[Avtech home page] (where this FAQ is located)

FAQ = "Frequently Asked Questions"
Contributors wanted! Enquire within. -- April 20, 1996

The Unusual Diode FAQ - v3.2

The photo shows a very large 35 kV, 55 Amp rectifier, circa 1960. Not your average diode.

Maintained by Michael J. Chudobiak, mjc@avtechpulse.com at Avtech Electrosystems Ltd..
My PGP public encryption key is available here. Please use it.

This page has been accessed 1570 times since Apr 03/96.

The home page of this FAQ is http://www.avtechpulse.com/faq.html


I. Introduction and CALL FOR VOLUNTEERS

Welcome to the fun-filled, action-packed Wide World O' Diodes. This FAQ is intended as a helpful tool for finding unusual diodes, such as germanium diodes or step recovery diodes. If you have any comments, suggestions, additions, contributions, PLEASE email me. Especially appreciated would be sections on laser diodes and various microwave diodes.

This FAQ used to be posted to sci.electronics and sci.engr.semiconductors around the first of each month, but it's grown to large for my newsreader, so it's staying exclusively on the WWW. It is stored at on the World-Wide Web at http://www.avtechpulse.com/faq.html

This FAQ is maintained by Michael J. Chudobiak, mjc@avtechpulse.com at Avtech Electrosystems Ltd.. My PGP public encryption key is available here. Please use it.

Copyright 1996, Michael J. Chudobiak


I don't know everything about all diodes, so VOLUNTEER CONTRIBUTORS would be greatly appreciated. If you see a type of diode that isn't included in here, and should be, feel free to write up a section for it, following the style of the existing sections. That is, include one or two paragraphs on what the diode is and how it works, and then list the manufacturers that actually make them. Please include their address, phone and fax numbers.

For instance, we could use sections on:

If you know any good books on diodes, let me know.

If you know something about diode manufacturer corporate history, and it isn't covered in the "What ever happened to ...." section, let me know. For instance, what happened to Mullard?

If you are a manufacturer of diodes, I'd love to have a copy of your data book. My address is at the end of this FAQ.

Anyways, this is a living document created in the public interest, so comments, ideas, and especially written contributions would be most appreciated.

Oh, if you have any good diode-related WWW links, especially to the manufacturers listed below, please let me know. Just email me at mjc@avtechpulse.com. My PGP public encryption key is available here. Please use it.

II. Changes Made Since Previous Versions

since v. 1.5

since v. 1.6

since v.1.7

since v.1.8

since v.2.0

since v.2.1

since v.2.2

since v.2.3

since v.3.0

Since v.3.0

III. Hot Links to Other Useful Diode WWW Sites

The study of diodes isn't a hugely popular area, and manufacturers have been a little slow to put useful info on the Web, so this hot-list is a little sparse at the moment. Please let me know of any useful diode-related links!

IV. Who makes / What are ....

IV.1 - Germanium diodes?

Germanium diodes find some use since Ge has a much smaller bandgap energy than Si, producing lower forward voltages. However, this smaller bandgap also makes Ge less useful at higher temperatures due to a higher leakage current. Ge diodes have been largely replaced by Si Schottky diodes for applications below 200V, and GaAs Schottky diodes above 200V.

IV.2 - Selenium and Copper Oxide diodes?

Cuprous-oxide-on-copper rectifiers were first used for the rectification of large currents in 1924. Selenium rectifiers were used extensively before Si power technology was sufficiently developed, due to its relatively simple manufacturing. Selenium was also used in solar cells and photoresistors. Selenium typically has a knee voltage of 0.5V, and copper oxide has a knee of 0.2V. Both have relatively linear forward I-V curves. Apparently the large-area sandwich structure of selenium rectifiers provides for excellent heat sinking capabilities. Also, they are supposed to be quite robust as far as tolerating excessive currents is concerned.

IV.3 - Tunnel diodes, backward diodes?

