the circuit provided by gyula


Cyan is TP1 and Yellow Tp2

PS sorry about fuzzy video, it appears there is no way for me to focus on video but on stills it auto focuses, Samsung Galaxy, nothing in settings and others are complaining as well by the looks of it.
   
Group: Tech Wizard
Hero Member
*****

Posts: 1193
Peter,  when all the possible tests that may come while the coil is on the core are over, and you remove the coil from the core to get an air core, it maybe would be a good step to repeat the test and experience how the tuning would behave.  Just a thought, if you disagree, I understand.

Some notices:
The 6.8 pF capacitor has 520 Ohm reactance at 45 MHz versus the 3536 Ohm of a 1 pF, so its shunting effect in both the input and output directions is higher (thus it can attenuate unwanted resonances) while it allows more input voltage to reach to the LC tank.
Because the 520 Ohm series reactance is still 10 times higher than the 51 Ohm terminating resistor (assuming the latter is still near to 50 Ohm at 45 MHz) this comparable voltage amplitude result shows the LC tank impedance must be higher than 51 Ohm I thought because after a voltage divison only a higher (i.e. resonant) impedance can 'augment' the divided voltage to be nearly equal to the input voltage.

The core characteristics will invariably change when the 0.5 T bias field from the 70 turn 1A coil will be present. 

One more thing while the coil is on the core: Perhaps a 3 dB bandwidth test would reveal the Q of the tank at 45 MHz? to do this, notice the amplitude of the resonant voltage across the coil, say it is 468 mVpp (from the latest scopeshot) and you detune the DDS from the exact 45 MHz down in frequency till the amplitude across the coil will be 0.707 times only of that at 45 MHz and log that frequency, then repeat the detuning upwards till the amplitude again reduces to 0.707 times and log the upper frequency too.  Now if you substract the lower frequency value from the upper one, you get the 3 dB bandwidth, BW.  Then the Q is received by dividing 45 MHz by the BW (BW should be also in MHz of course.)

Gyula
   

Group: Professor
Hero Member
*****

Posts: 3494
I've been using the circuit provided by gyula
It is one of the ways to determine the suitable parallel capacitance of the LC tank.  The resonance is not very sharp so your Q is low or you are way off into one of the SG's harmonics.

Don't disregard the phase information between the SG voltage and the voltage across the LC tank, since it indicates the rate of the energy transfer between the SG and the LC tank. You can visualize it better by putting your scope into XY mode.
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
Thanks for your time on this gyula & verpies also.

It will be very complex to get this system working in a resonant way, because also we should be heating the iron to above 150 Deg C and we will also have a 50Hz bucking 2 coil system operating as well as the bias field.

Quote
One more thing while the coil is on the core: Perhaps a 3 dB bandwidth test would reveal the Q of the tank at 45 MHz? to do this, notice the amplitude of the resonant voltage across the coil, say it is 468 mVpp (from the latest scopeshot) and you detune the DDS from the exact 45 MHz down in frequency till the amplitude across the coil will be 0.707 times only of that at 45 MHz and log that frequency, then repeat the detuning upwards till the amplitude again reduces to 0.707 times and log the upper frequency too.  Now if you substract the lower frequency value from the upper one, you get the 3 dB bandwidth, BW.  Then the Q is received by dividing 45 MHz by the BW (BW should be also in MHz of course.)

Sounds good i can do that next time i am in the workshop  O0


   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
Just ordered some power modules

2 off for constant current bias drive. and DDS / preamp
http://www.ebay.co.uk/itm/301737802598?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

and 400 watt 10 amp for power amp board
http://www.ebay.co.uk/itm/181863152703?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

and some ceramic fiber paper for insulation 1mm suggested by Grumage.
http://www.ebay.co.uk/itm/181863152703?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

and a SMA male/male coupler for DDS to Preamp
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
In the Czech patent he says the following
Quote
Applying energy to the strength 168,21x10-13 on the iron
core, formed by 56 Fe isotope , is called instability , which in turn causes
radioactivity , which amends Fe 56 Fe 54 while releasing two neutrons.

Now we previously worked out that the described output power was per atom.
Quote
This procedure yields about 32,040x10-19 J energy,

So what is the significance of this 168,21x10-13 because this is surely more energy than is released by 32,040x10-19.

