Connect with us

Tunnel diode rectification

Discussion in 'Electronic Components' started by Theo Markettos, Mar 7, 2005.

Scroll to continue with content
  1. Does anyone still make tunnel diodes? I'm interested in having a play with
    them. Particularly I'm interested in the fact that they pass a (relatively)
    high current at low forward voltage, and in the forward direction the
    forward current never goes to zero. Effectively I'd like to use it as a
    rectifier with no forward voltage - where a Schottky won't do. It doesn't
    matter that it has a low reverse breakdown voltage or valley point, as long
    as the valley floor is above zero.

    Is there a more modern device that will do this?

    Thanks
    Theo
     
  2. Tom Becker

    Tom Becker Guest

    ... a rectifier with no forward voltage...

    With a conventional diode in its feedback loop, I believe, an opamp can
    simulate a perfect rectifer. Would that do?
     
  3. Fred Abse

    Fred Abse Guest

    If you're prepared to pay megabucks, yes. Sometimes the turn up at
    hamfests and swap meets.


    I'm interested in having a play with
    A few milliamps, usually
    The valley floor has to be above zero, otherwise the device would contain
    a voltage source.

    "Reverse breakdown" isn't really a property of tunnel diodes, they conduct
    in the reverse direction more or less from zero.

    Tunnel diodes don't rectify, in the sense that I take you to mean the
    term.

    Fun devices, though. Probably the fastest switching device there is. Quite
    easy to generate pulses with picosecond rise times. If you're *really*
    clever, you can amplify with them.

    A word of warning: Don't try to test a tunnel diode with an ohmmeter. More
    of them have been damaged doing that than ever failed in service, and even
    if the device survives, it'll tell you nothing, since they conduct in both
    directions.
     
  4. mike

    mike Guest

    There's a variant of the tunnel diode called a back or backward diode.
    Think of it as a TD with extremely low peak current. Hook it up
    backwards and it makes a low forward-drop rectifier with a reverse
    breakdown of half a volt or so.

    Unless you need a fast rise pulse, stay away from tunnel diodes.
    Alternate technology has overtaken them for most other uses.
    mike

    --
    Return address is VALID but some sites block emails
    with links. Delete this sig when replying.
    ..
    Wanted, PCMCIA SCSI Card for HP m820 CDRW.
    FS 500MHz Tek DSOscilloscope TDS540 Make Offer
    Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
    Bunch of stuff For Sale and Wanted at the link below.
    MAKE THE OBVIOUS CHANGES TO THE LINK
    ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/
     
  5. Yes, a regular power FET will act as a rectifier with lower V drop than
    a Schottky. See the intro on P.6.
    http://www.onsemi.com/pub/Collateral/NIS6111-D.PDF1
     
  6. When I experimented with them, they seemed to have one nasty habit.
    They took every bit of stray inductance and capacitance and turned them
    into a delay line. Instead of oscillating with a sine wave, they
    decided that they wanted to do square waves instead. :-/
     
  7. Fred Abse

    Fred Abse Guest

    Yep. They do that. If you want sine waves, you need a high Q tuned
    circuit, and play with the bias so as it's on the middle of the negative
    slope (the bit you can't see on a curve tracer, 'cos it's too fast).
    RC circuits just make them switch.

    Cavities go very well.
     
  8. Fred Abse

    Fred Abse Guest

    Seen (aka burnt out) a few of those in mixers in my time :)
     
  9. True, good point.
    Am I misreading something from the graph at:
    http://www.americanmicrosemi.com/tutorials/tunneldiode.htm

    which seems to suggest that they're basically like a Zener except the
    tunnelling current means they have a lower resistance at low applied
    voltages than at medium ones? I suppose the problem is lack of scale on the
    left/lower parts of the graph. It does seem to suggest that they're
    actually better at conducting in the reverse direction than the forward
    direction - is this true? I can't seem to find any datasheets on these
    diodes.

    My interest is for a crystal radio, where the forward voltage drop of a
    silicon, Schottky or germanium diode means you lose a lot of signal, but
    power isn't available. I was wondering whether a tunnel diode might be more
    efficient? Are there any other types of diode that might do the trick
    instead?

    Thanks
    Theo
     
  10. mike

    mike Guest

    How about a Germanium point-contact diode?
    I've read reports that you can get the best performance with a real
    Galena Xtal and real cat whisker...as long as you're willing to futz
    with it.
    You often get better results if the Subject line reflects your actual
    objectives. Unlikely that a cat-whisker expert is gonna read a tunnel
    diode post.
    mike

    --
    Return address is VALID but some sites block emails
    with links. Delete this sig when replying.
    ..
    Wanted, PCMCIA SCSI Card for HP m820 CDRW.
    FS 500MHz Tek DSOscilloscope TDS540 Make Offer
    Wanted, 12.1" LCD for Gateway Solo 5300. Samsung LT121SU-121
    Bunch of stuff For Sale and Wanted at the link below.
    MAKE THE OBVIOUS CHANGES TO THE LINK
    ht<removethis>tp://www.geocities.com/SiliconValley/Monitor/4710/
     
  11. You're supposed to do it like this.
    http://www.oldradioworld.de/gollum/dt.htm

    These use a separate diode to rectify and filter the RF, then power the
    amplifier with it. So you are getting free power from the RF in the
    air. Just remember to use a long wire, the longer the better.
    Germanium diodes will rectify with a drop of only a quarter volt.

    Or you can use a solar cell that puts out a few volts to furnish the
    power to amplify the RF and/or AF. You can also buy one of the radios
    that has a solar cell and crank built in, and will run off the crank
    generator for several minutes or off the solar cells if it's sunny.
     
  12. Fred Abse

    Fred Abse Guest

    I think you're looking at the superimposed "conventional" characteristic,
    and implying reverse blocking. Tunnel diodes conduct in reverse more or
    less from zero. It isn't a very good idea to apply reverse bias as more
    than a few milliamps will destroy them.

    Tunnel diodes exhibit useful properties (ie. negative resistance) in the
    forward direction only.
    A Zener has a defined reverse breakdown voltage and is generally used
    reverse-biased. A tunnel diode has a reverse breakdown voltage of
    approximately zero.

    except the
    Yes, but they're never used in the reverse direction, nothing useful
    there. It's the negative resistance portion between the peak and valley
    that makes them useful. With a resistive load, increasing forward voltage
    through the peak point will result in a current step with picosecond rise
    time. Alternatively, you can use the negative resistance to overcome the
    losses of a tuned circuit, and oscillation will occur. That technique is
    usable at many GHz.
    There was *some* data in the old GE "Transistor Manual", if you can find
    a copy. Most information was published in the 1960s, and is probably out
    of print, now. They were used extensively as trigger switches in high-end
    oscilloscopes by Tektronix and HP.
    It would certainly oscillate, but I guess a transmitter isn't what you
    want :)

    Are there any other types of diode that might do the
    There's a thing called the backward diode, which is similar in principle
    to the tunnel diode, but which is designed as a (microwave) detector.
    They're probably as difficult to find as pork chops in a synagogue :)

    Is it worth going to all that trouble for a crystal receiver? Selectivity
    is just too poor for today's crowded spectrum.
     
  13. That's a good point. And it takes a decent long wire antenna, which is
    usually larger than most dwellings can accommodate.
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-