We study the tunnel diode on the example of 3I306M

In modern electronics, tunnel diodes are supplanted by components that are more convenient for solving the same problems. But why not experiment with an active element that was once considered one of the fastest?

Tunnel diodes are divided into those intended for amplifiers, pulse generators and key circuits. According to the datasheet , the 3I306 series diodes are designed for use in switching devices. The graph shows the dependence of the voltage drop across the diode on the current through it on a straight portion of the I – V characteristic:


The author’s characterograph is improvised, it consists of a signal generator, a 10-ohm resistor and an oscilloscope. In this case, an error occurs: one channel of the oscilloscope measures the total voltage on the entire serial circuit from the diode and resistor, and the other only on the resistor (the current can be indirectly determined from the second of these voltages). It is possible to calculate the voltage drop only on the diode by exporting the curves to a CSV file, and then generating the graphs in Python with matplotlib.

An example of the I-V characteristic of a tunnel diode on an oscilloscope screen:


Initially, the current through the diode rises to approximately 11 mA until the voltage rises to 150 mV, then sharply decreases to 500 μA and increases again. This is the area of ​​negative differential resistance where the current decreases with increasing voltage.

To study the operation of the diode in the switching device, the author connected it to two BNC connectors. Their cases are connected together, and a diode is connected between the central contacts. The signal from the generator with an output impedance of 50 Ohms is fed through the diode to the oscilloscope with the same input impedance:


The behavior of the diode is independent of the waveform. When the voltage exceeds the threshold, switching occurs. The author applied a triangular signal with a frequency of the order of 100 kHz. The current decays in 900 picoseconds, and the increase in 1.1 nanoseconds. Impressive, especially when you consider that the circuit consists of one part, not counting the signal generator. With a square-wave generator on timer 555, the switching takes about 100 nanoseconds.


But the magnitude of the output signal is small, since tunneling diodes operate at low voltages and currents.

Further, the author tries to use a switching diode for other purposes - in the generator. Here he will maintain undamped oscillations in the circuit:


The oscillation circuit initially consisted of one turn with a diameter of 9 mm and a capacitor of 2 pF. A 10 nF capacitor closes the generated oscillations to itself, not passing them into the power circuit. The supply voltage is 700 mV, after starting the generator continues to work when the voltage drops to 330 mV.


At first, the generator worked at a frequency of 295 MHz. When replacing the capacitor in the circuit with another one with a capacitance in pF, the frequency increased to only 300 MHz, which implies that the diode's own capacitance further lowered the frequency. Having calculated the coil inductance, the author further calculated the diode's own capacitance - 18 pF. The datasheet says that it does not exceed 30 pF, and this turned out to be so.

When observing fluctuations, it is important not to add additional capacity to the circuit. With a 10x oscilloscope probe, the capacitance is 10 pF, which is enough to further reduce the frequency. Therefore, the author closed the input of the oscilloscope to the case, having received another coil - measuring. Bringing it to the loop of the circuit, you can get a transformer without a core. The oscillation amplitude cannot be recognized in this way, but you can see how it depends on the supply voltage.


To increase the generation frequency, the author shortened the diode terminals and connected a capacitor with an axial pin arrangement directly to them. The coil is no longer needed, the inductance is provided by the outputs of the components. After applying a 700 mV supply voltage to the circuit, lasing began at a frequency of 581 MHz. How else to increase it? Take a cavity resonator?


Probably, it was not easy for the designers to work with tunnel diodes: the rule "we build an amplifier - it turns out the generator" was striving to be respected here. Therefore, the author has not yet tried to make an amplifier on such a diode.

The author took the output signal in the same way, and although it looks perfectly sinusoidal, it can be distorted, just at a frequency of 581 MHz the 1 GHz oscilloscope does not have enough resolution to detect distortion. Just as in the previous case, it is not possible to accurately measure the amplitude, which means that it will not be possible to compare this generator with the previous one.

Tunnel diodes are very “tender”: one of them failed during the removal of the I – V characteristic due to the amplitude of the signal from the generator too large, and the other from overheating during soldering. With the remaining eight, the author handled much more delicately. It is necessary to solder the diode at a temperature of no more than 260 ° C for no longer than 3 seconds and with a heat sink. The author does not have copper tweezers recommended for such purposes, 2 mm thick, but an aluminum clip, originally purchased for soldering germanium components, came up:


Diodes are also afraid of static, in addition, "diode testing by a tester is not allowed." After such an experiment, the author survived the diode, but during the test it did not ring in either direction. You need to determine the polarity from the illustration in the datasheet.

If you are going to experiment with tunnel diodes, purchase them just in case with a margin, but start observing these simple rules immediately. And then you will not lose a single one.