Hi,
Here's a small section taken from the encrypted file introducing the source of MEMM's (Magnetocaloric Energy Mover Machine) energy. There you'll see the math to calculate the Magnetocaloric energy exchanges occurring in materials.
MEMM is based on the Magnetocaloric effect, a very well known effect and discovered a century ago. Very few people, even scientists, realize just how much energy is exchanged during each quarter cycle in magnetic materials. For example, consider an iron toroid with copper windings. The effective permeability of a pure iron toroid is considerably high and requires a small amount of current; i.e., nowhere near 50 megawatts. The current in our example is a sine wave. Due to the permeability of iron the field fluctuates in our example between 1 and -1 Tesla's. We know that the Magnetocaloric effect for pure iron is 2 K per Tesla. In other words, when the field increases by one Tesla then the material heats up 2 K. When the field is removed the material cools down by 2 K. This has been used as an efficient powerful means of deep freezing. Back to our example, the signal oscillates at 100 KHz. So when the field goes from 0 to 1 T the material heats up 2 K. When the field falls from 1 to 0 T the material cools down. When the field goes from 0 to -1 T the material heats up. When the field goes from -1 to 0 T the material cools down. We will refer to each of these steps as one energy exchange; i.e., an exchange of energy. So there are four energy exchanges per sine wave. There are 100 thousand sine wave cycles per second. So there are 4 * 100000 = 400000 energy exchanges per second. The specific heat capacity of Iron is 25.10 J/(mole*K). J = Joules and K = Kelvin. Lets say our toroid is one cubic inch = 16.39 cc. The molar volume of iron is 7.09 cc/Mole. So 16.39 cc / 7.09 cc/mole is 2.312 moles. So 2.312 moles * 2 K * 25.10 J/(mole*K) is 116.1 Joules of energy per energy exchange. So at 400000 energy exchanges per second we get 116.1 J * 400000 x/s where x is energy exchange, which is 46.44e+6 J/s. J = watts * seconds, so 46.44e+6 J/s is 46.44 megawatts!
46.44 million watts of energy exchange per second occurring within one cubic inch of iron is astonishing, but very real! If we calculate the amount of energy exchanges in Gadolinium we arrive at almost the same value of 48.79 megawatts. The magnetocaloric effect for Gadolinium is 4 K/T
Kind regards,
Paul Lowrance
Hi tao,
I began designing the MEMM a little over a month ago and took a look at it and basically said, "Hey, this is the MEG!" Also another person by the name of Marcus or is it Markus uses the same effect. Since then the design is evolving into another form that should be more effective than the MEG. When the energy from MEMM is fully recirculated and confirmed you'll be one of the first to see the entire design freely published in extreme detail.
According to Naudin and others the MEG generates "free energy," but still has not been able to close the loop. After looking at Naudin's scope pictures I noticed he is making several errors in interpreting his scopes output energy, but these errors are mostly in the silicon iron version. Most of the scope pics on his Metglas version are very close and indeed "free energy." The problem in replicating the MEG is (if my theory is correct) in the UHF (ultra high frequencies). You can build 100 of Naudin's machines, which includes wrapping a large coil with over 2000 windings, and end up with 100 different machines. As you know, the capacitances involved in the windings can differ. Depending on the core material the magnetocaloric frequencies could typically be in the hundreds of MHz. The wrong impedance can kill such a device at UHF frequencies. Naudin used a modified carbon resister, which was created by high voltage I believe. That in itself can have some type of small high-frequency unidirectional characteristics. But replace that 100K custom carbon resister with a MOSFET and you'll most likely have a short at 500 MHz. The load is critical in absorbing the UHF energy. Also if the magnetic material does not have heat syncs and proper air flow then the material can go in temperature shock, which would prevent the machine from working too long.
Also there are other issues that come along with ultra high permeable materials when considering the magnetocaloric energies. For example when studying wave mechanics we see that such materials prevent nearly all the waves from escaping the core. This is akin to a wave traversing in ultra high reflective index and trying to enter a low reflective index material. The wave simply reflects. In such cases wires that are very close to the core are effective in absorbing the energy. Perhaps magnetic wire could help improve this issue.
So the above issues may have prevented Naudin from closing the loop. One thing seems for certain, his scope shots seem clear, showing enough information to conclude his Metglas device (not sure about the iron version) was generating "free energy."
You were asking how to tap into this magnetocaloric energy. I have never posted this secret, but it's perhaps time to at least post part of this bit of information. The answer is threefold:
The Secret:
1. Saturation. To be highly effective the device needs to saturate the core. A fully saturated core prevents the intrinsic electron spins from absorbing the magnetocaloric energy. Of course a fully saturated core is useless, but no realistic coil can saturate magnetic material. The core should be close to saturation.
2. Ultra high di/dt current that causes the net magnetic field to increase within the magnetic material such as Metglas. To do this you need a permanent magnet and current that both oppose. The field from the magnet needs to be stronger than the coils field. This flips the process and allows the device to collect the cores energy when the magnetocaloric effect is in is radiating cycle. The high di/dt causes a higher percentage of the electron spins to flip simultaneously, which in turn greatly reduces the cores ability to absorb MCE energy, which allows more of the energy to escape the core. Normally MCE energy is mostly absorbed within micrometers of the originating electron flip.
3. Ultra thin cores. You want the core to be as thin as possible. The thinner it is the less it can absorb the MCE energy.
A person, perhaps kator, mention danger with the MEG. There is no MEG danger with ultrasonics. The only danger with this technology is preventing the core from absorbing 50 megawatts of UHF energy and thereby escaping the core. Although the MEG is merely allowing an infinitesimal amount of the MCE energy to escape, which is probably safe, but it is best to shield core just in case. A 50 megawatt burst would fry any organic material!
Paul Lowrance
MCE = Magnetocaloric effect
Hi,
A quick peswiki page was created for the MEMM at
http://peswiki.com/index.php/Directory:MEMM
I'll try to keep it up to date.
Paul Lowrance
If I understand what you are trying to convey, the iron will heat up or cool down while experiencing a change of 1 tesla. But isn't the end result going to cancel itself? If a change of 1 tesla causes a change in temperature in one direction, and the opposite change of 1 tesla causes an opposite change in temperature; isn't the temperature change overall zero?
It appears to me that the changes in both directions are counted in the calculation and amount to the approx. 44 megawatt range. But it would seem that the overall result is still zero unless a method is found to realize this change where the effect adds together in one direction only.
Not trying to disagree, but just trying to understand the point. Can you help me? Have you found a way to realize the change in one direction only without the cost of power input?
Hi Liberty,
Yes, you are entirely correct about the MCE (Magnetocaloric Effect) in that the net temperature change is zero. Yet there's a great deal of energy being exchanged, back and forth, and the MEMM and MEG designs are machines that tap into that "free energy."
I have mentioned Iron on numerous occasions because it's the example most used in describing MCE. I don't know why since Iron has very little MCE at room temperature. Here are two reasons why Iron is a poor choice:
1. Iron has a very high Curie temperature and low MCE at room temperature.
2. Iron has large domains. If my theory is correct then you want materials with the smallest domains possible at room temperature, such as Metglas, which has nano size domains. Note that the smallest domains occur when the material is at or beyond Curie temperature, which is essentially ~0.1 nanometer size domains. So even Metglas is not the ultimate MCE material.
Regard #2, above, here's the reason small domains offer more potential energy. Example, take a thousand PM's (permanent magnets), evenly space them on a flat board, and put them on swivels so they can rotate. Then force all the PM's so they cancel each other out. You will note this requires energy to force the PM's to oppose each other. So if we look at one row of PM's, we would see a PM points south, then the next PM points north, and the next PM points south, etc. You will note this arrangement is equivalent to the smallest domains possible; i.e., each domain is the size of one PM. Now if you release all the PM's and slightly nudge them so they all align to one huge domain (one domain consisting of one thousand PM's) then you have the lowest energy state possible. You will note the PM's will snap into magnetic alignment and will release energy mostly in the form of sound and heat (friction). You have just converted PE (potential energy) to KE (kinetic energy). You went from the highest PE energy state, 1000 single domains, to the lowest PE energy state, one large domain. Note that when you completely saturate a toroid you are essentially creating one huge domain. It is possible to have domains of various sizes in between.
The above example was on a macro scale. IBM as studied magnetic materials on an atomic scale. It is true that the electron and even atom rotate / flip when you reverse the field in magnetic material. More precisely, the atom precesses as it rotates, unless you guide the flips with 90-degree fields, but that's another topic. What is happening as the electron flips is that it radiates electromagnetic waves and some of the energy is lost in friction as the atom rotates and precesses. Although I believe most of the energy is release in the form of radiation, not friction. Unfortunately most of this energy is absorbed by the material.
So in a nutshell, ultimately you want to go from the smallest size domains to the largest size domains. The material has highest magnetic PE when it is at smallest size domains and lowest PE when it is at largest size domains. When the material goes from smallest to largest domains it is converting PE to KE. You want to capture that KE.
Have I figured a way of preventing the magnetic material from absorbing that radiation? Yes, I believe so, as described in the four so-called "secrets" at my peswiki page:
http://peswiki.com/index.php/Directory:MEMM
Regards,
Paul Lowrance
Hellp PaulL,
this link might be of interest : Magnetocaloric effect in dysprosium
Dysprysium is a rare earth element. See wikipedia :
http://en.wikipedia.org/wiki/Dysprosium (http://en.wikipedia.org/wiki/Dysprosium)
http://www.osti.gov/energycitations/product.biblio.jsp?osti_id=6298920 (http://www.osti.gov/energycitations/product.biblio.jsp?osti_id=6298920)
Kator
Hi Kator01,
Thanks for the links. Dysprosium looks nice for certain conditions. Dy used to be ridiculously expensive, but it's dropped like a rock lately-- roughly the price of silver. It's probably a wee too expensive for common folks, but who knows at the rate it's dropping.
MnZn is a relatively cheap and has been used for for its room temperature Curie point. Hopefully in next few days I'll be testing some Amidon cores, which have some MnZn. Two of their best cores have Curie temperatures down to 120 C.
Another factor that improves MCE energy is the typical domain size of the material at room temperature. That why I really need to test a good Metglas core. So far I think it will be one of the best core materials for MEMM. Although Metglas is not that cheap is it?
Regards,
Paul Lowrance
Yttrium ?
Is this another possible substitute ?
Expensive, yet exhibits many rare earth properties.
Thanks for info dean. I found some very interesting core materials at :
http://www.cmi-ferrite.com/Products/Products%20Materials.htm
Some of the materials have first stage curie points as low as -25 C! The following materials seem promising: MN67, N40, MC25, CMD5005
Hopefully they can ship me 1 of each at a reasonable price, and we'll see
Regards,
Paul Lowrance
http://www.ryde.net/theo/mcaloric.html ;)
ok so that on the page of that link they say:
The topmost heatexchanger is the intake of cold fluid (seawater or similar).
doesnt that mean that there has to be a continous cold water flow from somewhere?
can we use the hot water for warming up our homes of something like that??
i dont hava cold seawater tube going in to my house and for the normal water supply would be way too expensife to let it run true the device all the time.
marco.
Hi,
I posted some great news regarding the confirmation of my MCE theory ->
http://www.overunity.com/index.php/topic,1569.msg14013.html#msg14013
Paul Lowrance
Hello Mr Lowrance,
the question is :
if you stimulate the core domains to get the "concentrated spin"-force effect,
under which conditions this force will enter the coil without to destroy it ?
You are in the nuclear(nucleus) sphere !
Sincerely
de Lanca
Hi lancaIV,
Most of the magnetic field in magnetic materials comes from the intrinsic electron spin. So it's not a nuclear spin. The radiation always enters the coils, but because the core absorbs nearly all the radiation it's usually small. Furthermore, most magnetic materials including ferrites radiate hundreds to tens of thousands times less MCE energy than amorphous & nanocrystalline material. The techniques I described will prevent the core from absorbing the MCE radiation.
Paul Lowrance
Hello Mr.Lowrance,
I am not concerned about nuclear/radio-active radiation,
but you describe for the one hand this power-values,
for the other hand a f.e. 10% conversion process lost=to heat transforming
will melt the coil/apparature in a very short time !
S
dL
Hi lancaIV,
It is impossible for the core to heat up with these designs. Actually the problem will be the opposite-- keeping the core from getting too cold.