Any thoughts

and strangely there are not enough digits after the comma to mean thousands 168,21?

Also from the FR 2nd patent he says
Quote
The second coil (4) is traversed by a sinusoidal wave of 21 MHz and 10-4 Tesla power which is a nuclear magnetic resonance activator allowing 180 ° rotation of the nuclear spine
iron atoms.

Interestingly he also states in the 3 coil system that the 3rd coil runs the length of the core.
Quote
The third coil (5) is a transformer primary which collects this induction energy at all points
of the bar established by the isotopic mutation iron atoms 56 in iron 54.

and an explanation of operation
Quote
The present invention uses a physical phenomenon that we have highlighted and which we call
'Isotopic Mutation ".
Description of the physical principle applied to the isotope iron-56: 56 The isotope iron contains 26
protons, 26 neutrons and 30 electrons, its total mass is 56.52 Mev, its actual mass is 55.80 Mev for.
The difference between the total mass and the real mass is 0.72 Mev, which corresponds to a
binding energy per nucleon of 0.012857 Mev.
If one introduces an additional 105 ev of energy to the iron core isotope 56, this one will have a
cohesion energy level of 0.012962 Mev nucleon corresponding to iron isotope 54. The instability
created by this energy input is determined a radioactivity that will transfer the isotope iron-56
isotope with 54 neutrons release 2 which transform a 9-minute Hydrogen by natural radioactivity.
This process will generate an energy gain of 20,000 ev since the mass defect of the resulting iron 54is only 0.70 Mev instead of 0.72 MeV for the iron isotope 56.
To bring to the iron core isotope 56 the energy necessary to perform the isotopic mutation, we use
the principle of nuclear magnetic resonance.
The 26 protons of the isotope 56 iron are at the origin of the lively nuclear magnetic moment of a
gyroscopic movement depending on the actual weight of the iron core. The weight loss caused by
the isotopic mutation phenomenon will change the moment gyro and release energy by increasing
the speed of rotation.
The physical phenomenon of isotopic mutation described above is applicable to all bodies of the
table Mandeleiev.
   

Group: Professor
Hero Member
*****

Posts: 3494
In the Czech patent he says the following:
Quote
Applying energy to the strength 168,21x10-13 on the iron core, formed by 56 Fe isotope , is called instability , which in turn causes
radioactivity , which amends Fe 56 Fe 54 while releasing two neutrons.

Now we previously worked out that the described output power was per atom.
So what is the significance of this 168,21x10-13 because this is surely more energy than is released by 32,040x10-19.
Any thoughts
I think the confusion stems from that piss-poor translation from Czech.
I think the word "core" should be "nucleus" and the word "amends" should be "transmutes" and "strength" could be "force".
If the phrase "applying energy" refers to adding energy to the nuclear binding force, then the 168.21x10-13 refers to the total nuclear binding energy or force.

IMO the device described in this patent was engineered first and then a slipshod theoretical explanation was added as an afterthought to complete the patent.

Interestingly he also states in the 3 coil system that the 3rd coil runs the length of the core.
Does he?
I read it that it "collects" only where the transmutation occurs.  But from EM induction theory we know, that spatial localization of induction just does not happen.  A winding reacts to flux changes from the entire flux penetrating the winding, regardless of its position along the path of this flux.  The flux does not even have to touch the winding ...see the A vector potential and all.
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
Quote
Does he?
I read it that it "collects" only where the transmutation occurs.  But from EM induction theory we know, that spatial localization of induction just does not happen.  A winding reacts to flux changes from the entire flux penetrating the winding, regardless of its position along the path of this flux.  The flux does not even have to touch the winding ...see the A vector potential and all.
Oh Dam my mistake.
hence the almost closed metal frame for magnetic path, so in fact this coil can be placed anywhere along the core  O0

Quote
I think the confusion stems from that piss-poor translation from Czech.
yes indeed makes a lot of sense
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
OK with the 6.8pf in series feeding the tuning cap in parralel with my 4 turn Inductor, i have a max pk-pk of 500mv @ about 45.6MHz at the moment, having done a real slow scan it looks like the bottom 353mV is hit around 42.1MHz and the top 353mV hit at around 49.4Mhz

Centre to top difference is 49.4-45.6 = 3.8Mhz
Centre to bottom difference is 45.6-42.1 = 3.5Mhz

So centre is not quiet centre but close enough considering the res of the capture is not perfect at this mv range.

