Author Topic: Any electrical circuit wizards out there? Voltage regulator to drive an LED  (Read 1246 times)

CNYCacher

  • friend
  • Senior Member
  • ***
  • Posts: 4,438
So, I have a need to create a circuit that will drive an LED very dimly but for a LONG time on one battery.  By dimly I mean it should be able to be seen only in pitch blackness from 50-100 feet away, and by a long time I mean measured in months or years if possible.

I experimented with some simple circuits using a large resistor 20k ohm, and a 50k ohm 15-turn adjustable resitor, and an LED with a 3-volt CR123 lithium battery.  This worked well for setting the dimness factor way down, and was adjustable, but the simpleness of the circuit proved to be the downfall of it, as the LED weakened along with the battery running down, and I need the brightness to be constant once I set it to my liking.

So being a hobby-level circuit maker and not formally trained in any of this, I surmised that what I want is a voltage regulator of some sort which outputted a set voltage, lower that the 3-volts provided by the battery, and yet high enough to drive the LED.  My thinking is that then the LED will hold a constant brightness until the battery dies down to below the voltage regulator output.  Lithium cells having a good pattern of holding their output for as long as possible before rapidly dying out, I think this will do very well.

I found THIS gizmo which appears that it will meet my needs, if in fact my needs are as I imagine them to be.

How I think it will work is this:  I hook the battery up to the ground and Vin of the chip.  I hook the LED with large adjustable resistor inline to the Cout and Vout of the chip.  I adjust the LED brightness with the resistor until I like where it is, and let the thing run along.  As the battery eventually begins to die out, its voltage will drop lower and lower without affecting the brightness of the LED, until the battery output drops below 1.9 volts, at which point the whole thing may shutoff or whatever, and hopefully this point will be a long time in the future.
On two occasions, I have been asked [by members of Parliament], "Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?" I am not able to rightly apprehend the kind of confusion of ideas that could provoke such a question.
Charles Babbage

CNYCacher

  • friend
  • Senior Member
  • ***
  • Posts: 4,438
So what's the deal. . . Am I totally off?  Do you think that this will work?
On two occasions, I have been asked [by members of Parliament], "Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?" I am not able to rightly apprehend the kind of confusion of ideas that could provoke such a question.
Charles Babbage

bedlamite

  • Hold my beer and watch this!
  • friend
  • Senior Member
  • ***
  • Posts: 9,829
  • Ack! PLBTTPHBT!
CR123A batteries are about 1300 mAh, so to make it last a month, you would need a drain of less than 2ma, which might not be enough for a constant light. Also, 123A's are much better at high current drain.  I would consider a C or D cell alkaline hooked up to a flashing circuit
A plan is just a list of things that doesn't happen.
Is defenestration possible through the overton window?

CNYCacher

  • friend
  • Senior Member
  • ***
  • Posts: 4,438
That may be the perfect solution.  Thanks a lot!

Only thing I worry about is it being too bright.  Do you think I could lower the brightness by adding a resistor inline with the capacitor and maybe using a lower value capacitor?

What I am designing you need to be able to see it in pitch black darkness, but any ambient light should make it invisible.  I guess I should just play around wiht it and see.
On two occasions, I have been asked [by members of Parliament], "Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?" I am not able to rightly apprehend the kind of confusion of ideas that could provoke such a question.
Charles Babbage

bedlamite

  • Hold my beer and watch this!
  • friend
  • Senior Member
  • ***
  • Posts: 9,829
  • Ack! PLBTTPHBT!
Start with a low intensity LED http://www.newark.com/jsp/Optoelectronics,+Lamps+&+Displays/LEDs/AVAGO+TECHNOLOGIES/HLMP-7000/displayProduct.jsp?sku=06F6859

Changing the capacitor will change the frequency, I'm not sure how a resistor will affect it.
A plan is just a list of things that doesn't happen.
Is defenestration possible through the overton window?

