I am now developing the SFT Mk5. The Mk4 version of the site is now archived.
When you hold two magnets with their opposite poles close to each other the magnets attract. I guess nearly everybody knows this. Similarly, if you hold your magnets with like poles facing (the same poles, North North or South South) they repel. What happens though if while they are repelling you insert a sheet of ferromagnetic (magnetisable) material, like a sheet of steel, in the gap between them? The answer, in case it didn't immediately occur to you, is that both magnets are attracted to the sheet. This is the core principle of this device.
So, in this particular embodiment anyway, this device has two bellcranks or 'rocker arms' that have magnets at their ends arranged so that the magnets face each other with like poles (North pole facing North pole or South pole facing South pole), thus the magnets repel each other. In the image at the top of the page you can see the rocker arms (orange) with their magnets (red). The rocker arms are arranged, each with an axle at their centre, positioned so that the magnet at each end faces the magnet on the opposing rocker arm. All being equal, both arms would balance on their axles so that they are parallel because the magnets on each side are repelling each other.
A rotor with small steel tabs or 'fingers' around its circumference is arranged between the rockers so that the gap between the magnets on one side has a finger in the centre of the gap. Since there are an odd number of fingers around the rotor, they are evenly spaced and the rotor is positioned so that its axis is in line with the magnets, there will always be a finger between the magnets on one side of the rotor and a gap (no finger) between the magnets on the other side.
The magnets on the finger side (right in the above image) will be attracted towards the finger while those on the other side (left in the above image) will continue to repel each other. Thus the rocker arms will rotate on their axle until the right-hand magnets are touching the finger and those on the left side are as widely spaced as they can be.
If the rotor is turned the situation reverses, now the magnets on the right spring apart while those on the left are attracted to the newly-introduced finger in their gap. Turning the rotor changes which side of the rotor (and rockers) has a finger and which side has a gap and consequently the rockers rotate between being closed (close) on the right and open (widely spaced) on the left to being closed on the left and open on the right.
This reciprocating motion is transferred by means of bellcranks and sliding blocks to drive the flywheel by way of a scotch yoke. The flywheel crank pin also limits the extent to which the rocker arms can rotate and thus prevents the magnets on the closed side from closing enough to make contact with the finger.
Well I don't know the answer to that yet. My feeling is that the work required to turn the rotor will be less than the work produced by the magnets. But it's only a feeling based on some experiments I have done and I could well be wrong.
Think about pulling a strong magnet off a sheet of steel: it is very difficult to pull it straight up from the steel (in line with the magnetic force) whilst it is considerably easier to slide the magnet sideways (transverse to its magnetic field).
There will be a magnetic drag on the metal finger: first it will get dragged into the gap between the magnets as it enters the gap, we can call this an assisting force because it is reducing the force needed to turn the rotor; then it will get dragged back towards the gap as it exits the gap, this is an opposing force, it is increasing the force needed to turn the rotor. These two forces will balance each other out and in sum will not change the total force needed to turn the rotor.
There will be a little friction in the rotor bearings and a (hardly measurable) air resistance, each of which will add a little to the work needed to turn the rotor.
Most significantly there will be eddy current drag in the metal finger from the field of the magnets. This is caused by a magnetic field that is induced in any metal object moving through a magnetic field. That induced magnetic field will oppose the field of the inducing magnet(s) and therefore causes drag on the metal, in this case increasing the work required to turn the rotor.
Will that eddy current drag cause the work needed to turn the rotor to be more than, the same as or less than whatever work the magnets do? That is the key question. I don't know and the physicists I have asked about it seem unable to answer (but their immediate, reflexive reaction is to say "1st Law of Thermodynamics: you can't get out more than you put in!") Maybe I'm asking the wrong questions.
If (and only if) it takes less energy to turn the rotor than is produced by the movement of the magnets then some of the output energy could be used to drive the rotor and, yes, we would have free energy or overunity or perpetual motion or whatever you want to call it.
My intuition tells me that it will take less energy to turn the rotor but we'll have to wait until I can get some sensible answers from the physicists or until I can build and test the device.