So we have a bandwidth of 49.4-42.1 =  7.3Mhz

and a Q of 45.6/7.3 = 6.24

PS tomorrow i will make my Bias coil and drive with 1 amp to see how things get affected  O0

EDIT
Maybe a different designed coil would be better, more spacing between the 4 turns or try 2 turns and a bigger tuning cap for our purpose here ??

I think i have enough litz left over to try a 2 or 3 turn coil with spacing between turns.
   

Group: Professor
Hero Member
*****

Posts: 3494
Maybe a different designed coil would be better, more spacing between the 4 turns or try 2 turns and a bigger tuning cap for our purpose here ??
Study this for the best coil proportions.

FYI: Below is a graph of different LC frequency responses for different Qs.




...and one of the ways the Q is calculated:

« Last Edit: 2016-02-05, 22:47:47 by verpies »
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
It would appear that litz is not much good at such high frequency,  because of the large amounts of dielectric material that cause low Q, looks like a better option is tube

and if we look at a skin depth chart stainless steel comes out way ahead i think followed by Nickle Silver then Steel & Tin.

Not yet sure how practical or how easy it will be to get hold of and bend.

S.Steel seems to be about 55um @ 45Mhz
   

Group: Professor
Hero Member
*****

Posts: 3494
It would appear that litz is not much good at such high frequency,  because of the large amounts of dielectric material that cause low Q,
No, because the different strands of the Litz cable are at the same potential, so the dielectric does not form a capacitor between the neighboring strands. 
Electric potential differences occur only between turns and layers, so that is where the dielectric loss does the most damage to the Q.

Therefore Litz still rulez.

Current crowding (proximity effect) is still a large factor so turn-to-turn spacing matters.
« Last Edit: 2016-02-06, 00:13:59 by verpies »
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
I am yet to find any calculation tables that go over 2MHz for Litz, below this frequency i can find data.

and at https://en.wikipedia.org/wiki/Litz_wire
Quote
Litz wire is a type of cable used in electronics to carry alternating current. The wire is designed to reduce the skin effect and proximity effect losses in conductors used at frequencies up to about 1 MHz.


and when you do a search for coils at 50MHz everyone seems to use copper or silver coloured tube


   

Group: Professor
Hero Member
*****

Posts: 3494
and when you do a search for coils at 50MHz everyone seems to use copper or silver coloured tube
That's an Argumentum ad Numerum
The Skin Effect does not stop just because the outside surface of a conductor is a part of a tube.

It is true that a waveguide, such as a coaxial cable, is less lossy than a Litz cable in transporting signals above 2MHz from one point to another.
...and internally silver coated waveguides are good for microwaves.

The purpose of our helical winding is not to transport a signal from one end of the winding to the other.  Instead its purpose is to create a variable magnetic field inside this winding.

   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
Fair enough
I will try making a spaced 4 turn with the litz i have left over and we can then compare  O0
   

Group: Experimentalist
Hero Member
*****

Posts: 782
Believing in something false doesn't make it true.
I am yet to find any calculation tables that go over 2MHz for Litz, below this frequency i can find data.

and at https://en.wikipedia.org/wiki/Litz_wire

and when you do a search for coils at 50MHz everyone seems to use copper or silver coloured tube



I think one of the reasons most of the time solid copper wire is used for higher frequencies is because the solid copper wire is easier to work with.   To wind a coil for 50 mhz you just take a piece of number 12 or 14 gauge wire and wrap it around a round bar or small pipe of the appropriate size and you have a coil that will hold it's shape when taken off the pipe or bar.  For litz you have to have some way to support the wire to get it to retain the shape you want.  Also at vhf frequencies you can tune the circuit just be spreading or compressing the turns of the solid copper wire coil.

It would be interesting to see what difference in inductance there would be between coils if one had litz and the other solid copper wire.  I haven't seen any charts that show that.

