Saturable Inductor amplifier
Conceived and designed by sv3ora


Magnetic amplifiers and logic, always fascinated me. Magnetic amplifiers were used for a large period in time (and still do in some applications). However, magnetic logic seems to had appeared for a short period of time, in an era where there was switching from vacuum tubes to transistors. There were some circuits and devices developed and in one case a full computer has been built out of magnetic logic. There are also documents on the net about magnetic logic that uses no diodes at all to perform the function. However, most of the documents I have found, do not give enough information of how to build practical devices out of magnetic circuits, as they do not contain all the details needed for a successful implementation. I believe that apart from experiments and demonstration devices, magnetic logic was never really given the chance to prove it's capabilities, due to the new coming transistor technology which dominated everything.

I have performed a number of experiments in the past, implementing my ideas or existing ones, using all shorts of magnetic things, like square loop toroids, pulse logic and core rope memories. This time, my RF background made me re-think of the whole thing and eventually  I concieved a new device, which I call Saturable Inductor Amplifier (in contrast to the saturable reactor amplifier). This "device of amplification" can be used to built magnetic amplifiers and magnetic logic devices, posing significant advantages over the magnetic logic devices that were developed in the past:
In this page, I am going to perform a number of practical experiments, to show how my Saturable Inductor Amplifier can be used to perform different functions in electronics. First of all, I am going to begin with a simplistic explanation of the basic principles of operation and afterwards experimenting with it in different circuits.

My Saturable Inductor Amplifier, is based on the combination of the principles of inductor core saturation and RF blocking (choke action). I was introduced in the core saturation principle by reading books about saturable reactors, and in RF blocking mostly experimentally, through the large number of HF radio experiments I have performed throughout the years. Essentially, with this device, I combined magnetism with radio.

Inverter logic gate

The easiest way to explain the operation of my Saturable Inductor Amplifier, is by looking at the circuit diagram above. The diagram shows an inverter gate I have built and tested, which works in the following way. When "dcin" is disconnected or tied to ground (logic 0), the "rfosc" RF signal passes through the right diode, it is rectified and it charges the right shunt capacitor, to create a voltage on "dcout" (logic 1). The RF signal from the oscillator, passes also to the middle capacitor and reaches the middle diode and the choke inductor. The positive portion this alternating RF signal, cannot escape through this diode, but it is also prevented from reaching the ground, due to the high inductance value of the choke. I say the positive portion
of the alternating RF signal, because the middle capacitor causes the RF signal to go below ground at half of the cycle and so, there is also a negative portion that can pass through the middle diode. This negative portion is sent to ground through the left diode, so as not to leak to the dcin port. So, a logic 0 at the input of the inverter gate, results in a logic 1 at it's output.

If DC is applied to the "dcin" (logic 1), it is filtered from any RF by the left capacitor and it passes through the middle diode, to the choke and the middle capacitor. The middle capacitor blocks any DC from flowing further, so DC can only reach the ground through the choke inductor. As DC passes through this inductor, it saturates the inductor core, dramatically decreasing it's inductance. This decrease in inductance opens the way for the RF signal out of the "rfosc" to pass through the middle capacitor and the (now saturated) choke to the ground. With most RF signal flowing to the ground, there is little or any signal left to pass through the right diode and the right capacitor. The right capacitor is then discharged through the load connected to the "dcout" and reaches logic 0.
Hence a logic 1 at the input of the inverter gate, results in a logic 0 at it's output.

For the circuit to work efficiently, the inductor must be as small in physical size as possible (small core), but at the same time to have a large number of turns. Thus, core materials such as the 75 (J) ferrite, are possibly not suitable, because they require a few turns to reach to high inductance values. This means that they require more current (and thicker enamel wire) to saturate at this low-turns number, which would make the gate inefficient. A good starting point, seems to be these standard 10mH 1W molded chokes that are sold in quantities for very low prices. If you use such or similar inductors, you won't have to wind anything yourself. A good thing about the circuit is that it can be scaled up into larger powers easily, by feeding it with a higher current/voltage RF source. The output diode-capacitor can handle lots of power if appropriate power components are used.

To be continued, be patient please...

Back to main site