The Tunnel Diode Really Works
October 1960 Radio-Electronics

October 1960 Radio-Electronics [Table of Contents] Wax nostalgic about and learn from the history of early electronics. See articles from Radio-Electronics, published 1930-1988. All copyrights hereby acknowledged.

The tunnel diode, with its unique negative resistance region, excels as an oscillator and switch. As described by Mr. Queen in this 1960 Radio-Electronics magazine article, is operating bias is critical - typically around 0.2V for oscillation. The article details a 27-30 MHz crystal oscillator using a 1N653 diode, requiring precise voltage division via a potentiometer (R1) and fine-tuning with R2. The load resistance must be less than the diode’s negative resistance (-40 Ω). A 16-turn inductor and adjustable capacitor (C2) optimize output. The circuit can self-oscillate without a crystal but becomes unstable. For switching, the diode's characteristic curve allows triggering via weak pulses or resistance changes. A light-sensitive version uses a solar cell in series with a relay - illumination increases voltage, snapping the circuit from a high-current (5 mA) to low-current (2 mA) state. Sensitivity can be adjusted so even a 40 W lamp at 18 inches triggers the relay. Reversing the solar cell's polarity enables triggering by interrupting light. The circuit’s extreme sensitivity allows activation by nearby AC fields, like a soldering iron.

The Tunnel Diode Really Works

By I. Queen, Editorial Associate

The tunnel diode is an amazing semiconductor. Articles on it and its applications have appeared in periodicals, but have often been too theoretical for readers who like to experiment. Exact voltage and component values are often missing, making it hard to duplicate the stated results. (The operating bias for a tunnel diode is generally very critical.)

The July issue of this magazine, page 26,carried an excellent discussion of basic principles of the diode. It showed that this device has a negative resistance sandwiched between regions of positive resistance. Fig. 1 shows a highly simplified form of the familiar tunnel diode curve. Current increases as voltage is increased up to about 0.1 volt, then decreases as voltage increases from 0.1 to a little less than 0.5 volt, when it again starts to rise with increasing voltage. Voltage and current values are typical for a 1N653 (Texas Instruments). Circuit theory indicates that such a device must be an excellent oscillator and switch. This article will give full details on a crystal oscillator for 27-30 mc and a switch that is triggered by light.

The Oscillator

The negative resistance region means that the tunnel diode is a ready oscillator, As explained in the July issue, the load line must intersect the region of negative resistance (Fig. 2). This load resistance is smaller than the negative resistance of the diode, specified as -40 ohms for a 1N653. E₁, the circuit voltage, is a small fraction of a volt.

No battery with such low voltage is generally available, so a divider is needed. Fig. 3 illustrates the crystal oscillator. Note the low value for R3 (which determines circuit resistance). Oscillations occur when the voltage across the diode is about 0.2. Start with a 100-ohm pot for R1. To set it, vary R1 (and R2 for fine adjustment) and listen for the signal on a nearby receiver tuned to the crystal frequency. Tune C2 for maximum output. For 27-30 mc, L is 16 turns of No. 18 wire, 14-inch diameter, close-wound.

You may also adjust this oscillator and observe circuit operation by measuring voltages. Start with maximum resistance at R1 and R2, then raise the voltage across the diode. (R2, the 20-ohm pot, is rather critical. Use 10 ohms for ease of control.) At about 0.17 volt (as measured with a 20,000-ohms-per-volt meter) you will note a sudden jump (to about 0.2 volt) and oscillations will begin.

This circuit will work with any overtone crystal from 27 to 30 mc without retuning. This is hardly the upper frequency range of a tunnel diode. With smaller tanks you should have no trouble pushing the range higher. If you remove the crystal, the circuit will self-oscillate with a rough and unstable signal.

A look inside the oscillator.

Switching circuit uses a tunnel diode. 

One resistor sits under the chassis of the switching circuit.

Once R1 has been determined for your particular diode and circuit, it may be replaced with a fixed resistor. A 47-ohm unit is just right for my oscillator.

For safety, limit diode current to a maximum of about 6 ma till you determine the current value for R1.

Construction is uncritical. In my unit, a battery and voltage-divider unit already on hand was used. That portion of the circuit could, of course, be mounted on the board with the other components.

Trigger Switch

When the load line intersects the diode characteristic in three places (as load 1 does in Fig. 4) we have a switch. As an example, suppose the circuit voltage is E₂ about 1.2 volts. Load line 1 resistance is about 250 ohms. When the circuit is switched on, it stabilizes at point A.

Now if a weak positive pulse is applied to the diode, the operating point will rise over the peak. It will move along the downward slope quickly (due to negative resistance) and arrive at C. (Point B is an unstable position, and is of little importance in this explanation). All this takes place instantaneously. Thus even a very weak pulse will trigger the circuit from A to C. In a 1N653 these points correspond to about 5 and 2 ma, respectively. (This is terrific amplification, if you want to look at it that way.)

The triggering push may be an externally applied voltage or it may be just a reduction in circuit resistance. To show this, refer to Fig. 5. Starting with maximum R1, lower the resistance. At about 5 ma the meter will suddenly deflect downward to 2 ma, although we are reducing R1. The operating point (Fig. 4) has moved up the curve till it reached the peak of the characteristic. At that instant (when it reaches the peak) it snaps to C. R2 (Fig. 5) is a limiting resistor.

Now increase R1. The slope of the load line becomes more horizontal until it coincides with load 2, at which instant the current will snap upward. For the 1N653 this occurs when the total series resistance is about 1,250 ohms. The current will snap from a low of 0.6 ma to 1.2 ma.

The circuit may be triggered by light falling on a photocell (Fig. 6). A low-resistance Hoffman S1-A solar cell is shown in series with the winding of a relay RY to be energized. Light falling on the solar cell increases the circuit voltage and triggers the circuit. The higher voltage corresponds to changing the load line from 2 to 3 (Fig. 4). When the new line touches the peak point on the characteristic, the circuit triggers and the relay drops out.

The photo shows an actual switching circuit based on the above experiment. The relay must have a low-resistance winding. Being unable to find a suitable unit, I modified a Sigma 5F-1000, which has two coils in series, for a total of 1,000 ohms. They are reconnected in parallel. to cut resistance to 250 ohms. Observe polarity of windings. The yellow lead of each coil is soldered to the black of the other. A 470-ohm resistor is shunted across the windings. This lowers the resistance and the sensitivity of the relay. Now the relay should be adjusted to pull in at 4.6 ma, release at 2.3 ma, approximately. Prime the circuit by pressing the RESET button, then lower R3 gradually until you are just below the trigger point. If triggering occurs, note the point and start over again, this time not advancing the control so far.

When properly set, the illumination from a 40-watt lamp at about 18 inches is sufficient to trigger the circuit and release the relay. Note that the solar cell is in series aiding with the battery.

This circuit may be made so sensitive that it can be triggered by switching on or off a nearby soldering iron or other appliance. The magnetic field of the ac does it.

If you reverse the polarity of the solar cell, you can prime the circuit with illumination on the cell. Then you can trigger the circuit by interrupting the light falling on the cell, even for a brief instant.