High-Speed LED Pulsing Experiments

This page is currently being populated with past experiment data.  All future experiments will be added shortly after their completion.

The plans for Edgerton are available for anyone to build their own high-speed flash.  I designed the flash by building a prototype and doing lots of testing.  Until now, I haven’t kept an updated log of all the experiments performed on the LED’s.  Maybe some people would like to see all the details behind the designing and prototyping of the flash.  Others may be interested in designing their own high-speed LED flash.

The complete experiment logs can be found at github.com.

2019-09-16 – Light Probe Testing

Goal:

  • Test the new active light probe provided by NQTRONIX
  • Potentially desctructive test budget: 0

Equipment:

  • NQTRONIX Active Probe
  • BK Precision Oscilloscope
  • Digital multimeter
  • Rev. 2 controller prototype with MIC2171-based flyback converter

Setup:

  • Single LED strobing for [0.5, 1, 2, 4] microsecond at a capacitor voltage of 125 V and MOSFET gate voltage of 16 V
  • Light was bounced off wall (nearby ambient light sources turned off, fluorescent overhead lights still on)
  • Probe was hand-held (distance from LED to probe not measured, we are just checking how the light probe responds)
  • Oscilloscope channel 1 is connected to the active light probe

Results:

  • LED failed during 4-microsecond test
  • Massive amount of noise in signal

Comments:

  • The LED failure was not expected.  This is the first test flashing using the new control board.  The only real change is the gate voltage (increased to 16 V).  I suspect that the gate voltage increase caused the failure.
  • The noise in the signal may be caused by the poor test setup (reflecting light off the wall, probe is hand-held).  The almost-complete integrating sphere will hopefully solve this problem.
  • It is also possible that the noise is actually the signal from the LED.  It’s pretty crazy and I may need to tame it.

2019-10-06 – Light Probe Testing

Goal:

  • Test Rev. 2 controller at lower capacitor voltage
  • Potentially desctructive test budget: 1 LED

Equipment:

  • NQTRONIX Active Probe
  • BK Precision Oscilloscope
  • 55-cm integrating sphere
  • Digital multimeter
  • Rev. 2 controller prototype with MIC2171-based flyback converter

Setup:

  • Single LED strobing for 1 microsecond at various capacitor voltages (60 – 70 V) and MOSFET gate voltages (8 V, 16 V)
  • Oscilloscope channel 1 is connected to the active light probe

Results:

Comments:

2019-10-13 – Light Probe Testing

Goal:

  • Investigate relationship between [light output & noise] and [gate & capacitor voltage]
  • Potentially desctructive test budget: 1 LED

Equipment:

  • NQTRONIX Active Probe
  • BK Precision Oscilloscope
  • 55-cm integrating sphere
  • Digital multimeter
  • Rev. 3 controller prototype with MIC2171-based flyback converter

Setup:

  • Single LED strobing for 1 microsecond at various capacitor voltages (65 – 150 V) and MOSFET gate voltages (7 – 16 V)
  • Oscilloscope channel 1 is connected to the active light probe
  • Oscilloscope channel 2 is connected to the capacitor voltage (to measure voltage drop, allowing cacluclation of average current)

Results:

  • Light output -v- capacitor voltage relationship remains relatively linear between 65 V and 150 V.  There is no significant drop-off
  • Gate voltage between 8 V and 16 V does not have significant effect on light output
  • Gate voltage does have significant effect on light output noise, especially during turn-on.  Turn-on noise can be eliminated and turn-off noise can be reduced by lowering gate voltage.
  • Gate voltage below 8 V has significant detrimental effect on light output.
  • LED was not damaged during tests

Comments:

  • The new Rev. 3 prototype controller is almost identical to the Rev. 2, except that the layout is much easier to manage.
  • By coincidence, the 8 – 12 V gate voltage used by the original flash design appears to be ideal.  The Mark 2 will have a regulated gate voltage and it will likely be set somewhere in that neighbourhood.
  • The 125 V capacitor voltage limit may be revisited as higher voltages appear to greatly enhance the light output.
  • This test is limited by the fact that only one LED was used.  The next experiment will be identical, but with a complete bank of 3x LED’s.

    2019-10-14 – Light Probe Testing

    Goal:

    • Investigate relationship between [light output & noise] and [gate & capacitor voltage] with a full bank of LED’s
    • Similar to 2019-10-13 testing
    • Potentially desctructive test budget: 3 LED

    Equipment:

    • NQTRONIX Active Probe
    • BK Precision Oscilloscope
    • 55-cm integrating sphere
    • Digital multimeter
    • Rev. 3 controller prototype with MIC2171-based flyback converter

    Setup:

    • Complete bank of LED’s strobing for 1 microsecond at various capacitor voltages (65 – 150 V) and MOSFET gate voltages (7 – 16 V)
    • Oscilloscope channel 1 is connected to the active light probe
    • Oscilloscope channel 2 is connected to the capacitor voltage (to measure voltage drop, allowing cacluclation of average current)

    Results:

    Comments:

Leave a Comment