What is happening with MCE is an energy exchange. There is no energy being created from nothing. Yes, we are converting PE to KE, but the magnetocaloric effect only heats up materials 4 C in Gd and 1 C for various amorphous and nanocrystalline cores. So that is one energy exchange; i.e., the material heats up. In a sine wave current signal there are four energy exchanges ->
1. Core heats up 4 C
2. Core cools down 4 C.
3. Core heats up 4 C
4. Core cools down 4 C.
#1 & 3 magnetic PE is converted to KE. #2 & #4 the KE is converted to PE. Do that 4 times per cycle and 100,000 cycles per second and you have 50 megawatts of energy exchange. So the material will never melt.
In the "free energy" design the goal is to prevent the core from absorbing the radiation in steps #1 & #3. Yes of course if the machine were to successfully absorb all 50 megawatts the coil (not the core) would melt down to center of the Earth, lol. ... I would not recommend anyone try and take 100% of the MCE energy.
There are constant energy exchanges occurring in nature in all matter that are far far far greater than 50 megawatts. My goal is to simply extract 1 KW, not 50 megawatts.
To extract 1 KW of energy from a one cubic inch core would require some really good air or liquid flowing over the core with thermal conducting fins to keep the core from freezing. A 1 KW machine would most likely require a larger core.
Regards,
Paul Lowrance
Hallo Mr.Lowrance,
FR2809241,Tarik and Belaid Yebda
page 2,3
S
dL
Hi lancaIV,
That patent is in French, which I do not speak. Is there a translation?
Paul Lowrance
Hi Paul,
I accidentaly came across a Radus patent on manufacturing magnetic ceramic materials at Westinghouse in the mid 1960s. Several details of the process are given, maybe of interest for you in some respect.
It is US patent US3294687. Freely download at http://www.pat2pdf.org/
Regards
Gyula
PS: You have got my answer to your personal message through this overunity.com a few days ago, haven't you.
Hi Gyula,
I'll look up that patent #. The site you provided did not find that patent.
I could have sworn I sent you a PM, but can't find it in my outbox, lol. I'll reply now.
Thanks,
Paul Lowrance
Hi Paul,
It does work,
Just put the numbers. NO US. 3294687
Hello Mr. Lowrance,
in this publication,not patent awarded(!)=Int.CL: H02K53,
there is explained a similar process to reach the "magnetocaloric" conversion,
with the photons as source.
magneto=electromagnetique caloric=thermique
Google presents some free translation-sites.
S
dL
You'll find it here, but it takes a minute or so to convert.
http://www.freepatentsonline.com/3294687.pdf
Thanks Joe, Paul-R, and LancaIV
I should sometimes slow down and be a little patient, lol. Paul, I tried that pdf file but it crashes my computer, lol. Is this a patent of a "free energy" machine?
The following reply was continued from JackH's thread ->
Hi Rob,
Nearly all common magnetic materials possess extremely small MCE energy exchanges, relatively speaking. Instead of 50 megawatts per 100KHz, 1T, 1 cubic inch as in Gd and 15 megawatts in Finemet your core might exchange just a few hundred watts of power. Note that a ferrite core might exchange 200 watts, but we might be hard pressed to deprive the core of just a fraction of a percent of its energy exchange and that would be using special techniques. Under normal conditions the amount of energy leaked from a core is practically zero. We never know how well a core performs until it is tested-- it might have small domains. Most nanocrystalline materials have small domains on the order of a dozen or more nanometers. As to whether or not amorphous materials have similar or even smaller domains is yet to be tested. The data regarding amorphous materials so far seems to indicate the magnetic moments are randomly oriented (with no applied field), which equates to domains the size of one or two atoms in diameter.
If true, then amorphous materials are the Holy Grail of "Free Energy!" There's probably a wide variance in amorphous materials. For example, two unique materials might appear similar in typical characteristics from permeability to saturation, etc., but their domain size might differ in magnitudes. Still no answer even after asking over a dozen people who specialize in magnetic materials regarding to the domain size range in amorphous materials.
What's very exciting is learning about Bill Muller's research, which he claimed that amorphous materials are the secret to "free energy." Has anyone any knowledge or experience in making Muller's amorphous material? IMHO this could be vitally important research. Metglas is very expensive and just a single source. It would be great if we could construct a simple process of making Bill Muller's amorphous material.
Regards,
Paul Lowrance
I've added a new section to the MEMM wiki project, titled "Why Domain Size is a factor in Potential 'Free Energy'"
http://peswiki.com/index.php/Directory:MEMM
Notice the images of domains.
Regards,
Paul Lowrance
The following was just added to http://peswiki.com/index.php/Directory:MEMM
Note, there would be no applied field to the above images. There are several factors that determine domain size. First is temperature. Second is magnetic strength from ferromagnetic atom to ferromagnetic atom. Third factor that affects domain size with no applied field is grain or crystal size. The grain walls make it more difficult for the domains to extend beyond. So basically it's a battle between temperature & crystal size versus magnetic strength. An increase in temperature makes the domains smaller. An increase in ferromagnetic affective density increases domain size. Decrease in crystal size can decrease the domain size.
Image A is one domain. In order to achieve this for an appreciable amount of material, say 1 cubic inch, you have to lower the temperature to near absolute zero. In such a case the magnetic moments would all flip in alignment because there would no longer be the vibrating atoms to prevent such an alignment.
Now consider magnetic material in Image D. We have small domains at no applied field. If we then apply a saturating field we end up with one large domain (Image A). Image D has high potential energy. Image A has zero potential energy. Therefore, that energy must go somewhere. When the electrons flip they give off radiation. There are techniques to lessen the magnetic materials ability to absorb the radiation.
In a nutshell, there is more potential energy in Image D than Image C. It requires energy to make Image C go to Image D.
Lets say we saturate magnetic material. Now our material is one large domain, Image A. So you might ask how does the magnetic material get back to Image D. When we remove our applied magnetic field it is the ambient temperature (vibrating atoms) that knocks / forces the magnetic moments to reverse. Note that this requires energy because the magnetic moment is in alignment with the net magnetic field. So it is ambient temperature that forces the magnetic moments (intrinsic electron spins) to unaligned. If it were not for ambient temperature then it would require energy from our coil to cause the magnetic moments to become unaligned. That is the reason magnetic materials near absolute zero have square loop hysteresis, high coercivity.
What is happening when the material changes from Image D to Image A when we apply a saturating field is the magnetic moments are snapping in alignment, thus giving off energy. This heats up the magnetic material as it absorbs the radiation. When the applied field is removed the vibrating atoms knock the magnetic moments out of alignment, which slows down the atoms as it requires energy. This cools down the magnetic material. Although, what if we robbed the magnetic material of its radiation. This means the material would not heat up, but it would cool down, meaning we gain energy. This gain energy is in the form of electricity.
As previously described, the amount of radiation in some materials is in the megawatts for one cubic inch of material with a 100 KHz signal that changes the net internal field up to 1 T. In the section below titled, "Relevant Post" we see an example of amorphous and nanocrystalline material, Finemet, that possess 15 mega joules of energy exchages per second, 15 megawatts. Such material requires but a fraction of a watt to generate such a net internal field within the core. This fraction of a watt is a catalyst for 15 megawatts!
Hi Paul,
Thanks for all the additional information.
I will digest this over the next few days.
Did you have any joy with getting samples for the Metglas C cores?
I have emailed them again requesting samples and filled in the form on their website so hopefully they will respond this time.
In the meantime I will build the pulse circuit with some minor updates to use better components.
Regards
Rob
Hi kingrs,
No, I haven't received metglas core yet. :-(
Is there any chance you could get the exact same parts that Naudin used? If not then perhaps there are higher performance parts. Perhaps a MOSFET with higher breakdown voltage and faster switching speed. Although if the exact same parts are not used then that further changes the capacitance, etc. of the circuit.
I hope you get the "smoking gun" before me. Are you a perfectionist like myself? I'm still messing with the testing process of various types of cores. A company sent me an interesting core that has a Curie Temperature of 43 C, permeability of 5500 and saturation of 1900 G. It was ordered for its extremely low Curie temperature. According to my MCE theory the magnetic moment boding strength is one of the major factors that determines domain size. So far this material is giving a lot of trouble. The temperatures can swing huge amounts one moment, but then hardly any the next. It might be for the reason that this core is very small and is actually touching the rod that goes down the center of the core. I use a thick aluminum rod as a single turn to decrease wire heat.
Regards,
Paul Lowrance
Hi mramos,
Naudin uses the BUZ11 MOSFET ->
Features:
Nanosecond switching speed
30A, 50V
rds = 0.04 ohms
High input impedance
Input capacitance is 1500 pF typical
Output capacitance is 750 pF typical
www.ortodoxism.ro/datasheets/fairchild/BUZ11.pdf
Looks nice. Fast switching speed, but somewhat low breakdown voltage.
I see digikey sells them for $0.84 in quantity of one:
http://www.digikey.com
just search for BUZ11
Paul Lowrance
Hi mramos,
20 cents each? What a deal! Do you have any specs on them?
Paul Lowrance
It's always a question just how much time should be spent on documenting versus just working on the experiments, building, and research. Lately due to my friends 20 year old prophesy I've spent a little more time on documenting. :-) Here are more "free energy" details -->
First, a few prerequisites and definitions.
PM - short for Permanent Magnet.
Magnetic materials - Most magnetic materials are either ferromagnetic or ferrimagnetic. They both generate magnetic fields, but ferromagnetic is stronger than ferrimagnetic. Ferrites are made with ferrimagnetic material. Pure iron, cobalt, nickel, etc. are ferromagnetic.
Electron orbital - The electrons are not particles, but really wave-particles. Even so, a lot of electrons do indeed have an equivalent orbital motion around the atoms nucleus. Simply stated, some electrons orbit the atom. Basically you can imagine this electron orbital as a coil of current.
Intrinsic electron spin - I'll abbreviate this as IES. If we zoom in a look at the electron we'll note there is an equivalent vortex of current. Basically speaking you can imagine the electron as a small coil with current. More precisely this imaginary current is spread out like a vortex. Essentially, IES is similar to the electron orbital except the IES is far smaller and more intense.
Magnetic field caused by all ferromagnetic or ferrimagnetic materials - The magnetic field caused by these materials mostly come from the IES, not electron orbital. I've read values of 80% IES.
Magnetic moment - This is a field caused by either IES or the electron orbital. If you have seen drawings of the Earths magnetic field then you know what the magnetic moment field looks like. See the attached image on this post.
MCE - This is the Magnetocaloric effect.
Eddy current - Please see the following web page -> http://en.wikipedia.org/wiki/Eddy_current
Electron flip - This is as described, the electron rotating 180 degrees and flipping. A great deal QM (Quantum Mechanic) physicists are under the impression the single electron does not rotate, but simply flips in an instant, in zero seconds. This is a false interpretation of QM. Experiments conducted by companies such as IBM have shown that the electron not only forces the entire atom to rotate, but it also forces the atom to precess as it flips / rotates. The actual electrons flip rate has been measured and it's typically a few nanoseconds, but can be significantly slower in electrically conductive magnetic materials.
Avalanches - This is an effect where a great deal of electrons flip. It is an avalanche effect where one electron will trigger another and so on until the avalanche dies out.
Applied field - This is simply a magnetic field that is applied to the magnetic material. This applied field can come from current in a coil or from PM's.
Magnetic energy - this is in reference to the energy associated with electron flips.
There are basically two main methods of extracting MCE energy. -->
Method #1 --- Using the Eddy currents as a tool
This is the method Naudin used in both of his designs. This method will not work on ferrite cores, as it requires the magnetic material to be electrically conductive at least on the micro scale. This is the easiest method.
Lets start from the beginning and with a very simple design. For simplicity lets use a design that does not have any PM's (Permanent Magnets) because PM designs introduce more complexity. We have a core with two coils-- coil #1 and #2. This design therefore requires a certain minimum amount of current running through the coil to make up for the lack of PM. Note that coil #2 is only for collecting energy. Our core is a toroid.
So current is flowing through the coil #1. The net magnetic field within the core is at level A. Now we want to increase coil #1's current as rapidly as possible. So coil #1 has increasing current and coil #2 is completely off. What happens is the IES's (Intrinsic Electron Spins) flip in avalanches. These avalanches are very slow because our magnetic core is electrically conductive. So there are avalanches igniting here and there. These avalanches cause Eddy currents, since our magnetic material is electrically conductive. So basically a great deal of the energy associated with the IES flip is given to the Eddy current. We see within magnetic material there's a storm brewing as the applied field increases. As the applied field increases there are millions of nano size avalanches and Eddy currents. The avalanches generate energy, which Eddy currents collect. The Eddy currents have an RL decay period, once they reach peak, meaning the Eddy currents decay at a changing rate, simply stated.
