Keep constant LED brightness between 3V and 1,8V

Discussion in 'Power Electronics' started by Goxeman, Jan 18, 2019.

  1. Goxeman

    Thread Starter Member

    Feb 28, 2017
    77
    2
    Hello,

    I have a battery powered device with two AA batteries. I have LEDs and the problem, as it is expected, is that the LEDs brightness change as soon as the batteries drain.

    The device works perfectly between 3V and 1,8V so I would like to use the all battery capacity, but it doesnt look good that the LEDs brightness is affected by the battery voltage level or even could almost turn off because you cant see them.

    Within my device I have a new PIC MCU (works between 3,5V and 1,8V I think) and everything is powered directly from the two AA batteries


    Thanks
     
  2. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    9,640
    2,330
    A DC-DC boost converter could raise the voltage to keep the LEDs at constant brightness. Unfortunately, the converter's current draw would probably drain the batteries quickly.
     
  3. Sensacell

    Moderator

    Jun 19, 2012
    2,309
    737
    What color LED? it matters because the Vf is different depending on color.

    The tricky part is that the battery voltage begins above the LED Vf, then ends below it.
     
  4. Goxeman

    Thread Starter Member

    Feb 28, 2017
    77
    2
    Yes, I dont consider that option because as you said the battery drain would be too fast. The option I am searching for has to be power efficient, that is what is keeping me thinking about the matter

    I am using 4 different colors, red (1,8V), green (1,8V), blue (2,7V) and white (2,7V). I could use all LEDs of 1,8V or all LEDs of 2,7V but in any case what is just affecting is the resistor value. At the end even if I use all LEDs 1,8V and with its proper resistor, the combination would be affected by the voltage drop
     
  5. danadak

    Distinguished Member

    Mar 10, 2018
    2,939
    640
    Constant current source insures brightness, but added circuitry to do that
    lowers overall effirciency.

    Another approach use a photosensor, and a UP with PWM, A/D, in a control loop.
    Based on photosensor output vary PWM duty cycle to control brightness.
    Thats easy to do, ATTINY85 could do that. Or PSOC -

    upload_2019-1-18_9-7-9.png

    LED, R, Photosensor offship, the rest is inside PSOC chip. Just used a
    fraction of overall PSOC chip resources.

    Regards, Dana.
     
  6. crutschow

    Expert

    Mar 14, 2008
    22,119
    6,406
    Below is the LTspice simulation of a low power, low voltage constant-current source.
    It uses a micropower op amp, and the low forward voltage drop of a Schottky diode as a reference voltage.
    The LED constant current is ≈160mV/R2.
    The simulation shows the LED current fairly constant down to a battery voltage of ≈2V with a 1.8V LED (the slope is due to the change in D2 forward voltage with current).

    Edited to use low cost (U$0.46 for 1k units), low voltage (1.8v-5.5V) op amp.
    upload_2019-1-18_12-45-46.png
     
    Last edited: Jan 18, 2019
    absf likes this.
  7. danadak

    Distinguished Member

    Mar 10, 2018
    2,939
    640
    Would be instructional to calc the total energy delivered in both methods,
    PWM switched current source vs analog linear current source. That would
    establish which gives most battery life for a fixed LED brightness.

    Its complicated -

    http://www.pathwaylighting.com/prod...Linear+vs+Logarithmic+Dimming+White+Paper.pdf

    https://forum.allaboutcircuits.com/threads/led-brightness-vs-pwm.83957/

    https://www.uv.es/~esanchis/cef/Caracteristicas/abd004.pdf

    Then there is a G factor associated with persistence of vision, eg, brain seeing
    LED on when PWM off, "filling in" light where there is none. Effectively reducing
    energy needed by PWM approach......


    Regards, Dana.
     
    Last edited: Jan 18, 2019
Loading...