I was watching one of Robert Murray-Smith's very early videos on his ThinkingAndTinkering channel on YouTube. He is a very interesting guy with a passion for energy systems, amongst other things, and has produced, literally, thousands of videos. In this particular one ( The Secret of Magnetic Motors) he was discussing the Wesley Gary magnetic motor and explaining that if you put a keeper (a strip of steel or iron) across the poles of a horseshoe magnet, if the keeper is thin and longer than the span between the poles, then there is a pole that appears at the ends of the keeper that are overhanging the magnet. The end attached to the North pole will also be a North pole and the end next to the South pole will be South. These poles appear because of flux leakage from the magnetic circuit. If you lift one end of the keeper off its pole (let's say it was the one on the North pole) the leakage flux pole vanishes and, so long as the keeper is still close to the magnet pole, then the end of the keeper is neutral, neither a North pole nor a South Pole. As you move the keeper further from the pole then the opposite pole appears on the end of the keeper.
So what's happening there? My explanation runs as follows: with a thick enough keeper the horseshoe magnet's field is entirely contained within the magnet and its keeper. If we use a thinner keeper then not all the magnet's field can be contained in it and some flux leaks out, giving rise to the flux leakage poles. When one end of the keeper is lifted well clear of its pole we've really just added to the length of the magnet, so if we lift the keeper at the North pole we have really extended the magnet at the South pole and our keeper extends the South pole of the magnet to the end of the keeper, which now presents a South Pole. As we slowly return the keeper to the North pole the magnet tries to complete its circuit, some of the South pole flux is still emanating from the end of the keeper but some also leaks back into the circuit. There comes a point where the flux leaking from the end of the keeper and the flux leaking back into the magnet's circuit are balanced, so the end of the keeper is now neutral, neither still South, from the influence of the magnet's South pole, nor yet North, from flux leakage at that pole.
Thinking about all that is what led to this. Below, under the heading 'Magnetism' I outline my model of magnetism. Really I have only recently developed it in order to explain, or at least give a metaphorical explanation of, what is happening in my device and why.
In this design I have kept it as simple as possible so that people can easily understand the principle. People understand mechanical forces much more easily than they understand electrical forces.
One physics forum to which this idea was presented quickly locked the thread. When my friend, who posted the thread, enquired with the mods why the thread was locked their response was, 'the device is useless and looks like a perpetual motion machine’. So I determined to keep this version of the device purely mechanical so that I don't get hung up in debates involving electricity, generation and so on. It is difficult enough to get physicists to even consider what is happening in this device because it "looks" like a perpetual motion machine and physicists know for sure such a thing cannot exist (but are still, apparently, able to recognise such an impossibility when they see an example.)
In a later version, if this one turns out to behave as I hope, I will surely incorporate a generator (or alternator), maybe solenoids or a triboelectric generator connected directly to the rockers and cutting out the flywheel and its related linkages, which may not be an ideal arrangement with regard to friction. However the crank wheel serves another important function, which is to limit travel on the rockers, preventing the magnets from sticking to the fingers, and I would have to find an alternative way to achieve that.
Absolutely you can. I am releasing the complete model with an 'unlicense' license (public domain).
I developed it in FreeCAD, an astounding piece of open source, user supported, free CAD software, so to do anything with the model you'll need to install FreeCAD. You'll also need to install the Assembly 4 Workbench and Fasteners Workbench in FreeCAD to work with the complete assembly.
Alternatively, there may be other CAD software that will import FreeCAD designs or you can import the STL files into most 3D CAD software and slicers etc.
The project is hosted on my GitHub at github.com/prajna-pranab/SFT-Mk5 from where you can download a copy (either clone the repository or click the green '<> Code' button and click 'Download ZIP').
Not having 3D printed anything before it has been something of a learning experience setting about getting this device built. Also, though I had heard about FabLab and had actually visited a FabLab in Manchester many years ago and taken a look inside my local FabLab (Aldeias do Xisto) I didn't really know much about how the FabLab system works. I dropped in to the Fundão Lab to see if I could get my first prototype printed. Unfortunately, the local lab had a week-long event happening when I went there and they were unable to give me much in the way of attention. I got the manager's email address and sent off an email about my project. There was no response, so the following week I dropped in to the FabLab again. João, the manager, only works there in the mornings because he teaches in the afternoons and i had missed him, so another hiatus for another week. The following week I went there in the morning and managed to catch João. He had a look at the model and was confident it could be printed but, although FabLab supports open source software and FreeCAD in particular, João confessed that he doesn't work with FreeCAD and is actually a trainer for Fusion, so it was no help to present him with the FreeCAD files and I would have to send him the STL files instead. So I went back and exported all the STL files.
Again, because of my inexperience with 3D printing, it turned out that the STLs were not arranged in
a useful way but João quoted me a price for the work and suggested I come in and familiarise myself
with the Slicer software. I was given a quick run around using Prusa Slicer and set it up on my laptop
before going home and laying out all the parts in Prusa. After a few iterations of doing that and
emailing the layouts to João for checking the layout was acceptable and the Lab began the printing.