At this moment our applied field is increasing, there are avalanches and Eddy currents. At the precise moment, and time is crucial, our coil #1 suddenly turns off and coil #2 turns on. A lot of electrons are still flipping and we already have a lot of energy built up in Eddy currents. We now have no current through coil #1. For simplicity coil #2 is connect to a resistor. So the resistor across coil #2 collects energy, which it dissipates in the form of heat. At some point the Eddy currents in totality will reach maximum and begin to fall. It is the job of coil #2 and its load (the resistor) to rob as much of this Eddy energy as possible.
Eventually the net magnetic field in the core will fall back to level A, as mentioned above, and the process repeats.
Method #2 --- The High Speed method
I'll document this method at a later time. Essentially this method requires non-electrical magnetic core such as ferrites. This method could possibly generate more power, but it requires extraordinarily high performing parts that can switch in roughly a nanosecond while allowing either high current or have high breakdown voltages. As in method #1, the core is always partially magnetized.
This method does not rely on the micro eddy currents. Rather, at high speed the coil current must increase (switch completely on) faster than a fraction of one flip speed. Since the core is non-electrically conductive the electron flips will occur at high speed, typically in a few nanoseconds. It's the job of the coil to generate one coherent simultaneous avalanche pulse. When the electron flip process has reached a certain rotation (roughly 90 degrees rotation) then it is time to collect the energy. Remember, just as in method #1, the core starts at level A net magnetic field. So the core is partially magnetized from the start. It is this strong net magnetic field that provides so much energy when the electrons flip. The magnetic field caused by the coil is but a fraction of the field caused by the magnetic material. That is why one cubic inch of Metglas oscillating at 100 KHz generates 15 mega joules of energy exchanges in one second (15 megawatts) per Tesla.
Note that the effective permeability in method #1 would be relatively low (~5 to 100) as compared to method #2.
Kind regards,
Paul Lowrance
Hi Paul and All,
I have just seen a new price info on AMCC-320 Powerlite core at the /MEG_builders/ group from Andrew:
..."I got in contact with the company that supplies the core as specified
on the JLN Labs site and they said the core is $107 as a set.
Their website is http://www.elnamagnetics.com and it is a MetGlas
AMCC-320 Powerlite core." ...
I did not manage to visit the elnamagnetics.com page, it is down, probably for the weekend.
Another issue I would like to comment is the type of power MOSFETs you wish to use for switching, especially in case of Method #2. Very fast switching-on will require a MOSFET able to survive several hundred (at least 600-700) Volts of back EMF from the coil and MOSFETs with such a high drain-source breakdown voltage are not cheap, especially with the simultanious need of the low rds.
One of the best firms on this field I know of is IXYS Corporation ( http://www.ixys.com/ ) and this link is their product family on MOSFETs: http://www.ixys.com/pfdmos01.html and Q2 Class HiperFET MOSFETs with Exceptionally Fast Switching category is a possible choice if a 25-30nsec on / off switching speed is the goal in PRACTICE. This performance needs appropiate gate drivers too, see http://www.ixysgatedrivers.com/
(Let me notice I have no any relation/connection with IXYS Corp!, surely there are some others on this field.)
Regards
Gyula
Hi Gyula,
I also tried the elnamagnetics.com website the other day and it worked, but now it appears to be down.
The google cached page still works but is of little use.
See this for MOSFET advice:
http://www.richieburnett.co.uk/mosfail.html
Its interesting about the gate current at high switching speeds being as high as an amp.
So the page recommends putting in your own back emf protection diodes (2 of them).
The bit about drive voltage is odd, I thought a mosfet gate voltage could not exceed 5volts let alone 10 - 15 volts.
Regards
Rob
mramos
100 KHz would a fine place to start. Although that frequency was used as an example to explain what is happening within the magnetic materials on the atomic scale.
Gyula
That's great news! Although their site being down is a prime example why a single source should not be relied upon when the "smoking gun" is released. The "smoking gun" will be easy to make and relatively inexpensive and highly effective in generating electricity, "free energy. " Can you imagine the entire world trying to order a core from a single company, lol???
Method #2 requires extraordinarily high performance parts. Therefore I think Method #1 is the best starting point. BTW, the MEG uses Method #1. Note that Method #1 as described in the wiki is simplified to describe what is happening within the magnetic materials. As stated, you can use a PM in Method #1.
Rob,
I'm curious, will you be using BUZ11 MOSFET's in your MEG replication?
Regards,
Paul Lowrance
Hi Paul,
I have 3 x 2SK3594-01 MOSFETs rated at 200V and 30 Amps,on resistance 0.05R
Very good switching time. On + Rise time is 37ns and Off + fall is 70ns
I also have some 25 x IRFU3707 in I-PAK format.
These are rated at 30v and 60 Amps, on resistance 0.013R
On delay is 8.5ns and the rise time is 78ns, off delay 12ns and fall time 3.3ns.
I have picked these two for their switching speeds and on resistance.
The first set of MOSFETs I bought were to go with a 500Watt 0-55 0-55 v mains toroidal transformer for a variable output PSU, but should work well in the MEG circuit. I am building a dual output 0 to 55V 4.5Amp PSU.
This is to charge the Joe Cell, I am also working on, to stage 3.
It needs about 100volts upwards to get a decent current to flow through the water.
Regards
Rob
Rob,
I can't see why those mosfests would not work. They look good. I don't think the differences in capacitance will make noticeable difference.
Gyula,
You mentioned your replication of the MEG using ferrite core. According to MCE theory the MEG design will not work on most ferrite cores because most ferrites are not electrically conductive. The MEG uses Method #1.
Naudin made two designs, iron and metglas, which are both electrically conductive. IMHO it would be extremely difficult to get "free energy" with an iron core, but possible. I found errors in Naudin's scope interpretation. At least the scope pictures I analyzed show his iron core version did not produce more energy out than in, but it is very clear that according to his scope pictures his metglas version produced more out than in. IMHO there is no way to discount Naudin's scope pictures of the metglas version, unless Naudin outright falsified the pictures. I've discussed this with other people in private regarding ultra high frequencies tricking the scope, the input source, etc. It's just not true and if anything Naudin's metglas produces more output than we can tell from his scope pictures. Of course, this all presumes Naudin's scope is not faulty. :)
Regards,
Paul Lowrance
Hi Paul, MrAmos,
I have ordered a AMCC-320 core from Elna Magnetics and should be with me by Thursday provided there is no hold up in customs.
I now need to look at building a pulse circuit.
Not sure wether to look at controlling it with a PIC or try the circuit that JLN used.
Ahh...I have just looked at the spec for the TL494C.
http://focus.ti.com/lit/ds/symlink/tl494.pdf
250mA output drive current, much better than the PIC.
Also the output voltage is determined by the supply.
1 to 300Khz frequency range.
Built in oscillator stabiliser.
Yep, it has the the lot.
Its tried and tested by JLN, so why re-invent the wheel.
I think I will invest in a 30v 5amp bench power supply too, more toys to play with ;o)
Regards
Rob
Hi,
That's really great news kingrs! How much was the AMCC-320 core?
Hopefully by tomorrow I'll be testing a silicon iron transformer under exceptionally small power conditions. Unless I'm incorrectly simulating this in my mind I'm guesstimating that a few microwatts should reach the level where even a standard silicon iron core could reach the "free energy" state. I know, whoopee, a whole microwatt, lol. :) There might be a whole lot of other forces involved at such microwatts, but we'll see.
Unless I made some errors the theory shows how efficiency is relative to the reciprocal of power. In other words, less power equates to more efficiency. Of course that is not considering the inefficiencies of the circuit and presumes the magnetizing field is always at optimum.
This seems to match Naudin's results, but it is not so evident -->
http://jnaudin.free.fr/images/meg21iof.gif
Notice the output (red line) exponentially increases up to 25 volts, but then drastically decreases. Then look at the input as it continuously increases with its exponential rate. Understandably there are a lot of factors involved, but we cannot ignore the fact that Naudin's power chart shows the input power approaching the output if we follow the pattern. I did not enter these graphs in a spline function, but I would guesstimate the two merge at roughly 75 to 100 volts input.
I know everyone wants to pump a ton of power in their core so they can get a lot out, but please consider lowering the input voltages as low as your circuit will efficiently tolerate. Another way of achieving less input power is to increase the amount of turns on the input coil, up to certain point of course depending on frequency-- you probably don't want significant reflecting waves.
Similarly, 10 small cores should be more effective than 1 large core.
In a nutshell, as you double to input voltages you are essentially doubling the coils current, unless your circuit is not linear. That quadruples the input power. Although, by doubling the current your are doubling di/dt, which simply stated will merely double the MCE energy. So you are quadrupling the input power, but you are only doubling the available MCE energy.
Regards,
Paul Lowrance
Hi Paul,
From Elna Magnetics the AMCC-320 core costs $107 + $64 shipping to UK, plus $3 handling.
The shipping is FedEx, but I guess if I ordered a few cores the cost per core would be a lot less. Shipping within the US will be a lot less.
The AMCC-500 core costs $140.
I did not ask about the AMCC-1000 (7Kg lump) but if the AMCC-320 works out, that will be my next order.
MrAmos, 1 hour to design and etch a PCB, now that is fast.
I have all the gear to design, UV process, etch and drill but it will take a couple of evenings to do one design.
Are you going to get an AMCC-320 core?
I will knock up a design on Eagle (free to use on small 4" x 4" or less boards) if anyone is interested.
Regards
Rob
I made my UV box from an old flatbed scanner that I could not longer use because the scsi card was ISA.
Put two UV tubes into the bottom, wired this to a guts out of a new 12v caravan light.
I use pre-coated laminates from MegaUK (http://www.megauk.com)
I time it on a stop watch and manually turn it off.
The tubes need to warm up a bit, so manual control is fine for this.
I have etched double sided panel PCB antennas using this and they are perfect.
Do you iron the toner onto the board from a transfer sheet?
My mate does this, but I prefer the UV method.
Where abouts are you based?
Regards
Rob
Hi,
I read the messages in this forum and then tried to reply, but everything went dead. Found out that my system suddenly got 9 viruses, lol. Anyhow, everything's reinstalled and back online, finally. :-)
You guys are beyond me in electronics. For now I just use a breadboard, sort of like plug and pray. Looking forward to seeing kingrs results. I'm still designing the circuit for the 1 uW iron transformer. This is turning out to be real fun.
mramos,
I understand, getting a car for your oldest has high priority. If you have some silicon iron U-cores around then perhaps you could replicate Naudin's iron version. I'm predicting that _if_ the input power could be lowered to a few microwatts that you could create a self-running machine. Efficiently working with a few microwatts could be tricky. I believe you can actually see a lit microwatt LED in complete darkness. Just need to allow your eyes to adjust. Hey, it will be the first freely published self-running machine, lol. From that point there's no place to go but up and next thing you know you'll be in the kilowatts. :)
Regards,
Paul Lowrance
Hi Paul,
I was wondering why you were quite.
Status:
I have ordered up a 30v 2.5A variable bench power supply to feed the pulse circuit.
I plan to drive the pulse circuit from 12V and the MOSFETS from the variable supply up to 25V.
This way I can prevent damage to the gate of the MOSFET and keep the chip consumption out of the power input equation, albiet, a small power draw.
My AMCC-320 core has been dispatched now.
I need to order some TL494CN chips and some heavy duty resistors, I think I have most of the other components.
Started on the PCB design for the circuit using Eagle, not sure about the placement of the MOSFETS though.
Hi MrAmos,
I used to have a panasonic KXP4420 laser, a few years back, cost me a small fortune 600GBP ($1000).
Now I have a cheap Samsung ML1510.
Works very well.
Regards
Rob
kingrs,
That's much better than the power supply I yanked from an old PC. It gives 30A @ 5V and 15A at 12V. It's not very good for noise and stability. They must use cheap regulators or something. When I need to keep the noise down I use batteries and various size caps across the source.
Paul Lowrance
Hi,
Someone from the MEG_Builder group asked me some questions. I thought it might help people to better understand the MCE process of collecting electrical energy. -->
-----------
There are several methods. Method #1 is the easiest. Normally MCE (magnetocaloric effect) heats up, cools down, etc. In electrical conductors such as iron and Metglas a lot of the MCE energy goes to micro eddy current bursts. Normally the eddy currents dissipate all the energy in the form of heat. If you pulse the core at the correct speed you will get a _coherent_ avalanche pulse. IOW, the avalanches are occurring at roughly the same time. You'll get eddy currents. When the Eddy currents reach peak then your receiving coil will attempt to rob as much energy from the Eddy currents. You do this by placing a load across the coil.
