Connect with us

Laser Diode Pulser

Discussion in 'Electronic Design' started by Al, Jul 16, 2003.

Scroll to continue with content
  1. Al

    Al Guest

    I few days ago someone put out a challange for more challanging
    questions in the news group. I responded with a request for a circuit
    that would excite a laser diode. The requirements for the diode are
    quite specific. I did get some responses.

    One of the responses led me to a circuit using an avalanch transistor.
    It is in essence the discharge of a capacitor via an avalanch transistor
    through the diode. It seems a bit uncontrolled.

    The specified maximum current for this diode is 10 Amps. The typical
    voltage at that current is 5.8V. The maximum pulse width is 180
    nanaseconds at a 1 KHz repetition rate.

    How can that capacitor discharge circuit be modified so one can ensure
    that the current never exceeds 10 amps, and very importantly, that there
    is no ringing. A curious omission from the specification is the maximum
    reverse voltage that diode can handle.

    Any comments?

  2. Al

    Al Guest

    Yes, I used the transmission line idea for a circuit used to test ICs
    for ESD (electro-static discharge) immunity. I used a vacuum relay to
    discharge specific voltages into the inputs of devices; the input was
    shunted by a 50 ohm load resistor. We would start at 500V and increment
    at perhaps 50 volts per step. It was a manual system where we discharged
    the pulse into the device and then measured the leakage current using a
    curve tracer. It much better, in my view, then using the human body
    model which consisted of a capacitor and a resistor. The square wave
    method was much more repeatable. I think I used a 10 ft length of coax
    to get a approximate 10 nS pulse.

    I did see look at the Zetex web site. Maybe I'll peruse it in a little
    more depth now.

  3. John Larkin wrote...
    Lessee, 1200V into 50 ohms, that's 24A, right? The step-recovery
    diodes I'm familiar with are low-current, low-voltage parts. How
    about giving us a few manufacturer's names and part numbers?

    - Win
  4. John Larkin

    John Larkin Guest

    Regular 1N5404-5407 parts sometimes work, or even better a
    fast-recovery diode like the FR80x family. The problem here is that
    the parts vary from batch to batch, and only some batches snap well.
    We suspect that packagers buy a box of wafers somewhere and sell them
    as anything from 5400's to 5408's, and they're not dependably the same
    silicon every time. To get our 1200 volts, we drive the diode through
    an inductor: foreward bias at around +50 volts (yes!) for roughly 100
    ns, til the foreward current hits maybe 40 amps, then switch the drive
    to -400 to reverse the current until it snaps and twangs. Fun!
    Impressive EMI!

    What's apparently happening is that, with the short foreward drive,
    the charges stay close to/within the center "I" region of the diode,
    so the reverse recovery is very local and fast. The Russian researcher
    Grehkov discovered this effect, as well as an even faster reverse
    breakdown mode, both in cheap power diodes. HP's stunning 1430
    sampling scope head, 12 GHz in roughly 1966?, did the
    short-foreward-drive trick on regular low-voltage SRDs to speed them
    up, too. McKwen did a terrifying pulser using 1N4008's as a shock
    line, driving another one in Grehkov's plasma avalanche mode.

    We've found a batch of FR804's that snap pretty good - bought the
    whole reel from Mouser - but we've since found a better, very
    consistant, heat-sinkable snapper, but that's not public for now.

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