Tunnel diodes exhibit a current "dip" in their forward I-V characteristics. That is, for a certain range of forward voltages the current actually falls, instead of increasing. This creates a negative differential resistance, making it useful in oscillators and switching circuits. The underlying quantum-mechanical tunneling effect is extremely fast. Leo Esaki, who developed the tunnel diode was awarded a Nobel Prize in Physics for his efforts.

A backward diode is a diode with an extremely low breakdown voltage, causing it to conductor better in the reverse direction than in the forward direction. Backward diodes are similar in structure to tunnel diodes and may show negative resistance, in which case they are usually called tunnel diodes. Backward diodes are also known as "Uni Tunnel Diodes".

Here's what Gabriel Paubert told me about backward diodes:

Some manufacturers:

IV.4 - Step Recovery Diodes (SRDs)?

When diodes are switched from forward bias to reverse bias, the diode still conducts for a very short period, since some charge is left in the device. Normal diodes remove this leftover charge very slowly, but SRDs are optimized so that the charge is removed rapidly, and the reverse conduction stops very abruptly. This abrupt change can be used to create very fast switching pulses, or to generate harmonics of the switching signal.

IV.5 - Current Limiting diodes?

Constant current two-terminal devices can be made by shorting the gate and the source of a JFET together. When the FET is forward biased, this results in a nearly constant current for voltages ranging from roughly 2V up to 300V (or the breakdown voltage of the device in question). In reverse bias, this kind of constant current device conducts as a junction diode (so one can oppose two such devices in series to regulate AC current).

Siliconix makes two-lead FET current-limiting diodes ranging from 0.24 mA (J500) through 4.7 mA (J511) in plastic packages, and from 1.6 mA (CR160) through 4.7 mA (CR470) in metal TO-18 packages.

National Semiconductor makes a three-terminal adjustable device, the LM134/LM234/LM334, that acts as a resistor-programmed current source diode, analogous to the 'programmable Zener' TL431. Adjustment range is 1.0 uA to 10 mA, and voltage compliance is from 1V to 40V (or 30V for some versions). The current is slightly temperature-dependent (this may be useful, or can be eliminated with a diode added to the adjustment resistor).

Motorola used to make current-limiting diodes, MCL1300 series, with 75V operating range and 0.5 mA to 4 mA current. I think they've stopped offering these.

IV.6 - Automotive diodes?

IV.7 - Replacement diodes?

IV.8 - Noise Diodes?

The avalanche breakdown process in diodes is inherently noisy (or random). Some diodes are designed to have a very well controlled avalanche breakdown characteristic; these can be used as white noise generators. If you aren't looking for something particularly fancy, a normal avalanche zener diode (not a tunneling zener diode) will work quite well as a noise source when biased in breakdown.

Personally, I wondered what noise diodes were used for.

This is what Marshall Jose, Marshall.Jose@jhuapl.edu told me:

This is what Bob Underwood, bobu@msm.com, told me:

And this is what Gabriel Paubert, paubert@iram.es, told me:

Here are some manufacturers:

IV.9 - Very High Voltage Diodes?

IV.10 - GaAs Schottky Power Diodes?

Gallium arsenide has a higher bandgap energy than silicon, so Schottky diodes made with GaAs will have higher breakdown voltages, lower leakage currents, and a larger temperature range than silicon Schottky barriers. However, GaAs diodes will also have a higher forward voltage, which results in a tradeoff. GaAs also has much higher electron mobilities than silicon, which will somewhat offset the higher Vf. The Vf for GaAs Schottky diodes becomes comparable to silicon for Vbr = 200 V, so silicon is used mostly below 200V, and GaAs is being introduced for high-voltage devices.

IV.11 - Transient Suppressor Diodes?

Suppressor diodes are used in combination with gas arresters and varistors to protect sensitive systems from overvoltages. They are a special kind of zener diode designed to withstand high pulse powers. This ability is achieved by a low thermal resistance and a large junction cross section. Since the large cross section causes a high parasitic capacitance they cannot be used to protect RF-systems. There are also bipolar suppressor diodes available which bear two antiserial suppressor diodes on one chip.

This section was contributed largely by Kai Borgeest.

For more info, check out Microsemi's Transient Voltage Suppressor application notes


IV.12 - Avalanche Photodiodes (APD's)?

Yuan Jiang, yjj@eng.umd.edu, says:

interesting note added by Dave Kirkby, davek@medphys.ucl.ac.uk :

Some good basic info is also available here.

Si APDs:

Si APDs packaged with receiver circuits:

Ge APDs:


IV.13 - Microwave PIN Diodes?

- contributed by Marshall Jose, Marshall.Jose@jhuapl.edu

PIN diodes are useful for switching and attenuating RF (radio frequency) signals. Basically, between the P- and N-doped regions of the diode is an undoped region referred to as "intrinsic" (hence the I in "PIN"). When a forward DC bias is applied the diode, a large number of holes and electrons are created in the I region, allowing forward conduction. If the bias is suddenly removed, these charge carriers will take some time to recombine and thus stop the conduction of current. This amount of time is quite a bit longer than the time a normal PN diode takes to cease conduction.

All this means that while the PIN diode is conducting forward bias current, it will conduct a high-frequency signal superimposed on the bias current, too -- even a large signal which would cause a momentary reversal of diode current! Furthermore, the high-frequency signal won't be much distorted. The net effect of the diode at high frequencies is that of a variable resistor, whose resistance decreases as the bias current increases.

(Note that the phrase "PIN diode" can also refer to a range of power diodes with a very wide near-intrinsic region, which supports a high breakdown voltage. These are not microwave diodes.)

IV.14 - Zener Diodes?

Every pn junction will break down in reverse bias if enough voltage is applied. A typical medium current discrete bipolar transistor has a collector-base junction which is doped fairly lightly (on the collector side) and will break down at a reasonably high voltage (perhaps 30V to 50V). This type of breakdown is called avalanche breakdown. It happens when thermally generated carriers in the depletion layer are accelerated by the electric field therein. If the field is high enough, the carriers are accelerated to high energies and they become capable of ionizing Si atoms in the depletion layer. The charge carriers from these secondary ionizations are in turn accelerated by the same electric field, and can cause additional chain-reaction ionization. The process resembles an avalanche (eg. on a snow covered mountain) hence the term. All junctions will exhibit avalance breakdown with sufficient reverse bias.

The emitter-base junction on the other hand, is generally very heavily doped. Before the field becomes high enough to cause avalanching, the junction will breakdown by another mechanism, called band-to-band tunneling, or Zener breakdown. In this case, the depletion layer is thin because of the much heavier doping levels. As the reverse bias is applied, a situation occurs where the conduction band on the n-side aligns and then drops below the valence band on the p-side. The exact voltage at which this is achieved depends upon the doping. When it happens, conditions are right for electrons to tunnel through the barrier. They suddenly appear on the other side of the junction, if the depletion layer is thin enough. This is called direct tunneling. There is another type of tunneling called indirect tunneling, or trap-assisted tunneling in which the electron tunnels into an intermediate trap level (or series of levels) before making it all the way throught the barrier. This one is less common in standard pn junction devices.

A "Zener diode" is made to break down at a specific voltage with a sharp reproducible characteristic. The diodes are designed to conduct the breakdown current evenly, and nondestructively. The breakdown mechanism may be avalanche breakdown or Zener breakdown, or a mixture of the two. If the diode breaks down at voltages of about 5.6V at room temperature, the two mechanisms are in equal measure. If the breakdown voltage is higher, the avalanche process dominates, and if lower, the tunneling or Zener mechanism dominates.

The temperature coefficient of the avalanche mechanism is positive, that is, at higher temperature, the avalanche breakdown voltage increases. The temperature coefficient of the tunneling breakdown is negative. At just the right doping level, the breakdown is a mixture of the two types in just the right proportion that the temperature coefficients cancel. This voltage is around 6 V.

For more, info, check out Microsemi's Zeners and Zero TC Reference Diodes application notes.

Practically everyone makes zener diodes, so I'll only list two:

IV.15 - Electron-Emitting Diodes?

from Paul Woods (paulw@hpcvnq08.cv.hp.com) at HP:

Have you heard of electron-emitting diodes? They are not readily available, in fact, I have only seen two or three references to them in physics literature. The most notable reference was in Philips Technical Review in 1987. Two researchers had made a diode with a very thin, heavily doped n-layer. When the diode was reverse-biased to the point of avalanche breakdown, a small fraction of the avalache electrons actually shot through the n-region and into a vacuum. They made this the e-beam source in a CRT and it worked pretty well. I am surprised that it has not, to my knowledge, made it out of the laboratory. It did have some unique requirements that probably made it expensive to produce. For one thing, the doping profile was very abrupt and required MBE which is slow and expensive. Also, to improve emission efficiency they reduced the surface work function by coating the emitting surface with Cs. I have heard that Cs is nasty to work with.

Curious things, aren't they? Sounds like a good thesis for someone. Perhaps they could be made practical using rapid thermal CVD instead of MBE.

Since nobody makes them commercially, here are some references instead:

IV.16 - Blue-Green Laser Diodes?

Everybody's invited to visit the Unofficial WWW Server for Blue-Green Diode Lasers!

IV.17 - Schottky Diodes?

No point writing something if you can just scam somebody's app note. Click here to read Microsemi's app note entitled, 401- Introduction to Schottky Rectifiers.

Everybody makes Schottky diodes.

IV.18 - Exotic-Semiconductor Diodes?

Here's a link to the NASA Lewis Research Center SiC (silicon carbide) page. Silicon carbide is just great for high temperature applications - up to 600 C, anyways. Silicon only goes up to 350 C or so.

Click here for info on III-V compound-semiconductor devices and research.

Here's some nice intro material on Mercury cadmium telluride (HgCdTe) photodiodes.

Here's what Fred Olschner, 72142.365@compuserve.com, told me about the availability of other wide bandgap diodes:

IV.19 - PIN silicon photodiodes?

PIN silicon photodiodes are quantum detectors sensitive to light from UV (200 nm) to near IR (1150 nm), gamma radiation, X-rays and charged particles. Photodiodes operate by the absorption of photons or charged particles which generates a flow of current in an external circuit. Photodiodes can be used to detect the presence or absence of small quantities of light and can be calibrated to measure the intensity of light extremely accurately. They can also be used for optical position sensing to measure displacement, angle, centering, surface uniformity and distance. See X-ray diodes for explanation of different operating modes.

IV.20 - PIN silicon X-ray diodes?

PIN Silicon X-ray diodes are detectors sensitive to X-rays, gamma-radiation and charged particles (alpha- and beta-particles). X-ray diodes operate by absorbing photons or charged particles. X-ray diodes have many similarities to photodiodes, but they are optimized for X-rays and have a suppressed sensitivity to light. X-ray diodes can be used with wide range of energies (from a couple of keV's to approximately hundred keV's).

X-ray diodes and photodiodes can be operated in current or charge (pulse) modes. In current mode the output current is measured directly from the detector. Current mode is used typically when event rates are very high. In current mode applications the output current is proportional to the intensity of the incident radiation with high accuracy. In pulse mode the single hits are transformed into pulses and then recorded.

X-ray diodes and photodiodes have also two different modes of operation depending on the bias voltage supplied to the detector. In photoconductive mode the diode is operated with high bias voltage. This mode is used for example in X-ray exposure and dose control applications. In photovoltaic mode no bias voltage is supplied to the detector. This mode is used in most applications including spectral measurement.

IV.21 - Silicon LEDs?