Picture a nano size group of atoms that flip. There are many factors that determine the flip rate such as magnetic field strength, but free electrons plays a huge role. The free electrons act as inductance, resist the flipping magnetic moments. (You can see this effect by dropping a neo magnet down a hollow Al tube.) This gives a micro eddy burst. So you could say its like a microscopic coil around the avalanche, which is a good thing so as to collect a high percentage of the MCE energy.
Under normal conditions you have millions of micro eddy currents that are simultaneously increasing and decreasing all over the place within the core. In other words, the bursts are not coherent. Micro eddy bursts do not last very long, which is why you need to pulse the core fast enough and then quickly absorb some energy from the eddy currents. Although, when the eddy currents occur at the same time then the bursts decay at a much slower rate, which is a good thing.
Where the energy comes from is fascinating. Without ambient temperature (vibrating atoms) magnetic material would align (saturate) and that's the end of the story. Even when you remove the applied field the core would remain magnetized. It is vibrating atoms that give low coercivity. So when you remove the applied field it is the atoms that _force_ the magnetic moments to break alignment with the net magnetic field. That requires energy, which is exactly why magnetic materials cool down when the applied field is removed. That is where MCE energy comes from. Even the NASA guy who contacted me agreed.
Trying to compute the energy relative to the field strength is perhaps not the correct method. Consider two PM's each on swivels, so they can rotate. The PM's are rotated so they repel each other. The magnetic fields cancel each other, so the net magnetic field is relatively low, just within close proximity of each PM. Now allow the PM's to quickly rotate so they align. You get energy _plus_ you get a net magnetic field, lol. Magnetic moments also rotate as IBM's experiments revealed. Normally this flip/rotation rate takes a few nanoseconds, but in electrically conductive materials such as iron and metglas it takes many microseconds.
-----------
Regards,
Paul Lowrance
Hi mramos,
I know what you mean, as I am fairly certain I saw your post. It seems to be gone.
I'm in Los Angeles County, CA., USA. I am still waiting for Metglas to ship me a small 2714AF core, but still no shipment. For now I am working on a cheap silicon iron core, ~half inch in diameter.
If you would like, you could wait to see kingrs results. Although on a personal level I am very confident that kingrs will succeed, but what if he does not? I don't see any reason for more than one person replicating the same experiment unless such a person can easily afford the costs. Hopefully kingrs will receive a _real_ metglas core.
The reason I am experimenting with silicon iron core, beside the fact my Metglas did not arrive as scheduled Oct. 17th, is that I am testing a theory that less power equates to higher efficiencies. As admitted this particular theory of power-vs.-efficiency is not set in stone, as it's merely a somewhat complex simulation done in my mind. It's a little complex so I could have easily made a mistake. Better to allow the computer to accurately reveal the results or just do the experiment, lol.
One thing that does not entirely make sense is Naudin's efficiency dropped like a rock toward 10 volt input. Of course his chip has a _minimum_ source voltage spec of 7 volts. I am wondering if Naudin fine-tuned his circuit for 10 volts to see the maximum efficiency. It sure would be nice to converse with Naudin!!!
Regards,
Paul Lowrance
mramos,
I forgot to post the link. It seems kingrs bought his metglas core through :
http://www.elnamagnetics.com
Paul Lowrance
Hi Paul and MrAmos,
My core is due to arrive on the 23rd of October (next Monday).
It is the Metglas AMCC-320 and costs $107 + shipping + $3 handling.
If you are feeling flush then you could opt for the AMCC-500 at $140.
They only like to use couriers as the US postal service was not reliable enough.
The 30V 2.5A variable PSU is due to arrive Thur/Fri this week.
I still need to order the TL494CN chips, and I have just discovered tonight that I only have double sided pre-coated laminates (PCB) left so I need to order up some single sided and some tech sheets for the UV work.
The schematic is nearly complete, just need to do the PCB board layout next.
Its all go at the moment, I'm quite excited about seeing this core in the flesh.
Paul, I think you are right about the input power being just right, according to my FEMM 4.0 simulations, the power needed to get the field to switch was tiny, maybe 10mW on a small core.
5W input on the AMCC-320 I would say was overkill, and should be in the region of say 100mW.
But that is with very small load on the output coils (half the input power I think), the simulation can only go so far with this experiment and I think a practical experiment may see the power input go to 1W.
I have to say I hate this site too, it runs the CPU flat out while the page loads, some crappy javascript running in background, pages sometimes fail, not a nice site...ahh well.
Regards
Rob
Meg simulation with 10mA, 0.137 V (1.3mW) in the bottom right input coil.
US Steel Type 2 core and N37 Neodimium magnets.
Core is 120mm wide and 100mm high and 50mm deep.
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2Fmegv1-2.jpg&hash=2fc17fc91ae05a53dfd4a51b70fdc17b0f2cf14b)
Hello Rob,
this 107$ for the Metglas AMCC-320,
could you give me the weight and material volume,please !
For the Mukherjee motor-gen invention(DE...) I would need such material,
but in granular/powder kind !
Sincerely
de Lanca
Hi mramos,
I might get a free metglass sample. It was supposed to arrive yeserday, but the arrival date changed to Oct. 20th. Not sure if they're playing with me. We'll see. You might be able to get a free sample. Ask for any size sample they have available for testing purposes. If you are not picky and if they have a spare around they might mail it to you.
I'm getting a small 2714AF. As far as efficiency is concerned I think small is better.
Regards,
Paul Lowrance
@All
I think the following may help those interested in the MEG, related devices, and further research. I know, I know ... spending too much time on documenting, :) but this is my only hope of defense if I were to suddenly pass on; i.e., if that happened then hopefully the disinformationist would not succeed in convincing people MCE energy is not real or worth much. Please never give up on the MEG. The pulse timing is vitally important! The amount of load resistance relative to number of secondary turns is vitally important. It would help if the load resistor appropriately changed during the received pulsed, hence it might be worth duplicating Naudin's conditioned resistors. Just make sure you use a common small R (as Naudin did) as a way of measuring the current. IMHO the input power to core is important. That is a lot of combinations. If you change the R 40 times, adjust for 40 different frequencies, and then 40 different input voltages that's 40 * 40 * 40 = 64000 combinations. Please do not give up on the MEG. Finding the exact correct situation for your particular setup could be akin to finding a gold mine, but it will be worth it. Experiment.
First, we know avalanche radiation exists. I prefer to call it "avalanche radiation" over Barkhausen because it's more descriptive and there's a little controversy about Barkhausen, but that's another subject. Point being, we know about the avalanches. We know it is typically UHF radiation for non-electrical cores and considerably lower for conductive cores such as iron. The amount of such radiation that escapes the core is very small. The reason it is small is because the avalanche occurs completely incased within the core and we know the fields have a closed loop, a magnetic short if you will. This easily demonstrated with FEMM. Also we may study induction simulations to learn that core radiation leakage is relative to the materials permeability.
Now to the point. At any given time while we are pulsing a core there are X amount of avalanches occurring that are unstoppable; i.e., if we remove the applied field the avalanches would complete. I refer to this as "Magnetic Momentum" (that's momentum, not moment), and not to be confused with Magnetic Viscosity.
The amount of magnetic momentum varies with material. There are a lot of factors, but the main factors are the materials MCE and its electrical conductivity. I predict that nanocrystalline materials such as Metglas and Finemet have high magnetic momentum.
---
I would like to differentiate the different between MCE energy and common induction. Envision thousands of tiny PM's (permanent magnets) on swivels that forms one big toroid. There is wire that wraps this big toroid to form a standard toroid coil. Basically we have formed a large scale magnetic toroid core with a coil. These tiny magnets are all aligned to form a closed loop-- essentially our core is saturated. Now at a constant rate randomly force say 100 PM's per second to flip. This will induce a net constant voltage. We know that the net constant voltage is not dependant on how fast _each_ PM flips. Rather the net constant voltage depends on _how many_ PM's _per second_ are flipped. So, the induction is relative to how many flipped PM's per second and MCE energy is relative to how fast a PM is flipped.
In other words, if each PM is flipped in 1 ms rather than 10 ms the net constant induced voltage will not change, but there will be more radiation energy. MCE is that radiation energy.
Note that each time a PM is flipped we'll see a dc pulse (a dc spike) in voltage. If the PM flips 1000 times fast, then the _net average_ voltage does not change; i.e., the voltage is 1000 times greater, but the pulse width is 1000 times shorter. So it flips 1000 times faster. The voltage will be 1000 times greater. If the voltage is 1000 times greater then power is 1000000 times greater-- P=V^2/R. Therefore, power is 1000000 times greater, the time is 1000 times less, the resulting energy is 1000 times greater. Energy = time * Voltage^2 / Resistance. If we were to look at this signal on a spectrum we would see that by increasing the flip 1000 times faster results in higher frequencies. If you flipped it fast enough you would have a high-energy gamma photon, and you better duck. ;-) E=hf
Regards,
Paul Lowrance
Hi Paul,
I don't quite understand where the excess energy is coming from.
Is it the change in temperature of the material (MCE) during the input pulse that alters (increases) the stored magnetic energy for when the pulse is removed and picked up by the output coil?
PSU has arrived, core due on Monday, schematic finished, doing track layout now, parts to order over the weekend.
It can output 32v at 2.9A so more than enough for the MEG:
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4724.jpg&hash=dcd0dbdbcbd70a13f15276f54f570ad01c7d3967)
Hi kingrs,
Quote from: kingrsHi Paul,
I don't quite understand where the excess energy is coming from.
It comes from ambient temperature. Have you ever tried to pull apart two PM's? It takes energy just as it takes energy for the vibrating atoms to break the alignment of the electron magnetic moments.
It's really simple. When the magnetic moments align we get energy-- MCE radiation. That's basic physics. If your pulse is fast enough then all the avalanches (MCE energy) is a coherent pulse. The avalanches give energy to eddy currents-- magnetic momentum. Note that eddy currents is not required for magnetic momentum. Magnetic momentum consists of the avalanches that have reached a stage where it cannot be stopped. At the correct timing you need your resistor load to gain as much energy as possible from that magnetic momentum. If you read my previous post you'll see the difference between induction and MCE energy.
Quote from: kingrsPSU has arrived, core due on Monday, schematic finished, doing track layout now, parts to order over the weekend.
It can output 32v at 2.9A so more than enough for the MEG
That's great. Remember though, more does not always equal better. Try a wide range of power input levels. See my previous post and you'll see it could take some time, roughly 64000 different variations.
Are you going to replicate Naudin's conditioned load resistors?
Regards,
Paul Lowrance
Hi Paul,
Thank you for the explanation, its nice to see there is a new refridgeration technology emergy using MCE.
Looking at JLN's scope shots he appears to be using 24KHz as the input signal, yet if you check his circuit diagram with spec. sheet for the TL494CN, his pulse circuit should start at around 40KHz and end at 100KHz.
Therefore I suspect the timing capacitor C1 may be a larger value like 5nF.
It would make sense to put a rotary switch in here to allow a range of values to enable a large number of frequencies to be covered, from say 1KHz to 100KHz.
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2Fmeg_pulse.jpg&hash=35fa1de8db4487c724c5855064b124cd9b82a550)
Hi kingrs,
Quote from: kingrsIt would make sense to put a rotary switch in here to allow a range of values to enable a large number of frequencies to be covered, from say 1KHz to 100KHz.
That would be a smart thing to do. Truthfully, I think Naudin was all over the place, spent a great deal of time on his MEG, varying all the parameters. I wish he posted scope shots of all variations of input voltages, frequencies, etc.
Do you plan on documenting and taking lots of scope shots? Also I think it's a good idea to eventually try and replicate Naudin's conditioned 100K load resister.
Regards,
Paul Lowrance
There's something about the MEG shots that I don't understand. I don't understand why the actuator current resembles a sine wave if the voltage is a square wave. Shouldn't it be either a saw tooth or an exponentially increasing saw tooth wave? If the core begins to saturate then the permeability decreases, which means di/dt should increase.
The only way it could be like a sine wave is if the resistance is significant. You know, an RL time constant. If the resistance is appreciable then initially the current will increase and eventually will level out due to resistance. You know what I mean?
Why would there be any appreciable actuator resistance? He shows the coil as 1.6 ohms and if the MOSFET is properly turned on it should be 0.04 ohms.
Regards,
Paul Lowrance
Hi Paul,
I am not sure about the replicating the output load resistor, a carbon resistor should be the same what ever you do to it.
I will just use a large wattage carbon resistor or series of them.
