Magnetic amplifier made from common 12 volt transformers and magnetic cores

[eneuro]

Attempting to splice aluminum wires?

AlumiConn Installation

But I haven't got such aluminum to copper conectors, so joining two srewed aluminum wires and pressing them on PCB with solder and insulating from air to avoid any oxidation should work in this prototype saturable inductor.

[iXod]

Idea behind using two cores in magnetic switch is to cancel this load inducted voltage in controll wires , so everything is ok at above magnetic amplifier circuit?

I've known this as a "saturable reactor".

Dave

[eneuro]

Finishing this research project too and I'm very close to final magamp PCBs like this 1inch2:



Etching this prototype tomorow and testing, than half smaller PCBs (similar to this one scaled down to size 12.5mm x 12.5mm -half inch )  for magamp switches rated at  HF low currents will be made 

We'll see if it works maybe even with common small cheap bead ferrite cores 

[AF6LJ]

Oh Boy does this bring back memories.
The days when industrial controls were large hunks of iron instead of fist sized chunks of silicon.

[eneuro]

The days when industrial controls were large hunks of iron instead of fist sized chunks of silicon.

Nowadays we have better materials for toroid cores 
Anyway, such magamp has some advantages-there is no way to break this thing by overvoltage if insulation holds, so no worry about VDS, VGS, transients, etc   

I've soldered my prototype low current HF magamp, but in slightly different configuration... as you might see this (sorry, not a great photo in artificial light conditions) without control wires, since my plan is saturate this thing using... neodymium magnet



Maybe, someone have idea what could be saturation magnetic flux density B of such ferrite bead?
Trying to estimate inductance Approximate Inductance of a Toroid in this thing to choose switching frequency to get decent impedance at not saturated state.

Lets try figure out, if I  will be able saturate this thing with strong magnet, so it could be nice to know expected saturation levels of ferrite beads, but unfortunatelly I haven't got any datasheet of this core-it was salvaged from.. CRT monitor flyback filter 

No chance to get such detailed information about this random ferrite core, like we have for this  PC90 material: Low loss and high saturation magnetic flux density power ferrite with close to 0.5T magnetic flux saturation 

I've found for NiZn ferrite 0.33T and in this publication 
Choosing Core for Flyback Transformer and Choke Designs they says about 0.36T saturation magnetix flux and 1.1T for iron powder, so probably something like 0.3T or less 0.2T  might be decent estimation for those unknown ferrite beads? 

Anyway, can't wait to experiment with this new magamp concept -we'll see what happends soon 

[calexanian]

Was it already covered that these systems are not really amplifiers but just saturate the transformer core constricting the amount of AC flux that is passed along to the load therefore reducing power?? Look up an Amplidine. Those were widely used in gun turrets until modern servo amplifiers were figured out.

[AF6LJ]

The days when industrial controls were large hunks of iron instead of fist sized chunks of silicon.

Nowadays we have better materials for toroid cores 
Anyway, such magamp has some advantages-there is no way to break this thing by overvoltage if insulation holds, so no worry about VDS, VGS, transients, etc   

I've soldered my prototype low current HF magamp, but in slightly different configuration... as you might see this (sorry, not a great photo in artificial light conditions) without control wires, since my plan is saturate this thing using... neodymium magnet



Maybe, someone have idea what could be saturation magnetic flux density B of such ferrite bead?
Trying to estimate inductance Approximate Inductance of a Toroid in this thing to choose switching frequency to get decent impedance at not saturated state.

Lets try figure out, if I  will be able saturate this thing with strong magnet, so it could be nice to know expected saturation levels of ferrite beads, but unfortunatelly I haven't got any datasheet of this core-it was salvaged from.. CRT monitor flyback filter 

No chance to get such detailed information about this random ferrite core, like we have for this  PC90 material: Low loss and high saturation magnetic flux density power ferrite with close to 0.5T magnetic flux saturation 

I've found for NiZn ferrite 0.33T and in this publication 
Choosing Core for Flyback Transformer and Choke Designs they says about 0.36T saturation magnetix flux and 1.1T for iron powder, so probably something like 0.3T or less 0.2T  might be decent estimation for those unknown ferrite beads? 

Anyway, can't wait to experiment with this new magamp concept -we'll see what happends soon 

I am not so sure you could get ferrites to do the same thing. My experience with ferrites suggests that if you saturate them with enough current (in my case RF) they get hot and their magical powers go away pertinently.

[calexanian]

I agree. I would first experiment with small traditional transformers. A split primary split secondary transformer will give you access to all the windings. Much like a small flat pack PCB mount transformer, say in the 20VA or so range should serve as a good test platform.

