The Magneformer-lenzless transformer ?

[tinman]

@MH-and all
I am going to start from the begining,so as all can see what it is im trying to show with the magneformer. As much as i have looked,i cant find the original device,so i have made a new one from scratch. Im guessing that the original is in one of my junk boxes out on the farm,as there just wasnt enough room in my new work shop to bring everything to our new home.

The core halves im useing are from an old flyback transformer,from either a TV or cpu monitor.
I now will make the pulse circuit for it,and it will be driven by my home made(kit)SG,and P/in will be in way of an SLA.This is so as the scope ground can be placed anywhere on the device for measurements.

Pictured below is the transformer part of the device ,that i made today. Pic 1 shows the transformer in its seperate pieces,and pic 2 shows how it all fits together. This way we can test it with and without the PMs in place,which are strong N48 neo's.

Once up and running,the first video will be about checking our DMM's for accuracy,and show how to smooth out the pulses across the DMMs-so as they read accurate.
In pic 1,the core looks like it is broken and is bent inward toward the open end's. But that is just because it is lying on an angle,and it is actualy straight. The cores are ofcourse ferrite ! grade unknown!.

[tinman]

Some results of the first run.
The neo's had to be swaped out for ferrite magnets,as they were far to strong.
Measurements are as follows after some fine tuning,and the resistor across the secondary coil is 81.7 ohms. Tank cap is not yet in place.

Without magnets
P/in=12 volts@ 7.79mA=.0934watts
P/out(flyback)=12.43 @ 4.45mA=.0553watts
P/out(secondary/resistor)=400mV/81.7 ohms=.00196watts
P/out total=.0553+.00196-.05726watts.
Efficiency is 61.3%

With magnets
P/in=10.4mA @12volts=.1248watts
P/out(flyback)=12.44 @6.8mA=.08459watts
P/out(secondary/resistor)=920mv/81.7ohms=.01036watts
P/out total=.08459=.01036=.09495watts
Efficiency is 76.08%

Both meters are of the same type,and were set at the mA scale,and large smoothing caps used on input and output.
Meters were then swaped over,and test ran again. The results are an average of the two test.
The two DMMs used have a .02mA difference.
The smoothing caps on the input were 1x 10 000uf high current cap,and two 4200uf caps.
Cap used on the flyback output was 1x 10 000uf high current cap.
No ripple detected across the two DMMs.

The first scope shot shows the device running without the magnets in place.
The second scope shot shows operation with magnets in place.
Scope shots taken befor fine tuned.

[gyulasun]

Hey Gyula

If the pulsed primary input field is opposing the permanent magnet, the core wont be saturated till the field of the magnet is flipped/reversed and then pushed further into saturation, if the input can deliver that much opposition to the cores permanent field. A ferrite mag can be demagnetized, or even reversed eventually, that is again if the input is enough. Brad isnt putting that much power in so saturation shouldnt be an issue and the magnets should last for quite some time.

Mags

Hi Mags,

Yes it is all okay what you wrote, I was mistaken in that I had believed literally the permanent magnet was really inserted into the coil while in fact the magnets were only attached to the cores but not inserted into the coil 

Gyula

[gyulasun]

...
Gyula:

I didn't know that a magnet would have such a low relative permeability.  You still might be able to take advantage of the polarization though as has already been stated.  Note the excitation from the primary coil is unidirectional.  This assumes the relative permeability is radically different depending on the direction of the external field.  It would be an interesting and easy test.  You just have to look at the slope of the current rise in the coil for same-direction and opposite-direction magnetic field generation by the coil wrapped around the magnetic core.  You cross your fingers and hope that you don't ruin the magnet.

I also wonder if the L-meter will be thrown off by the introduction of a magnet into the coil.  I assume they sample or sweep low-level AC frequency excitation for the coil under test and check the response to measure the inductance.  So if the magnet does indeed radically change in relative permeability depending on direction it may have a small heart attack (throw off the measurement algorithm).  It probably will read out as a high inductance - my guess.

Another point is that this is a transformer setup, not an inductance.  So assuming the core (any core) is coupling the energy properly, you _don't_ see inductance on the primary, you see the load, which is an LCR circuit.  So you see a wobbling resistance!  lol  Note since you are approximately at resonance, you are pretty much seeing the resistive component of the LCR circuit as the load of the primary.  So that means that Brad should see the voltage and current going into the primary winding as mostly in phase, assuming that the core/coupling is doing it's job properly to transfer the power.

MileHigh

Hi MileHigh,

Yes permanent magnets are almost fully saturated magnetically, many FE tinkerers are not aware of that and when they use magnets in a "closed" magnetic circuit where permanent magnets are used to "close" the magnetic path, they actually build an "air gap" into the circuit at places they insert the permanent magnet(s).

Well, L meters mainly use oscillators inside which excite the coil to be measured with saw-tooth like waveform so there are no abrubt amplitude change across the coil. Indeed you have to be careful when testing coils with magnets when using an L meter and move the magnet slowly in or out of the coil to avoid harmful induced amplitudes, to save the inside oscillator circuit. I did check air core coils with inserting different magnets into them and never had any malfunction in my L meter. The permanent magnet in itself when plugged into an air core coil does not change any of its original properties its permeability also stays the same. The L meter normally use a low power oscillator to drive the coils to be measured.

If you use a coil with ferromagnetic core and attach a permanent magnet to the core or just approach it closely with a magnet, then the permeability of the core can change a lot (it normally goes down) so the L meter shows a decreasing inductance value too. Taking sudden movements with the magnet causes the L meter to go out of range for some moments, then normally it returns to the new L value.

It is okay what you wrote in your last paragraph above, I would add that the load resistance  (earlier the 18 Ohm, now the 81.7 Ohm) establishes the loaded Q of the output tank circuit, and if this load does not change there is no 'wobbling resistance'. If there is no load resistance then indeed the LC tank has its own parallel resonant impedance and it is a high value (several kOhm or even higher) wrt the 18 or 82 Ohm loads, and it is also a constant value (if the switching frequency is stabil).

Gyula

[gyulasun]

Some results of the first run.
The neo's had to be swapped out for ferrite magnets,as they were far to strong.
....

Hi Brad,

My comment would be that the magnets still reduce the permeability of the C core on which the electromagnet coil is wound, this explains the higher input power taken when the magnets are in place. 
If you have an L meter, then you can check the inductance values first with the magnets in place (the setup is unpowered of course and leave the cores and the magnets possibly in the fine-tuned positions) and then removing the magnets you could adjust the core distances to arrive at the same L values for the coils like with the magnets and then test the input output.
Of course this may sound as an obsolote test now, the goal is to get more output with a reducing input, by using the magnets.

One more thing: if you happen to find with the L meter that the reducement of coils inductance (due to the magnets) is not the cause for the increased input power, then the explanation for the latter is what Magluvin also mentioned: more input is needed to flush out the PM flux from the core.  It is also possible that both the decrease in inductance and the extra flux from the magnets are indeed the cause for the extra input.  It would be good you would find your original setup of course and could check it again.

Gyula