I can see there being a possible issue with a wirewound resistor.
Also JLN's scope shots include the power for the pulse drive circuit so to show that the output efficiency as being better at the higher voltage may be a result of the driver circuit using a fixed amount and this fixed amount becomes less significant to the overall power consumption as the voltage input increases.
I think therefore the driver circuit needs to be run off a seperate supply which is what I shall do.
I do not think the coil is a 1.6 Ohm resistor, according to the circuit diagram, this is actually a resistor in series with the coil to reduce the current flow through the coil. I do not understand why this is required and will in effect cause a power loss in the pulse circuit in the form of heat. I can only think this is there to allow a larger input voltage.
I will take this off the board and therefore it can be added externally as an option.
From all my simulations the drive voltage for the pulse needs to less than a volt, therefore by seperating the drive supply and using my variable output PSU I can get a sub 1Volt pulse into the circuit.
The sine wave shaped drive current is probably a result of the resonance built up in the core.
I think when you first apply power to a transformer, there is a huge current surge until it starts to stablise, probably lasting a couple of cycles.
The coil's capacitance and core saturation can also shape the current.
I know very little about this topic.
Have a read of this:
http://www.butlerwinding.com/elelectronic-transformer/pulse_transformer/index.html
Regards
Rob
Quote from: kingrsI do not think the coil is a 1.6 Ohm resistor, according to the circuit diagram, this is actually a resistor in series with the coil to reduce the current flow through the coil.
OK, I just thought 1.6 R was just depicting the coils internal resistance.
I'd like to build this circuit in LTSpice and see what happens? Any estimates what the internal actuator coils capacitance is?
Regards,
Paul Lowrance
I just completed an LTSpice circuit simulation of Naudin's MEGv2.1 and found some very interesting results. I used v2.1 because Naudin shows the actuator currents, which is important. Below are three images of the circuit, the actuator current, and the output.
I found out that the unusual actuator current is due to mutual coupling and not capacitance. Actually capacitance has the opposite effect in that it causes the current to initially surge.
Anyhow, after doing some research on various size inductors and how series & parallel resistance and parallel capacitance changes between inductors I came up with some good estimates on Naudin's MEG. Given wire size and so on I calculated 1.4 ohms for his actuator coil. Naudin has 1.6 ohms. So it's safe to say the extra 1.6 ohms simply represents actuator resistance. Also the 37 ohms is the secondary resistance.
Amazingly I was able to find an actual BUZ11 spice library, so this is very accurate simulation as far as parts.
Initial simulations were showing completely different actuator currents than Naudin's. After getting into the details discovered the secret was mutual coupling coefficients. So I entered appropriate coefficients according to Naudin's MEG core. When they were entered you can see the results are almost exactly to Naudin's MEG with one major difference, the secondary output voltage versus actuator current. :) According to simulations, I had to bring the input voltage way down to get 280 mA actuator current, BUT Naudin's output voltage is 3.86 times greater! According the MCE theory this extra output voltage is caused by what I call Magnetic Momentum. What is very interesting is Naudin's scope shots shows the output as 28.98 W and the input as 2.373 W. That is 12.21 times greater. We know that power is relative to the square of the voltage. If we square 3.86 we get 14.9. That's very close to Naudin's 12.21 ratio.
I am not saying the above is proof that Naudin was telling the truth and that his scope shots are not faked, but I spent a good amount of time with LTSpice and the MEGv21 circuit, playing with stray capacitance and all. So far it is very interesting and worth investigating.
Attached is a zip file that contains the LTSpice file of the Megv21 and the BUZ11 spice library for anyone who wants to play with this. LTSpice is one of the best simulators. Best of all, it's free, created by Linear Technology, one of the best high end ADC manufacturers in the industry. Note that you will need the place the BUZ11 file in [your hard drive]:\Program Files\LTC\SwCADIII\lib\sub\
There are 3 attached images, which I think you need to click on to enlarge. One of the circuit. Another of the actuator current --> I(L2) purple graph. Another of the secondary output --> Orange is power, purple is voltage, turquoise is current.
Regards,
Paul Lowrance
Hi Paul,
I do not know how to calculate the capacitance of a coil, there are so many variables that I guess you would need to wind it then test it.
I downloaded the LT Spice and ran your simulation.
What an amazing piece of software this is!
I added an extra 1.6Ohm resistor to the circuit in series with the input coils and it seems to make little difference to the current, which seems odd.
You can see the back emf voltage in the input coil jump to 70V for a very brief period, then it bounces slightly at around 28V until the next pulse.
This means the mosfet would need to be able to cope with around 80 volts across source and drain.
How did you couple L1 and L2 to L3 and L4?
I have looked at the spec on the TL494CN chip again because I was worried that it would not be possible to have a dead time between pulses, so I started at looking at other function generator chips today to see if there was an alternative and the closest I came up with was the Exar XR-2206 chip:
http://www.exar.com/products/XR2206v103.pdf
Looks good but only has a single output so I would need to couple this to a flip-flop and then some output transistors.
So I went back to the TL494CN and I realised that I had mis-read the test wave forms.
It is possible to have a dead time between pulses by using the PWM input.
So its possible to vary the amount of dead time between pulses.
I found a rather good/inexpensive multi-meter for measuring frequency and duty cycle.
Tenma 72-7735
http://mcmb2b.com/cgi-bin/test/72-7735.html
It has an RS232 output too.
They also do an inductance meter 72-8155 but I cannot find a local supplier for it.
Regards
Rob
Hi kingrs,
Yes, LTSpice is amazing. It has shocked me several times so far in its accuracy. Too bad it can't predict Naudin's metglas MEG. ;) Make sure you get the latest update by clicked
Tools ->
Sync Release. One thing I love about ltspice is FFT spectrum displays. On a graph just right click and select FFT. You'll see a window full of options. Default is usually good. Sometimes you'll want to bump up the data samples. Click OK then you'll see the FT graph. The default display is DB vs. Log. Just left click (not right click) on the DB or KHz numbers to change this. I like linear in Cartesian display. This will give you the typical Amplitude & Phase. Another great option is testing for component noise. I used it to accurately design a 3 stage FET amp that had less than 1nV/Rt noise. In noise mode it will tell graph the voltage noises inherent in parts. LTSpice knows about all the different types of noises from thermal noise to shot. Lots of goodies.
Quote from: kingrsI added an extra 1.6Ohm resistor to the circuit in series with the input coils and it seems to make little difference to the current, which seems odd.
Actually I already placed the 1.6 R in the inductor. Just right click on the L1 or L2 and you'll see it has 1.6 ohms for the series resistance field. Same goes for L3 & L4 except it's 37 ohms. So you don't need to add another 1.6 R. Also you'll see all the inductors have parallel R and parallel capacitance.
The inductors are coupled with the Kx commands. I could have just did K1 L1 L2 L3 L4 0.99 for example, but I broke them up to give each coupling a unique coefficient so as to represent the MEG core since the two secondary coils are appreciably separated from the actuators.
This is going to be interesting to see the results. I think Naudin's 100K is closer to 35K, but people are saying it varies in resistance.
BTW, I tried replacing BUZ11 with IRF540 and what a difference. The irf540's are a lot slower than buz11's so the sim is a lot faster.
Regards,
Paul Lowrance
Hi Paul,
The BUZ11 is probably older than Methuselah and is very slow response wise.
The IRF540 is a lot quicker from start to end it takes just 120ns where the BUZ11 takes a painful 570ns, time to make a cup of tea, walk the dog, have a nap and its still just firing up.
So you would be far better off using a more up to date MOSFET like the IRF540, and of course there are better and faster mosfets than even the IRF540.
http://www.irf.com/product-info/datasheets/data/irf540n.pdf
rather than the old BUZ11
http://www.farnell.com/datasheets/45278.pdf
The IRFU3707 start to end is just over 100ns but is only OK for 30V so it would need protection from that back emf.
I have been thinking it may be worth protecting the device by straping a zenner diode of say 10 to 20V across the input coil.
Regards
Rob
Hi Rob,
My datasheets show the BUZ11 turns on twice as fast as the IRF540-- 30ns vs. 60ns. Although the IRF540 turns off a lot faster-- 180ns vs. 50ns. I'm thinking the turn on speed is very important in the MEG.
So don't discount those BUZ11's, lol.
I've been doing some FEMM simulations on the MEG. Due to the PM's the core field drastically drops at around 15 volts. I made an exact copy of the Metglas 320 core, 100 turns on actuator (MEGv21). Naudin's FEMM simulations are incorrect because he assumes the core's permeability is the same at 30KHz vs. DC. That Metglas core at 30 KHz has roughly the same permeability as FEMM's iron material. Naudin's power charts show a larger rise at around 20 to 25 volts input, which seems to agree with my FEMM simulations. Also anything after 25 volts does not help as much, which is probably why his power chart shows a huge decrease in efficiency at 30 volts.
I did sims on both Neo and ceramic magnets. The neo's will probably work well at higher input voltages and the ceramics work better at lower input voltages. This is because the neo's introduces a stronger traverse field near where the magnets and core touch. Traverse fields can hurt permeability. Ceramics might be a best first choice since they work better at lower input voltages and that's what Naudin used.
Regards,
Paul Lowrance
Actually I have it backwards. It's the turn off time that seems to be more important since the PM reverses the process. Still lets not discount the BUZ11 since that's what Naudin used and it is twice as fast turning on that the IRF540.
Paul Lowrance
Hi Paul,
Thats interesting, looks like every one and his dog makes a different version on the BUZ11.
RS have discontinued them all 4 of them.
They all appear to have different properties, and I bet if you bought two of the same and tested them they would show different speeds.
Price of the part seems to vary between manufacturers and it looks like its being discontinued all over the place.
Can you post your femm files.
Did you put in the material for the Metglas core (B/H curve), I had to do this the other day for a ferrite and it was not easy?
Core should arrive tomorrow and I will post a picture of it so you can see what it looks like.
I still need to order the bits for the pulse circuit. Are you getting a TL494 chip?
I noticed that the TL494IN is the one to go for as it works at -40c to +85c, (brrrrrr to damm thats hot!).
I need to check my scope still works and look at sorting out the interface to the PC for it.
I'm tempted to get a Stingray usb scope but the input burns out at +/- 50v and the potential offset probes cost more than the scope.
Regards
Rob
Hi Rob,
Can you believe NTE actually makes BUZ11's. It's NTE2389. Go to -->
http://nte01.nteinc.com/nte/NTExRefSemiProd.nsf/$$Search?OpenForm (http://nte01.nteinc.com/nte/NTExRefSemiProd.nsf/$$Search?OpenForm)
and enter BUZ11
Better yet is that there version is faster yet. TdON = 25ns typ., TdOff = 125ns typ. That's great because there's at least 3 stories nearby that have a large stock of NTE parts.
Also Jameco has part #870209 which is a BUZ11. There are two nearby stores that sell Jameco parts. According to Jameco they sell for 40 cents each at quantity of one.
http://mouser.com sells BUZ11's from a lot of different manufacturers.
http://www.digikey.com just has one version from Fairchild Semiconductor.
Quote from: kingrsCan you post your femm files.
Did you put in the material for the Metglas core (B/H curve), I had to do this the other day for a ferrite and it was not easy?
Attached is the FEMM file, but as previously mentioned all magnetic materials do not have the same permeability at 30 KHz versus DC. The permeability used by FEMM is DC permeability. Now what you could do if you wanted to be really accurate is create a FEMM properties of Metglas 2605SA1 using the metglas DC permeability graphs,
but you would then have to K-factor the coil current. The DC permeability of your metgals is 600000, but at 30 KHz is more like 10000. At 25 volts on the coil generates 280mA. So instead of setting the current to 280mA you would have to lower it to 0.280A * 10000 / 600000 = 0.0047A So in your FEMM you would be using the DC permeability of 2605SA1, but the current would be 0.0047 amps (not 0.280 amps).
Quote from: kingrsI still need to order the bits for the pulse circuit. Are you getting a TL494 chip?
I noticed that the TL494IN is the one to go for as it works at -40c to +85c, (brrrrrr to damm thats hot!).
I'll have to see if NTE has that or something close. Probably not, so I'll end up using a good old 555, lol. I kind of like those old buggers.
Rob,
You could possibly pick up a few 100 Kohm Thermistors? If you see the core temperature slowly decrease then you have it! I just bought some 100 Kohm thermistors for $2 each and some heat sync grease at radio shack for $2. I have two thermistors, one on the core and the other for canceling out room temperature changes. That way you can sense minute core temperature changes.
See the attached FEMM file.
