Strange results from very simply led lighting

Discussion in 'The Projects Forum' started by BrainFog, Jul 2, 2014.

  1. BrainFog

    Thread Starter Member

    Jan 24, 2011
    122
    4
    Hello everyone,

    Most of the time I am more of a lurker on this site so it is a little odd to be posting. :p

    Yesterday I decided to make a simple led light made up of 6 parallel rows each made up of a series of 3 (5mm) leds and an 80 ohm resistor to limit current. Connected to a 12v battery.

    It was my intent to run the leds at about 40mA at the cost of lifespan but past tests have shown they should be fine at this current.

    My weird problem is that one of the strings of leds is drawing far more current than it should and overheating, I cant work out why. I carefully checked and there are no shorts. All resistors (each 80ohm resistor is made up of 6 cheap 470 ohm resistors in parallel) have almost exactly the same resistance when read with an ohm meter. When powered the 3 leds drawing too much current have a voltage of about 10.4v and the other 5 stings have a voltage of about 9.9v.

    Seeing as all the resistors are identical how is this happening?
     
  2. ericgibbs

    AAC Fanatic!

    Jan 29, 2010
    2,503
    380
    hi BF,
    Isolate each LED and test its individual forward working voltage at say 25mA [ you could use 40mA, I wouldn't :rolleyes:]

    You will find there is a wide spread in Vfwd for LED's of the same colour.
    So in a parallel combination the lower Vfwd LED's will draw more current and may die.

    Try to match the Vfwds of the parallel LED's by testing each LED on its own test resistor.
    E
     
  3. wayneh

    Expert

    Sep 9, 2010
    12,115
    3,039
    It's true that LEDs vary and that may ultimately explain this, but I think something else must be happening. Drawing MORE current across a resistor should give a LOWER voltage on the low side, not higher.

    So I'll propose another explanation: Measure and examine your resistors again. Maybe try swapping the suspect resistor to one of your other LED strings, to see if the anomaly moves with the resistor or with the LEDs. Maybe the resistance is changing with high heat?
     
    Metalmann likes this.
  4. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,797
    1,103
    So there's something else odd here. Assuming the '12V battery' does indeed provide 12V and you have ~10V across a LED string, that leaves 2V across the 80 Ohm resistor; i.e. the current = 2/80 A = 25mA ?
     
  5. BrainFog

    Thread Starter Member

    Jan 24, 2011
    122
    4
    Thanks for your replies

    I carefully checked and the battery was giving off about 12.8v at its terminals. I also checked to make sure all of that voltage was being dropped across only the resistors and leds and it was.

    I disconnected all the solder from my project and tested all the leds individually. I was surprised that they did draw significantly more voltage than others. When using a 470ohm resistor in series with each individual led (providing about 20mA) I got voltages of 3.38, 3.46 and 3.32 while unused ones from the same batch I tested gave voltages like 3.25v.

    Maybe I made a mistake in thinking they were overheating. I thought that because after about 10 minutes of being on the faulty series became dimmer and changed from white to blue but after being left off for a while went back to normal. Does this sound like it is as a result of overheating?

    I know of something I think is called negative temperature coefficient (when resistors get very hot their resistance drops) Could a faulty resistor cause its resistance to noticeable change under load?
     
  6. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,797
    1,103
    The power dissipated in the 80 Ohms is only ~2V x 25mA = 50mW, and that is shared by 6 resistors (assuming you don't have a faulty low-value one in the bunch), so heating should be negligible. Check the resistance of each '80 Ohm' bunch.
     
  7. BrainFog

    Thread Starter Member

    Jan 24, 2011
    122
    4
    Just quickly checked the resistors in the faulty series and individually they are between 465 and 467 ohms according to my multimeter.

    All of the bunches of resistors had approximately the same resistance (within a few percent) when measured in circuit.
     
Loading...