My solution would be to use an Arduino microprocessor board like the Uno..
http://www.sainsmart.com/arduino-co...ent-board-compatible-with-arduino-uno-r3.html

I only lean that way because I haven't gotten into the PIC side of the world yet and know that the Arduino language is very simple to learn/do what you want here..

Couple that with a mosfet to switch the led on/off and you should be good to go..

 

Thanks guys, this has been a great help. Wayne, I'm obviously not doing real science nor do I care too much about total accuracy, the effect I'm interested in- very power efficient photosynthesis- would be so pronounced that it would leave a huge margin for error. So that blog entry has been most informative.

mcgyvr, I will certainly look into Arduino would you recommend a particular beginners resource?

 

Here is a very simple way to do it with a PIC micro;

  // C code to make exactly 5kHz, exactly 1% duty (PIC xtal = 4MHz)
  PR2 = (200-1);     // set TMR2 period to 200 uS
  while(1)           // loop and make the freq
  {
    if(PIR1.TMR2IF)       // if TMR2 rolled
    {
      PIR1.TMR2IF = 0;    // clear roll flag
      PORTB = 0x01;       // RB0 is output pin, now HI
      PORTB = 0x00;       // RB0 now LO again (was HI for 2uS)
    }
  }

You need to add a couple of lines of code to suit any PIC you choose, to setup port pins direction and enable TMR2.

 

Surprise-surprise- I did not think the LED would be visable, mine is yellow 1 W & about as bright as a 20 mA LED; @ 400 mA, 2μs on, 198μs off. Using 555 with 6 V battery, 10Ω current limiting R, independent ON-OFF pots, 2K & 100K, 1N914 diodes, cap .0033 μF, FET, 3055V N ch.

 

Decided to ink in sketch. IC TI 555P, MOSFET Motorola sur. mt. 3055V, 60V 5A?? N ch. Operation seems rock solid. Built on solderless BB. Battery Li Fe PO4.

 

Any chance you could scope the waveform? I thought it might be tough to get square corners (not that it matters) at that frequency without the drive transistors on the MOSFET gate as shown in Bill's circuit which I linked to. If the OP can do a reasonable job with just a 555, all the better.

 

I think you can do it with a CMOS 555 like shown. A couple of questions... Some for me and some for you. The way I read the article there is no power savings but the magic is how fast the plant can respond to light and store the excess. Is that how you see it?
The LED you posted doesn't spec peak current only continuous so I assumed it is similar to others I found with a spec, so I set it up for about 800 ma.. I dawns on me that with a 1% duty cycle the average light is very small. Does this mean you will use several LEDs?

"

When all of the light to make the equivalent of 50 pmol photons mp2 s-' was provided during
1.5 ps pulses of 5000 pmol photons mP2 s-' followed by 148.5 ps dark periods, photosynthesis was the same​

as in continuous 50 pmol photons mP2 s-'."

 

No scope camera, RE post 25, but sides are quite straight, at pin 3 there is a 1/2 V hook on top of leading edge, trailing edge falls straight down. V waveform of FET & LED quite clean except for .2μ s, 4V peak of damped oscillation on trailing edge.

 

Thanks Bernard, sounds good enough to me. Trusty old 555 comes through again.

 

A logic level N ch. MOSFET would be better, the MTD3055V just marginal at 6V gate, but it was already on BB as part of DC-DC boost experiment.

 

Thanks Bernard, sounds good enough to me. Trusty old 555 comes through again.

Darn. Missed again.

 

Yes and no.

But now I'm wondering about those gate drive transistors in Bill's figure 10.4. If this circuit doesn't need them, what would? Maybe a bigger (or paralleled) MOSFET with a bigger capacitance?

 

That's a nice FET. Low gate charge. I added the push pull driver because I was worried about trying to run a standard 555 with such a short pulse and the CMOS 555 can't drive to much.

 

Yes, the "surprise" I guess is that the simplified version works so well, certainly fine for the OP's application. I guess if the current and/or frequency went up, it would start to suffer.