Regards,
Paul Lowrance
Hi Paul,
Thanks for the FEMM files.
I have a temperature probe I can attach to the core, and using the supplied chart I can work out the exact temperature from the resistance change.
Well the core has arrived, very professional bit of kit.
The ultimate core "the AMCC-320":
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4727.jpg&hash=16cf3e381cb4af93a8a17944a579d0cb0ebc981f)
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4729.jpg&hash=1470ab06dd4ef0a7eca19fdacf5ba00cd9da1bbb)
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4732.jpg&hash=a54f37424711999ef2cb354d78929bc19e5050b5)
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4733.jpg&hash=d4425b78a6283ab2bb85dcbedc809b706ac5c007)
Regards
Rob
Great snapshots. Looking forward to seeing the data. Be careful inserting the PM's as Metglas is very brittle. That's probably another reason for using ceramics, initially.
What's the sensitivity of your temperature probe?
Paul Lowrance
Hi Paul,
Do you have any photos of your core?
You can upload to photobucket like I do for free.
Quote from: PaulLowrance QuoteBe careful inserting the PM's as Metglas is very brittle.
Yes, some edge bits are flaking off when you handle it, like chrome slivers.
I am very careful as to how I handle it as I appreciate it is a very fragile device.
Quote from: PaulLowrance QuoteWhat's the sensitivity of your temperature probe?
Tolerance I think is 0.2 C at 20C changing to 0.05 C at 40C.
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4736.jpg&hash=3634755ac53debc9dbad53e9c5bbd0c41687da21)
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4737.jpg&hash=90c7814afb8833c4fd5ff86432aa3cdd48c6a6d7)
Regards
Rob
Hi Paul,
I just ran your silmulation and noticed the input coils as being round in shape, surely these will be rectangular and tighter to the core than shown.
I do not think you will get the 100 turns in the small circle.
Also you have set the depth to 0.75 inches, this should be 50mm (2 inches).
I do not understand why the second input coil is set as air?
I would have thought it best to create a second current circuit and assign it to this but as a coil.
Sorry for picking holes in your sim.
Next step is to create a material like the Metglass, that means entering an BH curve.
Regards
Rob
Hi Rob,
It is fine to extend the coils, but it will make very little difference. If we understand how closed loop cores react to magnetic fields we know that the wire locations make very little difference unless you are talking about very high frequencies. Also it is the magnetic field from the wire that enters the core that makes nearly all the difference. FEMM simulation is DC current on a closed loop core, so extending the coils will only make a slight difference.
The reason I used Iron instead of creating a Metglas property is for two reasons:
1. The permeability of Metglas is not 600000 at 30 KHz. More like ~10000.
2. It takes a lot of time and it was a quick job; i.e., I'm spending far too much time on such things when I should be working on the "smoking gun."
The only way you could make the FEMM more accurate than mine is if you create a metglas property for the core and k-factor the coil current way down to adjust for the fact that the cores permeability drastically changes at 30 KHz. Also extending the actuator coil will help a tiny smidgen.
My Metglas samples arrived. :-) Two of them no less, lol! Although they are very small and round toroids, but that's fine since I am testing the MCE theory, not the MEG. Also I asked for a different material, 2714AF, which has nearly zero saturation Magnetostriction, <<1ppm, and annealed permeability of 1 million. Now if I can just get some time to work on it. I did not ask for the same material because you had already ordered it and it seemed highly unlikely they were going to give out a huge 320 core anyway. My goal is to first analyze the material to see if it is has high potential as the 2605SA1 (Naudin's core). The material I really want is Hitachi's Finemet. Man that stuff has "free energy" spelt all over it!!!
Attached is a very poor picture of the two MP3210P4AF cores. Sorry, my camera is very poor. There are two toroid cores and come in plastic cases. I removed one of the plastic cases so you see the core material.
Regards,
Paul Lowrance
Hi Paul,
Good news about your cores then.
Yes I read up about the Hitachi cores and it does mention Metglas in their litrature so I don't know if they are related.
So its onwards with the testing, I am putting together more orders for components, wires, plugs, boxes etc.
You can get multi-stranded enamelled wire which is design for high frequency circuits, and I was wondering if this could be used on the output coil to get the voltage down and the current up.
I will go with single core wire for now and put in a couple of taps so I will have some options.
I am really looking forward to getting this this up and running, quite exiciting really.
;o)
Regards
Rob
Hi Rob,
I think you're right that starting with solid wire is best first option. Stranded wire is probably better in the MHz region. At 30 KHz it would probably just add more wire resistance. Solid magnet wire with really good electrical insulation sounds good. After all we're talking roughly a thousand volts.
BTW, my Metglas 2714AF cores have an inner diameter of 0.78 inches.
About two weeks ago I place a sample request for Finemet, but Metglas contacted me and said, "I received notice from our sister company, Finemet, that you are also interested in nanocrystalline cores. Did you have a particular core in mind?" So Finemet is a sister company of Metglas. I was too embarrassed to ask for another sample, so I'll probably have to pay for a Finemet core. Do you know how much Finemet cores cost relative to Metglas?
Rob, now that you'll have a legitimate Naudin Metglas MEG do you think Naudin would reply to an email for possible pointers? It sure would be great if he could get involved at least a little.
Regards,
Paul Lowrance
Hi Paul,
I agree with MrAmos, be cheeky and ask them for a nanocrystalline core sample too.
Also see if they do a cut C core in that material.
I saw somewhere on the Hitachi website a cut C core that was about 1Kg in weight but I am not sure what material it was made of.
Regards
Rob
Hi Paul,
maybe could one use videotape as super thin core material ?
Or audio pure iron tape ?
There is then only a few mikrometer thick surface layer of this
magnetic material on the plastic tape background layer...
Or won?t this work ?
Thanks.
Regards, Stefan.
Hi Stefan,
I think the ratio of tape to ferrous material will be very large, so in effect you will end up with a core made of 99% pastic.
I'm leaving the building of the cores themselves to the experts at Metglas or Hitachi metals.
Hi Paul,
I have emailed JLN to ask what material he used for the permanent and whether he tried altering the DTC - dead time control on the TL494 to allow a gap in between the two input pulses.
What I suspect is that it may be able to get a different peak frequency using this variable DTC and possibly getting an even better COP.
At the moment the two input pulses are right on top of each other(follow one right after the other) and leaving a small gap may allow the output waveform to form its peak at a lower or higher frequency.
I am looking at testing both with and without DTC to see if I can improve upon the experiment that JLN carried out.
Using 555 timers for this will be complex and I suggest sticking with the TL494 as it is built for this push-pull process.
There are even faster/high current and more up to date versions of the TL494 but I think these might be a bridge too far.
Regards
Rob
Hello MEMM-audience,
Molina-Martinez used,following his patent description, Hyperco !
S
dL
Stefan,
Do you know of any high perm. magnetic tapes? That would be good for my Method #2. Method #1 is all about having the Eddy currents absorb MCE energy, which is why the MEG relies on electrically conductive cores. Most ferrites will not work for method #1 unless the ferrite has electrically conductive magnetic powder.
Rob,
Do you have any scope shots? Hopefully Naudin will reply because he should have books of tips. It seems Naudin spent a great deal of time precisely replicating the MEG. He'll probably recommend the conditioned resistors, which might be worth a try.
I'm still messing with taking MCE measurements on the Metglas. For some reason the office has been flooded with thermal noise fluctuations, intense electrostatic random fields. It took one day to narrow down and finally accept that my LM741 burnt up.
Anyone who's interested,
---
How to measure MCE:
Place two thick copper wires down the center of toroid core #1-- the thicker the better. Three feet long is fine. The two wires will always be in series. Through out the entire experiment you will run DC current through both the wires. The amount of DC current depends on the core material. You will need to bring the core to at least 1 T or saturation. Core #2 does not have any wires going through it, but it should be slightly above the two wires and close to core #1. There should be no wires touching core #1 or core #2. On top of each core is a Thermistor. You should place heat sync grease under and around the thermistor and core to help thermal conductivity. The thermistors go to an op-amp circuit. See attached image. A gain of 100 is fine for 100K Thermistors. One lead from each Thermistor goes to ground, while one lead from Thermistor #1 goes to op-amp negative input, and one lead from Thermistor #2 goes to op-amp positive input. This will eliminate any op-amp output changes with room temperature change, but a noticeable op-amp output change if the temperature changes in just one core. Therefore, if just core #1 temperature changes then we'll see the op-amp output change, but if the room temperature changes then we'll see no change on the op-amps output. Each Thermistor should be at least 100 KOhm.
Stage #1. The wires are connected so the DC currents create an applied field on toroid core #1. Also your op-amp filter should filter out higher frequencies. Two 22 uF capacitors work just fine; i.e., one cap from output to ?in, and the cap from +in to ground. Make sure the toroid cores are touching the least amount of solid. Stagnate air is one of the best heat insulations. You will need to completely cover the toroids to prevent any air circulating on the cores. It will take some time for the heat waves to settle down. Eventually you get a steady signal from your op-amp output. Note you'll want to use the op-amp balance pins. You can slow adjust a fine precision pot so the op-amps output approaches zero volts.
Stage #2. You will need to quickly swap the two wires so the currents in core #1 are canceled. This means there's no current in the core. IOW, current runs one direction through wire #1, but current runs the opposite direction in wire #2. You should always have a cap across one of the wires to dampen any spikes when you swap the wires. Use two switches such as a solenoid to swap the two wires. You want the wires to carry current as often as possible. The goal is to maintain the same temperature throughout the entire experiment. After you swap the two wires core #1 will need to be degauss. One method of degaussing core #1 is with a 3rd wire going through core #1. You will need to apply AC current (no DC current) through wire #3 and then slowly dampen the AC current until there's eventually no current. The AC current needs to at least be as high as the total current in wire #1 and #2. It should only take a few seconds at most to degauss the core. Although cores with square hysteresis curves may take longer. So now the core is degaussed and the temperature should drop. You need to write down how fast the temperature drops and to what degree. A graph would be great. It will probably take at least one minute before you'll see a good change in temperature and several minutes to reach peak.
When the temperature changes settle down, you can swap the two wires again so there's a net applied field in core #1. Now the core will heat up. You will once again record the temperature changes. This completes one full cycle. Now you can once again swap the wires and degauss the core for another cycle.
You want to make sure the core is truly saturated during stage #1 and completely degaussed in stage #2. You can verify this by running a tiny amount of constant ac current through core #1 and measure the voltage across the core. This will show the permeability factor. The voltage across the core should be maximum when the core is degaussed and minimum when saturated. Remember, the ac current must remain the same.
See the attached image. The POT (R10) is to balance your thermistors since no two thermistors are alike. You'll probably want to make R10 a large resistor in series with a large pot unless you have a pot with enough resistance. I drew the Thermistor (R5) as 98.7K merely as an example to demonstrate that two thermistors are not alike. R6 & R8 should be matched resistors or two adjusted pots since they need to be the same resistance. I find the old LM741 a good choice believe it or not because the 741 doesn't mind big capacitors. Surely there's always a better choice. The circuit drawing does not show the op-amp balance resistors. Op-amp pin 1 goes to one end of a 200 K pot, pin 5 goes to the other end of the 200 K pot, the center of the pot goes to a 200 K resistor, which goes to V+. If 200 K does not offer enough balance then try 100 K values.
Regards,
Paul Lowrance
Hello PaulL,
I simply can not follow what you are descrining here :
Place two thick copper wires down the center of toroid core #1
Please can you explain by a simple drawing what you mean when you say :
Place ... down the center - and later : the wires in series ???
It is really confusing because 2 wires in series ?? why not use one wire then ??
Regards
Kator
Hi Paul,
Yes I am confused too with the arrangement of the wires.
Don't you mean you wind a couple of hundred turns onto the toroidal core?
Using two wires (two coils) means a simple setup but you could use a 2 pole change-over switch and just use one coil.
Regards
Rob
Hi Paul,
By adding a resistor (150R) and capacitor(2nf) across the ground and coil inputs you can remove that nasty back emf:
Hi Kator and Rob,
This experiment is used to determine the amount of MCE in a core.
The attached image is a simpler version. There is a toroid (gray) and a switch (green) and two wires (red & blue). When the switch is position A the current runs down the red wire and then down once again through the blue wire. When the switch is position B the current runs down the red wire, but then up the blue wire. In position A there's a net current going through the toroid. In position B the current in the red and blue wire cancel each other out.
The idea is maintain a constant temperature. The amount of heat given off by the wires is equal to Turns * Current. The direction of current does not change the amount of heat given off by the wires. Therefore, the core will receive the same amount of heat from the wires regardless if the switch is in position A or B.