A few days later the prints were ready. This is what I ended up with:
The image includes a couple of rods for pushrods and axles that I bought too. The 2mm dia rod is silver steel, which is ideal for this purpose because it is manufactured to much closer tolerances than normal mild steel rod and it can be hardened to make it much stronger. I had hoped to obtain both rods in silver steel but the 5mm was not available locally. Most of the printed parts needed cleaning up, to remove the support structures and reduce the layer lines on the surfaces, so I spent a couple of days doing that with side cutters, a Dremel and various grades of sand paper. I ended up with this:
What I had not taken into account is that the model is very small and my power transmission linkage was rather exotic and over complicated and I hadn't designed in any adjustment to align the spacing between the rockers and the fingers around the rotor. All of this meant that it was not possible to assemble the model in a way that worked and I had to redesign things so that it was mechanically more sound. Which is what this new design aims to achieve.
Other difficulties with the first prototype is that it called for (ideally) electrical lamination steel for the tabs/fingers and I emailed, in the end, four different companies that specialise in electrical laminations, none of whom replied. For the current prototype I may have to resort to deconstructing a transformer or electric motor to get the lamination steel. Ordinary mild steel should work ok but I believe that electrical (silicon) steel should improve eddy current losses and it has a higher permittivity and lower coercivity. I was rather hoping that one of the lamination companies would offer me some advice on choosing the best material for the purpose and perhaps even the geometry that would be best employed in this application. I think they took a look at the project and thought, "Eccentric inventor, free energy, not a company, don't bother." Oh well.
So the current status is that I have finished the redesign and am now waiting for another appointment at FabLab to print the parts. I've had a quote for the printing and, if the FabLab staff do it, an estimation of three weeks. If they have a printer available then I can do the printing myself (albeit with a few pointers, I'm still pretty new to this) but I'm waiting for a response on that.
Meanwhile, I spotted that Worten (a local electrical goods retailer) have a Creality Ender-3 V3 KE printer on offer at €248 - probably the best entry-level 3D printer available, it has auto- levelling, easy filament loading, PEI build plate, XZ-Core, direct drive all metal extruder, a max speed of 500mm/s and acceleration of 8,000mms^2 - so if anybody fancies buying one and putting my name on it ...
"But free energy cranks are all the same, really. They come up with a contraption that is just complicated enough to exceed their powers of analysis - and then claim they have broken the laws of thermodynamics. Magnets are often involved, as magnetism is particularly poorly understood by such people. (Tesla often comes into the picture too, though thankfully not in this instance.)" - a senior member on scienceforums.net
Well it seems that it is not only 'free energy cranks' who don't understand magnets, physicists seem incapable of giving useful and reasonable answers to questions about magnets too. Examining this device they came up with all kinds of unrelated and unsolicited complexities but entirely failed to respond to simple questions with simple answers. In the process of trying to get them to understand and respond to simple questions I vaunted several examples of magnets appearing to do work. They insisted that magnets don't do work because, "All dipole magnets act like a current loop (permanent magnets, too) and the force on a current is IL X B. That’s a cross product - the force is perpendicular to the current and external field. Work is a dot product of the force and displacement. The work is in the common direction of the force and displacement. Since the force is always perpendicular, this dot product is zero. There is no work done."
Maybe the above response makes sense to a physicist but it makes little sense to me, so I've developed my own model of how magnetism works. It may not be a particularly good model in terms of physics but, for me anyway, it explains what is happening in this device, at least in metaphorical terms. So here goes:
Still writing this... please check back in a few days.
Via Nostr: prajna@nostrcheck.me
1. In the FreeCAD files you may find the device referred to as SMT rather than SFT, that is because I originally called it a Switched Magnets Transmission but I prefer now to call it Switched Flux Transmission, named because it transmits input energy by way of magnetic flux switching to the output, I make no further claim than that. I make it very clear that the device requires the input to be powered, it won't do anything till you turn the rotor. In this configuration it is certainly not a 'free energy device', it has no generator and no feedback of energy produced to the input. If it works as I (perhaps vainly) hope then I can begin to think about generation and feedback but for now it's a kind of 'thought experiment', maybe even a 'curiosity', as it has been described.
In case you're wondering what the background image is, I had to take a break from the
project because we took a day trip to Spain. But my mind didn't take a break and, while travelling
I redesigned the output linkage of the Mk4 device. When we stopped for coffee I grabbed a napkin and
sketched out the new design. This is the sketch:
