1

Everywhere I search, the thermal voltage in diodes is described either as

  • the microscopic property related to the average energy of electrons due to thermal motion
  • The constant in the ideal diode equation.

Now, I'd like to know the macroscopic significance of the thermal voltage. For example, something like it being the voltage across the diode in comparison to a diode a 0k if it were true.

It being a constant in the ideal diode equation is not so satisfying; I plotted typical values in the equation (https://www.desmos.com/calculator/fiwyqzw5ci ). There appears to be no discernible feature at is at the thermal voltage (black line), whereas the turn on voltage of 0.7 is a readily visible feature (blue line).

So, what is the measurable macroscopic significance of Vt thermal voltage?

ions me
  • 341
  • 1
  • 8
  • There's nothing magic about 0.7V - just change your vertical scaling to be something else, eg 1e-12 per division. – Kevin White Mar 30 '19 at 00:44
  • @KevinWhite Maybe it is arbitrary, but at normal voltage and current scaling, it is a good estimate of the forward voltage of a diode. – ions me Mar 30 '19 at 00:51
  • So for turn on voltage, you could give the macroscopic interpretation: if you are within normal operating range, the forward voltage will be around 0.7 volts. Question is: What is the equivalent description for thermal voltage? – ions me Mar 30 '19 at 00:53
  • 1
    What is normal? - I often operate diodes at pico amps. Under those conditions the "turn-on" voltage is nowhere near 0.7V. The Vt does describe the slope of the current vs voltage curve. For an ideal diode the current will double for every Vt increase in voltage. – Kevin White Mar 30 '19 at 01:03
  • @KevinWhite That sounds a lot like an answer. – Spehro Pefhany Mar 30 '19 at 01:43

1 Answers1

0

Given a normal silicon diode at room temperature, with abrupt junction transition from N_doped to P_doped, this table suggests the logarithm_current/linear_voltage behavior of a diode:

Idiode Vdiode

1mA 0.600 volts assued

0.1mA 0.600 - 1 * 0.058

0.01mA 0.600 - 2 * 0.058

1uA 0.600 - 3 * 0.058

0.1uA 0.600 - 4 * 0.058

0.01uA 0.600 - 5 * 0.058

1nanoA 0.600 - 6 * 0.058

10mA 0.600 + 1 * 0.058 (Maybe; if diode area is large enough to hid Rbulk)

2.718mA 0.0600 + 1 * Vthermal

1.1mA 0.600 + 0.004 millivolts (this is useful threshold for some distortion)

analogsystemsrf
  • 33,703
  • 2
  • 18
  • 46