Regards,
Paul Lowrance
Quote from: kingrs on October 28, 2006, 08:30:29 PM
Hi Paul,
By adding a resistor (150R) and capacitor(2nf) across the ground and coil inputs you can remove that nasty back emf:
Yeah, that's a good idea. I just place a cap across the red wire so when the switch breaks the current the cores energy can dissipate across the cap. I think 150 ohms like you said is fine. Also the caps internal resistance is good enough. The lower the R just means the energy dissipates at a slower rate with a little extra ringing. The ringing helps degauss the core a tad. ;) Although you really need a separate wire through the toroid & circuit to properly degauss the toroid. Do you have a good circuit that would work for degaussing?
Regards,
Paul Lowrance
Hi Paul,
The core should not need to be de-gaussed.
If the field is permanently alligned in some way then surely you have a bad core material.
I suppose the TL494 could be used as de-gauss tool, connected to a "H" bridge so that the single coil can be used.
This way the polarity can be flipped at several khz.
I have been looking at the various PWM chips that Texas Instruments manufacture.
The TL598 looks like a better choice than the TL494 in that it has a totem-pole output stage. So there is not pull-up or pull-down resistor required which should reduce power consumption but more importantly the switching time on/off and off/on will be faster and hence will save power in the switching of the output stage.
Also they have some more recent devices with what I think is called "TrueDrive" for a high speed, high current drive into the output MOSFET. I think these are also available as a device that sits between the pulse chip and the mosfet.
All these bits may add up to a few percent of input efficiency but if the output is twice the input then it would mean that its not too important at this stage and can be looked at later once the concept is proved to work.
Regards
Rob
Regards
Rob
Hi Rob,
Any word from Naudin?
...
The Metglas cores have square hysteresis curves. If you saturate the core and remove the applied field most of the field will remain. Don't you think it's important to degauss the core for the MCE experiment?
Regards,
Paul Lowrance
Hi,
I updated the MCE experiment
http://peswiki.com/index.php/Site:MEMM#How_to_measure_MCE
Regards,
Paul Lowrance
Hi Paul,
Paul Lowrance wrote:
QuoteAny word from Naudin?
No nothing yet, he is usually very good in getting back to me, so I can only think he is deep into some research at the moment.
Paul Lowrance wrote:
QuoteDon't you think it's important to degauss the core for the MCE experiment?
If the flux polarity can be switched at 24khz and above then cannot see why you should need to worry about degaussing it.
As I understand it, you only degauss something that can retain a magnetic field, like materials used for a permanent magnet or a piece of audio/video tape.
e.g.
If you place a screwdriver into a coil of wire (say 500 turns) and pass DC current of a few amps into coil for one second the screwdriver will be magnetized and will stay this way for many months.
I once made up a coil on a piece of 15mm copper tube for a mate of mine who is a plumber, and he thinks its great that he can now magnetise all his tools using a drill battery pack and this coil.
Try placing a powerful rare earth magnet on your core, remove it and then see if the core can attract a small piece of soft iron.
Regards
Rob
Now I see. You thought I was talking about the MEG when I was talking about the MCE experiment. In the MCE experiment the core must be degaussed to get accurate MCE readings. The purpose of the MCE experiment is to determine how effective a magnetic core is for "free energy." Anyhow, the MEMM wiki should clarify this. :-)
http://peswiki.com/index.php/Site:MEMM
Rob, in your conversations with Naudin did it seem he encouraged people to replicate the MEG?
I just cannot understand why Naudin did not publish any details about his attempts to close the loop. His MEGv21 documentation was superb, but the other versions were not fully documented.
Regards,
Paul Lowrance
Hi Paul,
No, I knew you were talking about the MCE experiment.
You have a Metglas toroid which should have the same properties as the Metglas cut C core.
The toroid should be able to handle 25 to maybe 100khz pulsed current.
Therefore I would have thought that you should not need to degauss the core between switching polarity.
Re. JLN MEG
I think it was the fact that the resistor did not heat up quickly enough to show the level of power shown in the scope.
It looks like that was his main concern that stopped him from exploring it further.
He agrees with Thomas Bearden's claim and managed to get the same results.
He also says that a COP >1 will only be confirmed by a closed loop.
Reading between the lines, I think he would be very pleased to see someone finish what he started.
He managed to power a 9w fluorescent lamp from 3w input, so I would have thought that it should be possible to work out the equivalent power to produce the same amount of light or supply it with enough mains power to create the same amount of light using different balast resistors and measuring the current on a true RMS meter.
regards
Rob
Hi Paul and anyone else following this thread,
Just thought I would update you on my progress at the moment.
I have ordered the following:
enamelled copper wire:
1 Kg of 0.4mm
0.5Kg of 0.5mm
0.5Kg of 0.6mm
test equipment:
Tenma 72-7760 DMM
Tenma 72-960 LCR meter
components:
4 x TL494 PWM controller
2 x TL494CDG4 PWM controller
2 x UC28025DWG4 PWM controller
4 x 2SK3586-01 100V 73A 0.025R power MOSFET
2 x CASE, DESKTOP ABS+AP 225X165X90
5 x 1N5820 schottky diodes 3A 20v
10 x 1R resistor 0.5W
50 x 6R2 resistor 0.5W
4 x 15V 5w zenner
Parts outstanding is 2 x LCD temperature meters, single sided PCB, PCB drafting sheet, tin plating powder.
I have opted for two types of controller chip, the TL494 which you know about and the UC28025.
The later is surface mount with 0.05" pitch so will save drilling the 16 holes in the PCB.
I have chosen this because:
1. It was in stock at Farnell (quite important).
2. Up to 1.5A of output current to directly drive the MOSFET gate (uses a totem-pole outputs).
At the moment I am trying to figure out the best way to mount the output coil formers onto the core.
I have some Kapton sheet and various plastics I can try.
Then I need to build a rig to sit the C core into for winding.
Looking on the TI site there are some PWM controllers that can obtain a 97% efficiency with the ideal conditions.
Regards
Rob
Hi Paul,
Latest update of parts ordered:
10 x ceramic magnets 50mm x 19mm x 10mm
4 x ceramic magnets 50mm x 19mm x 5mm
The gap in the AMC-320 core is 35.5mm, so I will use 2 x 10mm (two outer) and 3 x 5mm (in middle).
I can have magnets cut from tiles but the machine setup cost of 38 GBP means I will test with the 19mm wide magnets first then look at fine tuning when things start to look promising.
I want to start the testing of the pulse circuit next on breadboard before I commit it to a pcb artwork.
I need to work out the timing capacitors for the 6 way switch on the pulse unit, (bit of trial and error I think), as the ranges need to overlap just slightly.
Aiming for 100Hz to 100Khz (maybe I should have ordered a 12 way switch now).
Regards
Rob
Rob,
Sounds good. I commend your professionalism. I wanted to add a side note for those people who do not have much money to spend that anyone can build the MEG on a shoestring. I'm the master of shoestring experiments, lol.
I'm discovering some really interesting effects with such nanocrystalline cores such as Metglas. From the BH graphs provided by Metglas we can see it requires very little transverse fields to saturate the core. A transverse field is not a longitudinal field. The toroids coil causes a longitudinal field. Although any stray field such as Earths magnetic field is considered a transverse field. Earths 0.5 Oe field is enough to saturate the Metglas core. Such transverse fields could cause undesirable effects in the MEG. Perhaps one solution is to encase the MEG core inside a relatively thick iron round enclosure. Another possible issue is radio waves. Any radio wave could cause similar problems. Unfortunately a big chunk of iron is not going to block magnetic fields above a few KHz-- a few hundred KHz more realistically.
Such problems are causing havoc on my MCE experiments. One moment I can measure strong MCE, but the next it's completely gone because the core snaps back into saturation caused by stray fields. As mentioned in my previous post, it is vitally important to completely degauss the core in order to perform the MCE experiment. Especially with cores that have high squareness such as Metglas cores.
Regards,
Paul Lowrance
Hi MrAmos,
I have a lot of diffent pic chips and an ICE-Breaker development board and Microchips PICkit 2 Development Programmer/Debugger.
Its sorting out the program to generate the pulses using the interupt timers that will take time.
Also I worked out the steps are going to get a bit lumpy at the high end of 100Khz even using a 20MHz chip.
I may go down that route later on.
Hi Paul,
Yes I wound about 40 turns onto the AMC-320 core and hooked up to my new LCR meter, and I noticed that when I rotated the core slowly around on the worktop the Q value and inductanced changed slightly by about +/- 2%. Very odd to see this happening.
I can do the test again and give you the actual values.
Meter uses 1Khz and 120Hz for its reading.
Regards
Rob
Hi Rob,
Wow, + and - 2% is lot for just changing the cores orientation relative to the external fields. And that's not even removing the stray field, just changing orientation. These nanocrystalline cores are amazing! Now if you had a large piece of iron, perhaps even silicon iron, that could _completely_ surround the Metglas core. I'm wondering how much that could help. Your Metglas core is large, so the iron core would have to be huge. Not sure how effective such a huge iron core would be at 1 KHz much less 30 KHz, but it's worth a try.
Personally I wouldn't bother with it for now. I'm thinking that if for some reason you are still unable to precisely replicate Naudin's scope shots then it might help to find a material capable of removing stray fields.
Regards,
Paul Lowrance
I've been using a method in an attempt to eliminate stray DC magnetic fields. I have two large toroids. One is silicon iron and the other I have no idea. These two toroids are stacked on top each other. I then place my Metglas core inside the two toroids.
The above will momentarily work, but it has problems. I took a compass and placed it inside the staked toroids. I then rotated the compass and notice it did not change. The toroids were blocked the Earths magnetic fields, which is in agreement with FEMM. This works fine if your toroids are degaussed. I then placed my toroids near a PM simply to slightly magnetize the toroids. I then removed the PM far away and repeated the experiment. This time when the compass changed to fixed direction when placed inside the toroids. This meant the toroids were applying a magnetic field on the compass.
This could momentarily block DC magnetic fields if you degauss the toroids, but any stray field could re-magnetize the toroid.
That covers DC fields. AC fields is another story. I'm thinking a small Faraday cage could block most AC fields. The Faraday cage needs to be non-magnetic wire. Any magnetic wire such as iron causes too much induction. Copper wire should work. Also the toroids need to be away from the Faraday cage walls. Since I live in Los Angeles, CA. there's a lot of radio ways at just about any frequency. We know the Metglas cores are affected to fields in the MHz.
Regards,
Paul Lowrance
Hi Paul,
Sorry, don't get too excited yet, it looks like the meter takes time to give an accurate reading.
I tried it again and it is not conclusive, in fact placing a ceramic magnet into the core has little effect on the inductance and Q value.
I placed the core onto a 1' cube polystyrene block in the middle of the kitchen floor away from metal objects and found that even after waiting for the meter to settle the reading barely changed when the core was rotated though the 360 deg.
They are as follows:
Inductance = 11.04mH
Q value = 42.0
Coil = 42 turns of 0.56mm enamelled copper wire wound on the middle on one of the "C" halves.
Things may look different when the core is operating at 25kHz.
Regards
Rob
Rob,
That's why I was surprised that simply changing the orientation changed the inductance that much. A few percent sounded like a lot from just simply rotating, but whatever. :-) Rotating just changes the transverse alignment, but the Earths and stray fields are still there. Now you say it _barely_ changes ... that's fine. I have been focusing on nulling out Earths magnetic field and stray radio fields. They do indeed make a difference when performing the MCE experiment. :-(
Regards,
Paul Lowrance
Quote from: kingrsI tried it again and it is not conclusive, in fact placing a ceramic magnet into the core has little effect on the inductance and Q value.
Rob,
If that's your final conclusion then there must be something wrong with your meter. I don't think there's a core on Earth that remains unchanged by placing a magnet on it.
I'm curious what the problem is.
Regard,
Paul
Rob,
I had a thought why you are getting varying results. As previously stated many times, these Metglas cores have square BH curves. Also the cores require very little stray fields to saturate. I've seen even slight 60 Hz stray fields cause the core to saturate. Once the core is saturated then it doesn't matter if you remove all the stray fields-- the core remains saturated unless you degauss it. The Metglas BH graphs show the proof how sensitive these cores are to transverse fields. BTW, my metglas core is more sensitive to transverse fields than yours ... lucky me.
