2013-12-26 High Power LED Flasher – Strobe

This is a continuation of the 2013 Dec 18 blog.  It was getting too long, and so is this blog.

Update Dec 24 – I searched for high current pulse gen schematic using Google and I didn’t come up with anything near enough to what I’m looking for.  I saw a lot of high tech and high priced equipment for very specialized applications.  I saw some stuff for high voltage generation.   I saw only a few schematics and of those only one seemed to be anything close to what I would call simple enough for the average experimenter.  That was an avalanche pulse generator.  And it can’t supply enough power to give the 10 watt LED any kind of brightness.  I think my next step will be to go through the books I have and look for something out of the decades when transistors were most popular, like in the 1970s and 80s.  Be back soon.

Update Dec 26 – Yesterday, during my encounter with some sparkling bubbly stuff [hic], I perused some old books and came up with an old schematic from the ’70s for an emergency auto flasher that uses a 6 volt lamp.  I replaced the lamp with the 10 Watt LED and built the circuit on a thin piece of plywood.  It uses a PN2222A for Q1 and a TIP32C PNP for the LED driver transistor.  Other than that, it’s a typical flasher circuit with two transistors and a 10 uF capacitor that  determines the flash rate along with the two base bias resistors.  The schematic shows the transistors as being RS-20xx, which are old Radio Shack numbers. Just about any NPN will work for Q1 and any PNP power transistor for Q2, as long as it can handle well over 1 Amp.  Alternative types are the BD138 or BD43

It runs down to 9V, but the speed is slow at low voltages and speeds up as the voltage increases.  I think that if I put a few k resistor across the LED it might help make it run at lower voltages.  I turned the PS’s current control up to maximum, which is 1 amp, but the voltmeter still dips slightly when the flash occurs, which means the power supply is going into current limit mode, so the flasher is drawing over 1 amp peak.  The high current pulse is too fast for the current meter to respond, the needle just flicks up a bit.  I put a half dozen 10000 uF capacitors across the power input, and a 180 ohm resistor between the caps and the power supply.  I adjusted the PS until the voltage across the caps was 10V, and measured the current.  Even though there were 60000 microfarads on there, the current still jumped around a bit, but I got a bit under 2.5 milliamps average current.  I can safely say that the current is 2.5 mA or less at 10V, which is the maximum rating of the capacitors.

The intensity of the flash is extremely BRIGHT!  Like, exceedingly, overwhelmingly bright.  I covered the LED with a piece of black electrical tape, and lifted the corner so the flashes came out away from my eyes.  They hit the piece of paper on the table and between the flashes the reflection off the paper left a square of dark in my eyes.  It’s bright!

Update Dec 27 – I took the power flasher along with me to the restaurant to eat with a friend, and I grabbed a 9V battery and some clip leads.  I wanted to show my friend how bright it was.  I clipped the battery on but it refused to flash.  I added the 10000 uF capacitor, and it still refused to flash.  Either the battery was low, or the flasher wouldn’t flash on 9V, which was what I found when I tried reducing the power supply voltage down below 10V.

The idea that I came up with was to replace the 9V with one of my 1.5V to 9V DC to DC converters.  The converter is perfect for this application.  The flasher’s average current is less than 3 mA at about 10V, so that comes to about 30 milliwatts.  The typical Joule Thief puts out about 66 mW, and draws about twice that from the single AA cell.  So if the converter is about 50% efficient then the current from the battery should be about half of the normal JT current.  That means the Joule Thief will easily have plenty of output to run a single power flasher, or even enough to run two of them.

This morning I drew up a schematic of a preliminary circuit.  It consists of two sections; the power flasher itself, and the Joule Thief power supply.  This afternoon I built up the Joule Thief power supply section.  It boosts the voltage to 9.5 to 10.5 volts.  It does have a quirk, though.  I used a 2200 uF 16V capacitor across the output to act as a reservoir to allow the flasher to draw a very high current for a fraction of a second.  When this large cap charges up the JT power supply shuts off and the voltage drops down, then it kicks in again, sort of like an oscillation.  I’m going to have to investigate further.


7 Responses

  1. Kirk says:

    Thats going to be one bright flasher, what will it be used for ? It would surely keep the squirles out of the attic.
    From one extream to the other yesterday 0c tonite is droping to -39c plus wind chill = -44c.

    • admin says:

      Squirrels? Attic? I had holes in my house’s vent screens and I patched them up. I was more afraid of bees than squirrels; last place I lived had bees in the wall. The exterminator came and banished them. Then every summer, something came oozing out of the wall above the door. I think it was honey from the bee hive. I washed it off the door, but every time it got hot, the stuff would start oozing out of the wall and running down the door.

      I think the best use for it would be a strobe to make stop motion — or else an insanely bright flasher. It’s bright, but the average current is only 2.5 milliamps, and I’m working on getting it to run from a single 1.5V AA cell, using a Joule Thief for the power supply.

  2. Kirk says:

    Thats very good that you can get the current draw so low. The house next door has been a home for the squirrles for many years, but the owner is not bright enough to just put a cheap white strobe light in the attic, the squirrles would leave and not come back. He also has a huge wasp nest in the attic.

    • admin says:

      The low average current indicates that the LED is on for a very short period of time, probably only a millisecond or two out of each of the 1 second flash intervals.

      Some of the flashers using a regular 5mm LED that I’ve built draw only a half milliamp or even less. The batteries in these may last more than a year.

  3. Kirk says:

    The low current draw reminded me that I saw it here http://www.talkingelectronics.com/projects/200TrCcts/200TrCcts.html#9D
    Don’t know if this would be close to your application.

    • admin says:

      That is the same circuit, with minor changes, as the one that’s the third schematic above it, which runs on 1.5V. And after experimenting with the similar flasher from bowdenshobbycircuits.info, I found that the 1.5V TE circuit with two transistors was about as good as the one on Bowden’s website, which uses three transistors. This is in my blog, just search for bowden.

      But one thing should be changed. The transistor that drives the LED is shown as a BC557, which is not capable of high currents. A much better choice is the BC327-25, 2N4403, or the PN2907A.

  4. Kirk says:

    I think this is going on my list of things to make.

Leave a Reply to Kirk Cancel reply

Your email address will not be published. Required fields are marked *

© RustyBolt.Info/wordpress
CyberChimps