Magnitizing magnets??

[Joe Gwinn]

If the LC does oscillate, wouldn't the simple circuit tend to demagnetize the just magnetized device since the current flows in the reverse direction? That is, perhaps the simple circuit works because the Q is low enough so that the reverse current is a fraction of the initial forward current and so has little effect. The small amount of reverse current still might be enough to damage the capacitors, requiring the use of the diodes. If the Q is that low, then the circuit cannot save much energy since it is mostly gone before the reverse current occurs.

No, there is no reverse current through the coil. The circuit routes the current into partially recharging the capacitor to the original polarity. The key is that the voltage across the coil reverses when the coil tries to maintain the original impulse current, the capacitor having run out of electrons.

It's a very clever circuit, and simple. A SPICE model would be informative.

[Mike Sulzer]

No, there is no reverse current through the coil. The circuit routes the current into partially recharging the capacitor to the original polarity. The key is that the voltage across the coil reverses when the coil tries to maintain the original impulse current, the capacitor having run out of electrons.

It's a very clever circuit, and simple. A SPICE model would be informative.

Yes, the circuit in the patent does that. But the simple circuit I am referring to is the one where you just discharge the capacitor through the coil. That has the potential to oscillate: the capacitor gets charged in the reverse direction, and then current flows in the reverse direction. Now I suppose in practice this might not happen because of the SCR switch; I think it should should turn off at some point stay off, presumably when the current gets very small after reversing the voltage on the C. I think any circuit has to handle issues such as these, and the one in the patent seems like a clever way to do it. Of course, he is not the first person to use diodes to properly and safely get rid of energy stored in inductors!

[Mike Sulzer]

No, there is no reverse current through the coil. The circuit routes the current into partially recharging the capacitor to the original polarity. The key is that the voltage across the coil reverses when the coil tries to maintain the original impulse current, the capacitor having run out of electrons.

It's a very clever circuit, and simple. A SPICE model would be informative.

Another thing. It would seem that this circuit needs a lot more capacitance to get the same peak current. Not only do you need two caps, but I think they have to be bigger. This is because the voltage across the second capcitor as it charges reduces the voltage across the coil compared to if it were just connected to ground. It would take careful analysis to determine how big this effect is. (It has to do with the rise time of the current through the coil which of course has some inductance.)

[Joe Gwinn]

Yes, the circuit in the patent does that. But the simple circuit I am referring to is the one where you just discharge the capacitor through the coil. That has the potential to oscillate: the capacitor gets charged in the reverse direction, and then current flows in the reverse direction.

Oh. Naturally, I thought we were discussing the patent circuit.

A simple charged capacitor suddenly connected across a LR circuit will happily oscillate, unless the R provides sufficient load to damp the oscillation.

Now I suppose in practice this might not happen because of the SCR switch; I think it should should turn off at some point stay off, presumably when the current gets very small after reversing the voltage on the C. I think any circuit has to handle issues such as these, and the one in the patent seems like a clever way to do it.

The current through the SCR will not at first reverse, so the SCR will continue to conduct. The voltage on the capacitor will become fully negative when the inductor has exhausted its stored energy, and the current will stop. This will cause the SCR to turn off, if the current remains below as SCR-dependent threshold long enough. In practice, the oscillations may be fast enough that turnoff doesn't happen until the oscillation has largely died down, unless the SCR is equipped with snubber circuit of some kind.

What people try to do is to choose the coil winding dimensions such that the assembly is critically damped, so all the stored energy is absorbed by coil resistance in the first pulse. The problem is this is hard to reliably achieve in practice, especially if one wants to be able to use various coils, so there always needs to be some kind of snubber circuit. Eddy current loading also affects this - The R of importance is actually the AC resistance.

The simplest snubber would be a big diode from the top of the inductor to the top of a big heavy high-power resistor in parallel with the inductor. The diode is oriented such that the initial surge (when the top of the inductor is positive) is blocked from the resistor. When the inductor starts to supply current, the voltage at the top will become negative, the diode will conduct, and the heavy resistor will absorb all the remaining energy. This resistor needs to be physically large, so it can absorb the entire impulse without harm.

Of course, he is not the first person to use diodes to properly and safely get rid of energy stored in inductors!

But he is the first to think of how to make the snubber recycle the energy, versus dumping it into a big resistor. That's why he was awarded a patent.


Coming back to the original question, J M Fahey is trying to build an industrial magnetizer that works on a household power supply. Fortunately, he can space operations out (25 speakers per 8-hour day is one bang every 20 minutes), so it is practical. My guess is that the one SCR, one diode, and one heavy resistor et al will cost about what two SCRs and two diodes will cost, and will be less trouble or load on the household power.

[Joe Gwinn]

Another thing. It would seem that this circuit needs a lot more capacitance to get the same peak current. Not only do you need two caps, but I think they have to be bigger. This is because the voltage across the second capacitor as it charges reduces the voltage across the coil compared to if it were just connected to ground. It would take careful analysis to determine how big this effect is. (It has to do with the rise time of the current through the coil which of course has some inductance.)