If your core is already saturated then that could easily explain why even a PM has less affect on inductance. I have been fighting with these Metglas cores for several weeks now. They are so pernickety it's a nightmare as far as MCE experiments go. As far as inductance measurements ... I'll leave that task up to you. ;-)
Regards,
Paul Lowrance
Hi Paul,
The Ceramic magnets have now arrived, and they are a perfect sliding fit, a single turn of pvc tape around the center magnet and they form a tight fit.
I did a re-test on the meter and it looks as if it may be the pvc tape binding the core is relaxing, so this may explain the gradual changes.
Certainly placing a magnet in the core is now changing the Q and inductance.
| Q Value | Inductance mH | Mag Centre | Mag on Leg |
| 39.1 | 11.75 | N | N |
| 38.5 | 10.73 | Y | N |
| 36.0 | 11.25 | N | Y |
"Mag on leg" means the magnet stack is attached to the left leg pointing in towards the core centre.
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4765.jpg&hash=0353ef41066b42313d6fa9d2a3d2a59fd2db5062)
Regards
Rob
Hi Rob,
Nice picture!
Surely you are tired of the induction measurements, but if by chance you already know, how much current peak-peak did your induction meter generate while measuring your Metglas core?
I'm wondering if there are induction meters that first attempt to degauss the inductor before measuring the inductance? Metglas cores are highly unusual in that -->
1. They saturate at practically no applied energy. I wouldn't be surprised if they saturate from a person sneezing. ;-)
2. There permeability is so intensely high that it would require the induction meter to apply very low amount of AC current. Otherwise the meter is just saturating the core. The meter will still report induction even if the core saturates because of the materials high permeability during magnetic flip while measuring inductance.
Regards,
Paul Lowrance
Are there any lower priced and more readily available suitable lower permeability alternatives to Metglas cores available in the UK?
Regards
Hoppy
Well, Naudin did a cheap silicon iron version, but I showed errors in his interpretation of the scope shots. Surely it's possible to get "free energy" from silicon iron, but it seems very difficult and not sure the present MEG design can do it.
The Metglas MEG is another story. Unless Naudin faked the Metglas MEG scope shots it shows "free energy," but according to the theory the timing and everything has to be just right. It's unknown just precisely how critical everything is. Hopefully Rob will find out, especially if he can get in contact with Naudin. :-)
Regards,
Paul Lowrance
Hi,
No reply from Naudin :O(
The core can be ordered from the Elna magnetics in the US, for about ?100 for the AMC-320.
Still waiting to see if I can avoid the import duty of an extra ?20 with FedEx because it is a sample core for testing.
I will see what I can find out about the LCR meter but I would imagine the power is going to be very very low or the PP3 9V battery will not last very long.
I could use a second meter to measure the current draw off the supply battery.
Regards
Rob
I wonder what happened to Naudin.
Rob, I just noticed your avatar pic. Very nice! Did you just change it?
Regards,
Paul Lowrance
Hi Paul,
Yes I did just change it, something a bit different I suppose.
The meter consumes 40mA according to the manual but I guess that I would need to couple it to a scope in case they are charging up a capacitor to give it a wallop of say an amp for 1 ms.
Regards
Rob
Hi Rob,
Wow! That's really saturating. :) I quickly counted 42 turns on your MEG. Might be off a few. Is that close? I did this in FEMM using my core material, which requires about 1.7 times less applied field to saturate than your core. A few weeks ago I created a FEMM model for my material. At 42 turns my core saturates at just 8 mA. Although it only requires 2 mA reach half saturation.
These cores are awesome! I'm convinced there's enough stray 60 Hz current to saturate these cores just by touching one of the winding wires. Anyhow, even if an induction meter generated say 0.5 mA the core would still be saturated unless you first degauss the core because these cores have high coercivity-- square BH curves. So first, the L meter would have to degauss the core and then apply the 0.5 mA AC.
You bring up a good point about a possible cap in meter that might generate an initial current surge. Personally I wouldn't spend any time on measuring the cores inductance and Q for now. If for some reason you cannot replicate Naudin's scope shots then it might pay off to build a simple degasser. I built one out of a 10 turn high precision pot and a 60 Hz AC voltage source. You only need to make sure the initial current, depending how many turns on your core, is enough to fully saturate the core, and the final current is definitely low enough after turning the pot all the way. The idea is apply 60 Hz AC current and then slowly decrease the current. You could do the same with a fancy circuit, but I'm always in a hurry. Although, given that these cores are ridiculously sensitive it is very difficult to keep these cores degaussed. As you know, the human body is a big capacitive antenna. Just getting close to the wires generates AC current. And what about radio signals? There are dozens of radio stations near by. So the wires could pick up these signals.
Regards,
Paul Lowrance
Hi Paul,
Bit off topic, but I have just finished fitting a blue LED illumination disc between the fan switch and the ceiling in the bathroom.
Made it from a GU10 LED lamp (18 blue LEDs) and some 10mm acrylic sheet (shaped using a table router). Had to break all the glass away from the PCB boad using a drill vice, desolder each LED and then shaped the PCB to fit into the back of the switch.
Looks very good (bright blue glow) when you switch on the fan now.
I have to agree with you about the low current saturating the core, from all my FEMM sims it was nearly always around the 10 to 20mA of current at 500 turns for this size of core even for silicon steel.
I am fosusing on building the pulse circuit now and I need to work out the capacitors required to create the overlapping freqencies.
I have log and lin 10K pots, so one of them should provide a linear sweep over the frequency, not sure which so I ordered both.
Regards
Rob
Quote from: kingrsI have to agree with you about the low current saturating the core, from all my FEMM sims it was nearly always around the 10 to 20mA of current at 500 turns for this size of core even for silicon steel.
Supposedly your core has permeability of 600000 as compared to like 5000 to 10000 for SiFe. :-)
Quote from: kingrsI am fosusing on building the pulse circuit now and I need to work out the capacitors required to create the overlapping freqencies.
I have log and lin 10K pots, so one of them should provide a linear sweep over the frequency, not sure which so I ordered both.
You're still replicating the Naudins MEG? I think the MEGv2.1 is a good start since that seems to be the only Metglas MEG that Naudin shows all the necessary voltage and current scope shots.
Regards,
Paul Lowrance
Hi Paul.
I may try this to build this after the MEG is complete:
http://www.ee.surrey.ac.uk/Workshop/advice/coils/BH_measure.html
Then I can plot the curve for the frequency that JLN came up with.
It would be very interesting to see how the ferrite magnet in place alters the relative permability of the core compared to a vacumm.
In theory the permeability should increase dramatically.
Regards
Rob
Hi Paul,
Just working out the frequencies now.
You can roughly calculate them but there is no pattern that I can work out to calculate the exact result.
TL494 push pull breadboard testing pics:
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4768.jpg&hash=9d41acf6f571fb62d779b5d89b19f29004fe40ca)
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4769.jpg&hash=98b902586df066cc7c885982e0e5f0bc16e65d07)
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4770.jpg&hash=0c894c2eeee67956a4d28001b8fceb24ea00383d)
Fan light I was telling you about.
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4766.jpg&hash=88afbc1325c136b98a4a024cd1bf70f99d887eff)
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FDSCN4767.jpg&hash=e079e68b66413400ad38c2364f6b41be2eb5c0f0)
Regards
Rob
Hi Paul,
According to a write up for the TL494 the frequency for push-pull setup should be:
http://focus.ti.com/lit/an/slva001d/slva001d.pdf (http://focus.ti.com/lit/an/slva001d/slva001d.pdf)
f = 1/(2 * R * C)
But as usual it does not quite work out correctly.
It is possible that there is some capacitance on the breadboard itself but I cannot see this being more than 100 pf.
Its out by about 35% on the larger capacitance (88nf) and gets better, 27% for 10nf, on the lower the capacitance so that rules out the stray capacitance idea.
I will add a supply capacitor across the chips supply rails just in case its that.
Last thing I need is an unstable push-pull pulse circuit.
Its lower limit looks like to be about 365Hz where the spec. tells you that its 1kHz.
Pity, I was hoping to get down to about 10Hz.
Regards
Rob
Hi Rob,
I suggest using a 1uF timing capacitor at Pin 5 of the TL 494 to get to the lower 10Hz frequencies.
From the writings of your link, Page 8, in Fig. 6 you can see the 1uF value for the lower 10Hz operation and this goes with a 200-300 kOhm timing RT resistor, preferably a potentiometer of 220 kOhm. With these two values you can go up to maybe 1kHz from the few tens of Hz, so you have to switch in a lower value timing capacitor to reach higher frequencies. A good solution at the potmeters to connect a 22kOhm one in SERIES with a 220kOhm to get fine tuning possibility and to avoid switching the potmeters when higher frequencies are needed.
If you cannot get near to the lower 10Hzs with the 1uF/220kOhm, then I suggest bending out Pin 5 and Pin 6 pf the TL494 horizontally to disconnect them from the board and solder directly the cap and potmeter wires to these pins.
rgds
Gyula
Hi Rob,
That's odd. Are you sure your cap isn't 88 pF? What's your timing resistor value? My datasheet shows the TL494 having the following range -->
minimum 0.47 nF and maximum 10000 nF for timing cap
minimum of 1.8 Kohm and maximum of 500 Kohm for timing resistor.
That should give a range of 0.1 Hz to nearly 600 KHz, but that's probably not the valid frequency range. I don't see how your circuit could have anywhere near a few nF stray capacitance.
What's your Vcc? Valid ranges is 7 to 41 V.
http://focus.ti.com/lit/ds/symlink/tl494.pdf
Regards,
Paul Lowrance
It looks like the value of the discharge resistor causes some issues if it is too high for the larger caps. 20k or above with 0.47uf and it jumps all over the place.
At the higher frequencies like 1kHz and above it is not so much an issue.
With testing I can get down to about 10Hz but I have to use a resistor of about 1k + 10k pot.(big frequency swing though).
Also I have managed to get the duty cycle working, 47% down to 0% (off), although the meter flakes out at about 5% and then shows it as 99.8%???
I can see on the scope that it is about 2% but the meter tells me different.
24khz on the scope with storage so I could get a photo, sorry for the poor quality.
Notice the alternating pulses, it does work so far!
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FIMG_1413.jpg&hash=8b0821305f036fbbabc4d531f1ecbd223b4029e5)
734kHz scope trace, maxed out with no caps.
(https://overunityarchives.com/proxy.php?request=http%3A%2F%2Fi100.photobucket.com%2Falbums%2Fm25%2Fkingrs%2FIMG_1414.jpg&hash=6102264f9071c18294dfbfb722aa2c3a72dd365f)
Regards
Rob
Pulse output results so far:
12 ranges tested with the capacitor leg lifted off the breadboard and coupled direct to the capacitors. Some of the capacitance values where I combined 2 or more capacitors have been measured with a TENMA 72-960 LCR meter.
Resistance swing is 9.95kR - 52.8kR using a 47kR trimmer and a 10kR resistor.
1. 200kHz - 533kHz 22pF,
2. 80kHz - 300kHz 110pF,
3. 25kHz - 110kHz 468pF,
4. 12kHz - 35kHz 1200pF,
5. 6kHz - 20kHz 2200pF,
6. 3.5kHz - 12kHz 3.9nF,
7. 1.2kHz - 5kHz 10nF,
8. 600Hz - 2kHz 22nF,
9. 300Hz - 1000Hz 47nF,
10. 155Hz - 500Hz 89nF,
11. 100Hz - 290Hz 146nF,
12. 40Hz - 132Hz 386nF
Other ranges I also tested that will not be included in my 12 frequency range pulse unit:
25Hz - 80Hz 600nF
13Hz - 40Hz 1.1uF
5.7Hz - 17Hz 2.5uF
Note that the ranges 11 and 12 and the 3 above all had a some level of double pulsing:
Example, the first output of the TL494 - Q1 would show two pulses widths together and the output for Q2 would at the same period show two spaces, which is fine, except that the frequency is now altered slightly.
It is quite rare, only occurring once every 2-3 seconds or so.
This is not bad going considering the chips minimum frequency is specified as 1Khz and I am pushing this down to 5Hz.
I think I have improved the stability by applying a 0.1uF capacitor across the GND and Vcc.
Next step is to test a MOSFET coupled to it and then design a PCB for it all.
The front panel will have:
12V in (pulse supply)
power LED
0-30V in (MOSFET supply)
A out
B out
f Range select (1-12)
f fine tune
duty adjust
output control switch(push-pull / push)
I will leave a space in the top left of the panel for a 2-line LCD display for future expansion. (frequency and duty info display)
So far so good.
Regards
Rob
Count me in on the replication project! :)
I ordered my Metglas C-cores today, hopefully they will come in the next couple weeks.
Excellent work Paul!
Eldarion