[calexanian]

Oh. and I forgot to mention. Be careful of the voltage being generated by the AC side of the transformer back on the DC control side. Don't want to blow up your lab supply with high voltage AC being thrown back at it. You should have proper protection via blocking diodes, and a filter network of some kind to keep things safe.

[eneuro]

My experience with ferrites suggests that if you saturate them with enough current (in my case RF) they get hot and their magical powers go away pertinently.

Yep, but RF frequency is orders of magnitude higher than switching frequency <<1MHz and those ferrite bead are designed to have low losses at this lower frequency and filter high frequency crapp and dispate this energy as heat, so no supprise for me 

In this setup ferrire bead will not transfer high currents, but low current and magamp will work as a switch detecting presence of strong magnetic field from neodymium magnets.
I've checked datasheets of a few cheap small  neodymium magnets and they have easy 0.4T at 0.7mm distance from magnet.

[eneuro]

A split primary split secondary transformer will give you access to all the windings.

Maybe title of this thread is somehow talking about transformers used for magamp, but I'd rather like to use custom magnetic cores with far less number of turns, thanks to higher switching frequency (but of course no a few MHz RF frequency  ).

Anyway, in this latest experiment with ferrite bead we need NOT ONLY access to windings, but also to... its core, since there is no third control DC coil there, my idea is use pure neodymium magnets to saturate this thing 

BTW: I've updated thread topic and added magnetic cores, while of ocurse not only 12V transformers now are in my interest 

[eneuro]

Yep, thanks to lectures by prof. Walter Lewin
8.02x - Lect 16 - Electromagnetic Induction, Faraday's Law, Lenz Law, SUPER DEMO
https://www.youtube.com/watch?feature=player_detailpage&v=nGQbA2jwkWI#t=1103

physics works and my magamp increases load current since induction falls in those  ferrite bread magamp windings when strong magnet creates decent bias magnetic field inside toroid 


Similar effect in this simple full-wave SMPS driven at the same 400kHz frequency and low current-when strong magnet is applied to its secondary ferrite bead, induction collapses and induced EMF is very low, not able provide even fraction of mA to power up white diode 


Of course saturating core in classic SMPS converter is not a best idea since its primary inductance will collapse too, while those windings are usually well magnetically coupled, but in the case of magamp made even using ferrite bead it looks like there is no difference if this core were saturated by third DC bias current applied in the middle of this thing (as shown a few posts above in classic magamp with diodes to improove sensitivity) and ... strong magnet, which can have very usefull practical applications 

Preparing for more spctacular demo with more power  full-wave DC-DC converter 

[eneuro]

I would first experiment with small traditional transformers.

I was able saturate this bigger (choke I guess) E trafo core, while I've one coil was in series with those leds above (anti-parallel configured to detect AC, not DC only), however problem with those E cores is that coils a in the middle and we can only try expose this thing to external magnetic field from both sides, which means to be able apply symetric external field for each half of AC waveform, we have to put two magnets from opposite sides and.. we have a problem-magnetic field in the middle of this E core will cancel if we use the same magnets, so only part of the coil will see strong magnetic field.



In the case of, when those two magamp coils are not in the middle, when external field is apllied in the middle (by magnets or DC voltage) than magnetic flux is divided and flows to each side of magamp, so when additional diodes are present it makes symetric saturation easy at the same level of additional magnetic flux in whole AC period, so it looks like there is advantage when those (transformer) coils are not in the middle of E core.

Anyway this choke has much higher inductance when not saturated, while this simple ferrite bead one side has probably less than 100uH, since it blocked 100kHZ AC at similar way like this known SMD 0.1 Ohm 100uH inductor at 100kHz.

So, yep this choke (each coil has) much higher inductance than SMD inductor and ferrite  bead tested, but magnetic flux inside this thing is asymetric (the same single SMD inductor), when applied strong magnetic filed to ONLY one side of this transformer (SMD inductor), so probably we should be able to see that one of those LEds make more bright light, since those inductors will saturate asymetrically when we'll be close to its maximum magnetic flux density (saturation) 

Low power magamp controled by strong magnet looks quite interesting, however we have to know where apply external magnetic field, since in one direction it will be added to internal magetic flux created by its coils, but otherwise they will try to cancel each other, so we can still get this magamp core saturation effect when magnet field will be very strong, but when those magnetic fields are in correct direction, then the same effect can be achived using smaller magnet.

It was interesting experiment-I need to order a few stronger neodymium magnets to experiment more with this thing, but early results are very good and this thing can be usable even with ferrite bead