Gewehr98

  • friend
  • Senior Member
  • ***
  • Posts: 11,010
  • Yee-haa!
    • Neural Misfires (Blog)
I used the same flashing circuit Bedlamite linked to in my last car.  It flashed a small red LED mounted in the dash whenever the ignition switch was off.  (Low-budget car alarm simulator)

Radio Shack has integrated that flasher circuit within an LED's base, creating a flashing LED. I've used those flashing LEDs in several projects, and they're still easy to obtain. That's good, because the original LM3909 LED flasher IC has long since been discontinued.
"Bother", said Pooh, as he chambered another round...

http://neuralmisfires.blogspot.com

"Never squat with your spurs on!"

cfabe

  • friend
  • Senior Member
  • ***
  • Posts: 513
I agree with the suggestion of a flashing LED. To be noticable to the eye, you need to have the LED on for about 20ms. I would suggest a flash of about 50ms every second. This would give you a 20x increase in battery life (ignoring standby current) over a constant on LED. I would suggest looking up some 555 timer circuits. If you set up the circuit with a 555 timer and select a chip designed for low power consumption, be careful with your RC circuit (select a high ohm resistor, like 1meg) you will be able to have this thing flash for a very long time.

The problem with the constant-on circuits you've considered, both the voltage drop resistor and the linear regulator, is that while they do "dim" the LED, they still consume a considerable ammount of power in the device itself, wasting it as heat; particularly the linear regulator. You could use a switching regulator for better efficiency, but really a flashing circuit will be the best solution if flashing is acceptable for your application.

CNYCacher

  • friend
  • Senior Member
  • ***
  • Posts: 4,438
So instead of making it dimmer, I would just make it flash so quickly that it is barely noticable. . . interesting idea.  I actually did want to make it flash, so maybe I will go with this idea.

The context is that people are trying to find something I have hidden.  I am going to mark it with this LED.  They will be lead to a certain spot in the woods nearby, and have to stand there looking around for "something".  They are instructed that they need to turn off all their flashlights and headlamps and stand in the woods in the pitch black in order to see it.  The idea is that their eyes will need to be acclimated to the pitch black before they will be able to see the light.  That's why I originally thought a very dim light would be best, but I can see where a very quick flash might work also. . . not sure though.

The important part is that it should not be visible unless you shut down all your lights and adjust your eyes to the pitch black for long enough to get spooked.  If you can see the flashing LED with a flashlight pointed at it, the effect is worthless.

This is for a geocache, by the way.
On two occasions, I have been asked [by members of Parliament], "Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?" I am not able to rightly apprehend the kind of confusion of ideas that could provoke such a question.
Charles Babbage

Alberttaylor

  • New Member
  • Posts: 2
So, I have a need to create a circuit that will drive an peimar led lights very dimly but for a LONG time on one battery.  By dimly I mean it should be able to be seen only in pitch blackness from 50-100 feet away, and by a long time I mean measured in months or years if possible.

I experimented with some simple circuits using a large resistor 20k ohm, and a 50k ohm 15-turn adjustable resitor, and an LED with a 3-volt CR123 lithium battery.  This worked well for setting the dimness factor way down, and was adjustable, but the simpleness of the circuit proved to be the downfall of it, as the LED weakened along with the battery running down, and I need the brightness to be constant once I set it to my liking.

So being a hobby-level circuit maker and not formally trained in any of this, I surmised that what I want is a voltage regulator of some sort which outputted a set voltage, lower that the 3-volts provided by the battery, and yet high enough to drive the LED.  My thinking is that then the LED will hold a constant brightness until the battery dies down to below the voltage regulator output.  Lithium cells having a good pattern of holding their output for as long as possible before rapidly dying out, I think this will do very well.


I found <a href="http://www.national.com/pf/LP/LP3992.html" target="_blank">THIS[/url] gizmo which appears that it will meet my needs, if in fact my needs are as I imagine them to be.

How I think it will work is this:  I hook the battery up to the ground and Vin of the chip.  I hook the LED with large adjustable resistor inline to the Cout and Vout of the chip.  I adjust the LED brightness with the resistor until I like where it is, and let the thing run along.  As the battery eventually begins to die out, its voltage will drop lower and lower without affecting the brightness of the LED, until the battery output drops below 1.9 volts, at which point the whole thing may shutoff or whatever, and hopefully this point will be a long time in the future.

Make use of varying resistor
« Last Edit: December 09, 2012, 03:59:57 AM by Alberttaylor »