Light emission from silicon is quite a challenging prospect, since silicon is an indirect bandgap material (i.e. when an electron crosses the bandgap and emits a photon, it must change its energy level, obviously, as well as its momentum. A direct bandgap material does not require a change in momentum. This means that photons are much more likely to be emitted in a direct bandgap material.)

However, people are founding ways around this problem. The following text is from the UK's Defense Research Agency web site:

V. Info wanted for these diodes ....

I've been asked who makes the types of diodes listed below. If you know, please email me at mjc@avtechpulse.com .

VI. What ever happened to ....

I'm not 100% sure about all of these changes, so please feel free to correct me.

VII. What are some good books on diodes?

I've included Library of Congress catalog numbers on most books. That should make them easier to find.

Good undergraduate-level texts on the physics of diodes:

Graduate-level texts on diodes:

Graduate-level texts on Diode Switching:

A modern graduate-level text on diodes and transistors:

Some excellent and useful books on the physics of power diodes:

Zener Diodes:

Good introduction to power diode applications and circuits:

Other broader-range books that have been recommended to me:

VIII. Acknowledgements, and Who To Contact

If you CONTRIBUTED, your name could be here! Instant FAME! Unstoppable career advancement!

The maintainer of this FAQ is Michael J. Chudobiak, mjc@avtechpulse.com. My PGP public encryption key is available here. Please use it.

This is not an official document of Avtech Electrosystems Ltd, I post it personally. Just to be clear, Avtech does not sell diodes, but we do sell high-speed pulse generators, pulsed constant current sources, pulsed laser diode drivers, and other test instruments.

Feel free to send me contributions. Please E-mail them to me in plain text or HTML.

Yuan Jiang, yjj@eng.umd.edu, wrote the section on avalanche photodiodes.

Dr. Barry L. Ornitz, ornitz@emngw1.emn.com, pointed out that Custom Components makes tunnel diodes.

Kai Borgeest, Borgeest@tu-harburg.d400.de, recommended the book on Zener diodes, and contributed the section on transient suppressor diodes.

David Gillooly, mfield@ix.netcom.com, pointed out that GAD Semiconductor makes GaAs power rectifiers.

John Scarpulla, tjohns2@tus.ssi1.com, recommended several useful textbooks, and provided most of the zener diode section.

Marshall Jose, Marshall.Jose@jhuapl.edu contributed the section on PIN diodes and the explanation of what noise diodes are for.

Dave Kirkby, davek@medphys.ucl.ac.uk contributed the bit about APD modulation.

Paul Woods, paulw@hpcvnq08.cv.hp.com, contributed the section on electron- emitting diodes.

Bob Underwood, bobu@msm.com, added some comments to the noise diode and transient suppressor sections.

John Whitmore, whit@hipress.phys.washington.edu expanded the current-limiting diode section.

Gabriel Paubert, paubert@iram.es, added some comments to the noise diode and backward diode sections.

Fred Olschner, 72142.365@compuserve.com, added a section on wide bandgap diodes.

Jussi Koskinen, jvkoskin@snakemail.hut.fi, added the sections on PIN silicon photodiodes and PIN silicon X-ray diodes.

Michael J. Chudobiak

at Avtech Electrosystems Ltd..


---  Avtech Electrosystems Ltd.  ---------------------------- since 1975 ----

PO Box 265           ph:  1-800-265-6681 or 613-226-5772      Box 5120 Stn. F
Ogdensburg, NY       fax: 1-800-561-1970 or 613-226-2802      Ottawa, Ontario
USA 13669-0265           email: info@avtechpulse.com          Canada  K2C 3H4

Nanosecond Waveform Generators for general purpose, R&D and OEM applications

Pulse Generators - Laser Diode Drivers - Sample and Hold - Pulse Transformers
Impulse Generators - Monocycle Generators - Pulse Amplifiers - Accessories

(C) Copyright 1996, Michael J. Chudobiak