impulse magnetizer

[surdopickups]

we're almost there?

[surdopickups]

easy to be found, and to buy,

[surdopickups]

gentlemen, we accept offers

[surdopickups]

after use, the object is thrown away

[surdopickups]

http://www.icar.com/pdf/referenza_magn.pdf

not my circuit, although in some respects resembles,
but for those interested in the link is a pdf file circuited, the circuit diagram.

[surdopickups]

http://web.mclink.it/MC5690/capitolo6_2.html

[Mike Sulzer]

JM and Joe,

Yes, critical damping is what you need, but the "resonant frequency" must be low enough so that the pulse is long enough to meet the condition Joe described. This is different for different magnetic materials, and so one device might not most efficiently cover all jobs.

Just 2 small details:

1) although a capacitor and an inductor by definition do form a resonant circuit, you must never ever let it resonate !!!!!

a) because diminishing alternating field is the best guaranteed way to demagnetize.

b) if you use electrolytics, the strong inverting voltage and current will make them explode like grenades, no kidding.

2) so there are 2 solutions to this:

a) you calculate circuit constants, and add external resistance as needed to turn it into a critically damped circuit, , which by definition lets only one pulse pass, and only one way.

b) you put a big diode in parallel with the coil , so the initial large current peak keeps rolling as long as needed for the coil resistance to absorb it.

The big brother of what's used to damp relay pop and kickback.

3) my first magnetizer, brute force 3 x380V , 47A peak so some 25kW power used a Siemens "contactor", a brute force 3 phase relay typically used to start-stop big electric motors, elevators, etc.

And even so worked only on the AC side of things, before the full wave 3 phase rectifier.

Once while trying that machine prototype I tried to commute DC with a Frankenstein Movie type knife switch ... the blades opened but a huge arc kept shorting them ... and a small Sun started to form around it.

I had to run to the street and pull the 3 main 380V fuses, still under load of course.

Today all switching is done by replacing 3 rectifier diodes with large SCR ... but the main trick is the flywheel diode which keeps everything under control.

A similar trick is used on Capacitive/Impulse ones: an SCR grounds one end of the coil (the other one is always connected to +V) but again the flywheel diode helps.

[J M Fahey]

Giancarlo Franceschetti - Magneti Permanenti

Very good
Here's a rough translation:

CHAPTER VI - THE MAGNETIZATION

Magnetizing pulse

It's advantageous to reduce the number of turns forming the coil and increase the current as much as possible to minimize the inductance and avoid oscillatory phenomena which would cause demagnetization of the magnet.
The magnetization in itself does not depend on the duration of the pulse magnetization; however, especially in the case of metallic magnets that have a very low specific electrical resistance, the eddy currents generate a magnetic field opposite to that applied. The fig. 30 shows the effect of the induced currents during magnetization: curve a) for short pulses as the current increases a reverse flow of the induced currents can decrease the magnetizing flux, unless, curve b), the current pulse is long enough for the induced currents to decay before the pulse itself. In practice good design for magnetizing the pulse duration is about 0.01 seconds.

The principle is shown in fig. 31:

1) the capacitor C is charged to the voltage V in a time determined by the resistor R0
2) the capacitor is discharged on the coil L; if R is greater than 2 Φ (L / C) the pulse is unidirectional without oscillations, otherwise it must be mounted in parallel to the coil, the so-called "diode recycling" to dampen the negative half-wave. This latter feature allows the use of the magnetizer also as demagnetizer disengaging the recirculating diode in an oscillating circuit.