I'm perplexed. In Steingrover (4,258,405) there is exactly one energy storage capacitor, designated as object 8 in Figure 2.

[Mike Sulzer]

My guess is that the one SCR, one diode, and one heavy resistor et al will cost about what two SCRs and two diodes will cost, and will be less trouble or load on the household power.

I agree the cost would be similar. I think you charge the C through a resistor of large enough value to prevent blowing a breaker. I think that is independent of the hardware after the C.

[Mike Sulzer]

I'm perplexed. In Steingrover (4,258,405) there is exactly one energy storage capacitor, designated as object 8 in Figure 2.

Don't be. I am just wrong. Figure two really does fix that potential problem in figure 1.

[Mike Sulzer]

Here is a similar idea for a somewhat different application.

Patent US4363067 - Transistor chopper protection circuit - Google Patents

[Joe Gwinn]

Here is a similar idea for a somewhat different application.

Patent US4363067 - Transistor chopper protection circuit - Google Patents

So it appears; I'll read it.


By the way, I reread the Steingroever patent, and was reminded of a significant advantage, that an undamped circuit (even with later pulse energy recovery) has 2.7 times the magnitizing current of a critically damped (one impulse) circuit for the same capacitor, and loses no energy when fired, so the size of the pulse is not much constrained by the power supply.

[Joe Gwinn]

Another thing. It would seem that this circuit needs a lot more capacitance to get the same peak current. Not only do you need two caps, but I think they have to be bigger. This is because the voltage across the second capcitor as it charges reduces the voltage across the coil compared to if it were just connected to ground. It would take careful analysis to determine how big this effect is. (It has to do with the rise time of the current through the coil which of course has some inductance.)

You were talking about Figure 1. I'm pretty sure the two capacitors are the same size, and no larger than needed to generate the magetization current.

This is a form of resonant transfer, and is very efficient. If you google on resonant transfer, you will get a lot of not quite relevant hits. I learned the trick in the 1970s where the underlying technology of T1 digital lines was documented in the Bell System Technical Journal (BSTJ). T1 ran at 1.544 megabits per second over wires many kilometers long, which was a breakthrough in its day. The T1 networks are all gone now, but the T1 signal structure is widely used, and will never die - too much embedded base.

[Mike Sulzer]

You were talking about Figure 1. I'm pretty sure the two capacitors are the same size, and no larger than needed to generate the magetization current.

With the advantage of an undamped circuit you mentioned above, I think that is right. Also, with that advantage, it might be that the two switch, two diode circuit is worth it even for occasional use because of the savings in C for a given energy release..

[jack briggs]

I bought a MAG 24c about 1 1/2 years ago when my capacitive discharge magnetizer bit it. It works real well, but you have no control over how much a magnet is magnetized. It just saturates it to full charge. I knock the gauss down with a ceramic mag when a certain pickup model calls for it.

I'm asking the factory if the 24-C could be plugged into a 15A variac to vary the charge. Seems feasible, no?
I have a WWII-era Bendix demagnetizer to degauss alnico, but it would be nice to have a variac work with the 24-C and find the charge level that works, dial it onto the variac and not have to degauss to get a prescribed charge.



cheers,

[Jim Darr]

I'm asking the factory if the 24-C could be plugged into a 15A variac to vary the charge. Seems feasible, no?
I have a WWII-era Bendix demagnetizer to degauss alnico, but it would be nice to have a variac work with the 24-C and find the charge level that works, dial it onto the variac and not have to degauss to get a prescribed charge.

Jack,

Clever idea. Let us know if the manufacturer says it is feasible to use the Variac.

I've always charged on the 24-C to full charge and then backed it off with a plate demagnetizer using wooden spacers to get my ideal charge for a specific recipe.

I switched from Neos to the 24-C in order to get a consistent and even charge each time...I've been pleased with the results. But your idea, if it is possible, should, at least in theory, save an extra step.

Thanks

Jim

[EtLa]

Supposedly Leo Fender used a magnetizer made from a truck starter to magnetize Tele magnets. Does anyone know how he did this?

ken

I think Leo was using a car magnto magnetizer that run with car battery like mine. H. G. Makelim Co. 1957 magnetizer

[jack briggs]

Jack,

Clever idea. Let us know if the manufacturer says it is feasible to use the Variac.

I've always charged on the 24-C to full charge and then backed it off with a plate demagnetizer using wooden spacers to get my ideal charge for a specific recipe.

I switched from Neos to the 24-C in order to get a consistent and even charge each time...I've been pleased with the results. But your idea, if it is possible, should, at least in theory, save an extra step.

Thanks

Jim

Jim,

Here's a pic of the Bendix demagnetizer I picked up last year.

I uses a variac to adjust the level of demagnetization. I emailed Magnetool about the feasibility of using a variac in front of the 24-C and the response was that it could be used to vary the output.

cheers,