LED Switching Oscillations

Discussion in 'The Projects Forum' started by Jack Bourne, Apr 21, 2013.

  1. Jack Bourne

    Thread Starter Active Member

    Apr 30, 2008
    39
    0
    Hi,
    I am currently trying to drive an LED with pulses of widths 10 to 50ns. It is a high current device (500mA drive current) except Ive encountered a problem.
    As you can see from this picture, http://i.imgur.com/TCtDdUp.png, I am having some trouble with oscillations caused by switching of the device.
    It is being driven using a MOSFET circuit as shown here http://i.imgur.com/Djzbe8r.png (it also has a 50R resitor in parallel with Vin.

    Do you have any ideas as to the cause of these oscillations? Im not entirely sure theyre optical or whether they are electrical interference.

    Any help much appriciated
    Thanks
     
  2. Jack Bourne

    Thread Starter Active Member

    Apr 30, 2008
    39
    0
    I may have found a solution.
    Searching Google for mosfet gate oscillations turned up this article:
    http://tinyurl.com/c8er42c

    In this article, the author offers a solution of including a resistor and ferrite bead in series with the gate.

    Has anyone used this solution before and if so, did you get any good results?
     
  3. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Ferrite beads on the gate leg, and sometimes the drain, is extremely common in professional designed SMPS devices.

    So it can help. :) You might also be getting measurement artifacts.

    The main thing I would ask is WHY you need to drive a 500mA LED at MHz frequencies?

    If you are using PWM to control the LED brightness then frequencies below 50kHz will be plenty.

    If on the other hand you are deliberately designing some very high speed 100 MHz optical comms etc, then there there are a lot more issues than just "the scope is picking up some ringing".
     
  4. Jack Bourne

    Thread Starter Active Member

    Apr 30, 2008
    39
    0
    This is indeed an optical communications project and yes I do need to drive it at those frequencies. This link should transmitt information at a range of 25m underwater with a beam half angle of 15degrees; this is why I need such a large power output and hence drive current

    These are not measurement artefacts, the output was from a p-i-n diode with a high gain factor.

    I believe the oscillations were in fact a combination of EMI and optical output oscillations.

    I have tried several combinations of resistors and ferrite beads. With a 1uH ferrite bead and a 4ohm resistor, there is no more ringing of the gate and drain voltage but at the price of a reduced slew rate.
    Later this week, I shall test the optical receiver with the driver running for EMI.

    Any other advice is much appreciated
     
  5. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    Thanks for providing more information. :)

    I think if you are driving 500mA in the 10's of MHz you need good RF design principles, and direct support from someone with expertise in RF power design and layouts etc would be ideal. It's definitely not a beginner hobby project.

    What kind of comms protocol or modulation will you be using? And why the need for such high frequencies? Is this intended to send video data?
     
Loading...