Page 209 - Troubleshooting Analog Circuits

Appendix E

VF VS. IF on Various Diodes

List of Diodes

A. SR306 Schottky rectifier
B. 1N87G germanium
C. HP5082-2811 Schottky
D. Big old stud rectifier
E. 3S14 rectifier
F. 1N4001 silicon rectifier
G. 1N4148 / 1N914
H. HER103 ultra-fast rectifier
I. 1N645
J. FD300 low-leakage diode
K. LM194 / LM394 VBE
L. 2N3904VBE
M. 2N3904Vo
N. LM3046VBE
0. 4N28 diode, pins 1 and 2.
P. RedLED
Q. Green, yellow, or super-bright red LED.
Notes
FD200 and FD600 are similar to the 1N4148 curve.
Transistors have collector shorted to base, for curves K, L, N.
All data at 25 degrees C.
LEDs must be in darkness for accurate low-level data.

Comments on Semi-log Plots of VF versus Log of IF

Note all the different slopes of diodes! I wouldn’t want to show such a confusing set
of plots, except life really is confusing . . .
The 1N4148s have a slope of about 115 mV per decade, compared to about 65 mV
per decade for some of the Schottky diodes, and 60 mV per decade for the transistors,
and intermediate values for other types of diodes. Nobody ever tells you about these
widely varied characteristics!
Note that the transistor VBEs all have the same slope and are better (steeper) than
most diodes. However, they are only good to about 6 V of V,,,,. (The LM394 has a
built-in base-emitter clamp diode and thus cannot be reverse-biassed.)
Note that the 2N3904’s c-b junction (curve M) has an inferior slope-worse conduc-
tance than the b-e junction. Consequently it has a higher Vf at high currents, but
worse leakage at low voltages.
Note that a red LED may have as little as 1 pA of forward current when biassed with
0.6 V forward! But the LEDs must be kept in the dark, to avoid photo-currents.
Note, I started measuring some of these Vfs with a curve tracer. I realized later that
the V+ were wrong-the curve-tracer was badly out of calibration. I black-flagged it

I96