Carroll


---------------------------
Just because it is on YouTube does not make it real.
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
A New spaced coil only had enough homemade litz for 3.25 Turns though

49.4-45.9=3.5
45.9-42.5=3.4


bandwidth 6.9MHz
Q = 45.9/6.9=6.65

Not that much change really

Also made a Bias coil 40 Turns of 1mm wire, 22 turns first with 18 turns on top, 2cm wide,
measured inductance is 48uH, now this measured inductance on the core has me worried, there seems to be something amiss .
I am using this calculator to design my coil magnetic field, for iron i should be using a relative permeability of 200 and that's how i arrived at 40 turns for 0.5T field strength but according to the same calculator my measured inductance should be 8mH
http://www.calctool.org/CALC/phys/electromagnetism/solenoid

Now if i put a relative permeability of 1 into the calculator instead of my cores 200 i get roughly the measured inductance, but then my magnetic field strength is only 0.002T which is way too low, so i am currently not sure whats wrong here but it looks like my field strength is way too low and i need to find some other calculator software to check this out.
at 1 Amp should give me

I've increased the winding diameter to 30mm to allow for a heater on top of the core.

« Last Edit: 2016-02-07, 11:17:06 by Peterae »
   

Group: Administrator
Hero Member
*****

Posts: 3959


Buy me some coffee
Oh boy it gets worse, according to this site

https://en.wikipedia.org/wiki/Permeability_%28electromagnetism%29

Iron (99.8% pure)       Permeability µ [H/m]= 6.3×10−3    and relative = 5000

so now in that calculator for 40 turns i have 12T field and an inductance = 201mH

and then if i go here and enter my values to estimate the inductance of an air core
http://www.circuits.dk/calculator_multi_layer_aircore.htm

It agrees with my measured inductance, so this begs the question why is my bias coil behaving as if there's no core present??

In other words why does my core appear to have the same permeability of free space, is it because the coils inner diameter is 1cm bigger than the outer diameter of my core.  :D


EDIT just measured it's inductance off the core at 39uH @10khz & 48.9 on the core @ 10Khz

if i test it at 120 Hz i get 61uH off the core and 155uH on the core @120Hz
   

Group: Professor
Hero Member
*****

Posts: 3494
A New spaced coil only had enough homemade litz for 3.25 Turns though
Was the turn spacing equal to the wire thickness?

49.4-45.9=3.5
45.9-42.5=3.4
bandwidth 6.9MHz
Measured like this?

Q = 45.9/6.9=6.65
Not that much change really
The Q might be dominated by other circuit components, e.g. the capacitor or leads.

Also made a Bias coil 40 Turns of 1mm wire, 22 turns first with 18 turns on top, 2cm wide,
measured inductance is 48uH, now this measured inductance on the core has me worried, there seems to be something amiss .
I am using this calculator to design my coil's magnetic field, for iron I should be using a relative permeability of 200 and that's how i arrived at 40 turns for 0.5T field strength but according to the same calculator my measured inductance should be 8mH
According to this, the permeability of 99.8% iron should be around 5000.

Now if i put a relative permeability of 1 into the calculator instead of my cores 200 i get roughly the measured inductance, but then my magnetic field strength is only 0.002T which is way too low, so i am currently not sure whats wrong here
The Eddy currents in a conductive core can decrease the measured permeability.  Check if the measured permeability changes drastically with the frequency of inductance measurement.  Negative strong frequency dependency suggests, that the Eddy currents in the core are repelling the AC magnetic filed created by the inductance meter.

It agrees with my measured inductance, so this begs the question why is my bias coil behaving as if there's no core present??
Eddy currents or measurement error is all I can think of.  Eddy currents repel the magnetic field (or freeze it).
Especially the Eddy currents in the steel bracket because it has lager dimensions than the rod.

In other words why does my core appear to have the same permeability of free space, is it because the coils inner diameter is 1cm bigger than the outer diameter of my core.
No.  Try putting a piece of ferrite inside that coil and see if the inductance shoots up.  Ferrites do not support Eddy currents.  Also Eddy currents do not exist at DC so you can expect that the constant magnetic field developed in the core/rod at DC will be radically different from an AC field.

I've increased the winding diameter to 30mm to allow for a heater on top of the core.
That's good.  Larger coils immerse the rod in a more homogeneous field towards their center.  See the picture below:


« Last Edit: 2016-02-07, 12:25:07 by verpies »
   
Pages: 1 2 3 4 5 6 [7] 8 9
« previous next »


 

Home Help Search Login Register
Theme © PopularFX | Based on PFX Ideas! | Scripts from iScript4u 2024-11-23, 07:07:46