# Diode reverse-bias resistivity?

Discussion in 'Electronic Basics' started by eromlignod, Jan 17, 2008.

1. ### eromlignodGuest

Hi guys:

I'm an ME, so I have a basic knowledge of semiconductors, but not as
in-depth as you EE's.

When a diode is connected in reverse (the direction where current
doesn't flow), does it behave like an insulator or like a
semiconductor? In other words, does it have an extremely high
resistivity, like a dielectric, or does it just have a high
resistivity and pass some very small amount of current, like a
semiconductor?

I'm talking about at voltages below breakdown.

Thanks for any replies.

Don

2. ### Greg NeillGuest

Check the datasheet specs for the terms Leakage Current
or Peak Reverse Current. Typically on the order of a
few microamps for small silicon diodes in the region
before breakdown. This can be modeled as a high
resistance of tens of megohms.

A sample data sheet:
http://www.diodes.com/datasheets/ds28002.pdf

Beware of another property, junction capacitance, which
can be on the order of several picofarads. This can
permit AC components of a signal to leak through even
when reverse biased.

3. ### John PopelishGuest

They pass a current called the saturation current that can
be very small, depending on the materials and doping of the
junction. But for some, especially high current or low
voltage devices this reverse leakage can be in the milliamps
or even hundreds of milliamps. Typical leakage for 1 amp
silicon PN junctions is in microamps.
For each particular device, you find this information on the
data sheet, which sometimes just lists a maximum value.

Here are a couple examples with graphs of reverse leakage
versus voltage and temperature. Not that below breakdown
the current is less than proportional to the reverse voltage
(not resistive that follows Ohm's law, but closer to a
constant current).

figure 5 on this one:
http://www.vishay.com/docs/88503/1n4001.pdf

figure 2 in this one:
http://www.onsemi.com/pub/Collateral/MUR8100E-D.PDF

much higher leakage for Schottky low voltage type, figure 2:

4. ### eromlignodGuest

Great Info. Thanks for the replies.

Should I assume that a transistor in its "off" state would exhibit
similar behaviour?

Don

5. ### John PopelishGuest

For sufficiently vague concepts of "similar", yes. There is
definitely a reverse biased diode (collector to base
junction) leakage current component that dominates the
character of the leakage if the transistor has a reverse
biased base emitter junction. If the transistor has a small
positive base emitter bias it might still be considered to
be off, but an additional leakage current component (gain

The main similarity is that the current is less than
proportional to the voltage over a considerable voltage
range, till breakdown is approached. Then it takes off and
becomes more than proportional to voltage.

6. ### John LarkinGuest

Typically a nearly constant current, with a much less than resistive
slope, as JP states. Schottkies are leakier and more "resistive" than
silicon PN junctions. The reverse leakage current typically doubles
with every 10 degree C temp increase.

It can vary wildly between "identical" devices, especially power
diodes.

Some glass diodes are photosensitive, too!

John

7. ### JoergGuest

When I was a kid that discovery made me jump for joy. No more plunking
down more than a buck for an LDR. Just take a diode or transistor from
the TV scrap heap, scrape the black paint off the glass body, and bingo!

8. ### whit3rdGuest

Oddly, all (four, I think) answers (in other posts) are correct.

(1) If you don't have sub-picoamp current meters, a good
quality instrument diode is an insulator.

(2) If you DO have sub-picoamp current meters, a good
quality instrument diode leaks nearly constant 'satuation current'.
The diode equation scales all diodes into a simple function,
and the saturation current is the scale factor, I_s

(3) if your diode has significant surface area, and has
aged and/or gotten 'dirty' there can be large leakage
currents on the exterior crust of the semiconductor, not
really covered by the diode equation at all, and the manufacturer
has some data on how bad this can be for his design and
with his chosen materials.

(4) It varies from device to device (i.e. a rectifier is 'good enough'
if the leakage is a milliamp, but few are that bad, and NONE
are that bad straight from the factory).