Increasing DC voltage for series LEDs

Discussion in 'General Electronics Chat' started by Garoad, May 16, 2012.

  1. Garoad

    Thread Starter New Member

    Apr 24, 2012
    26
    0
    1) I'm considering increasing the voltage from a 12v DC power source (maybe approximately double it) to support more LEDs in series (less parallel wiring to worry about). I've done some research on this, but there seems to be a wide "variety of facts" on this topic. Does anyone feel this is too high or approaching a needlessly dangerous level?

    2) Assuming this is fine, apparently there's more than one way to do this. How would you recommend going about doing this? (Obviously with minimal wasted power, since the supply is likely to be 8 D cells, which I'd like to last at least a few weeks with as many LEDs as I can reasonably fit on the wires.)
     
  2. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    You must have a current limiting resistor, this is always true.

    Other than that, use as much voltage as you wish, as long as the resistor wattage can handle it.

    LEDs, 555s, Flashers, and Light Chasers

    The closer the total Vf of the LEDs is to the power supply, the more efficient it is. However, batteries do change voltage over time, while the LEDs are mostly constant, allow some head room, say 2V. You can also use a constant current source to keep the LEDs a constant intensity even with battery changes. The link I gave is a simple tutorial on the first 2 chapters, and covers most of this subject in detail, complete with schematics.

    There are circuits (SMPS constant current) that will handle this with improved efficiencies, but some electronics required.
     
  3. Garoad

    Thread Starter New Member

    Apr 24, 2012
    26
    0
    Bill - your reference guide is awesome, and it's appreciated. I'm sure I'll be implementing a few of those circuits eventually. (Chapter 8 sounds very similar to another idea I've had, sounds really cool. The only downside is I didn't have the challenge of figuring it out for myself!)

    The main difference between what I'm going for and the example circuits there is that I'd like a few more lights (in the range of 30-40+). Also, I was considering using one or a few 2N6027 PUTs if they can handle the loads.

    I don't mind a little electronics, I've been learning about 555s and so on (via book Make: Electronics). Within reasonable space requirements (this will go in an office in December), a little complexity will make it more interesting (and impress more people - haha). But I'm new at this, so I need to start off simple and work upwards.


    Q. I am a little confused by the absolute resistor requirement, though (maybe I'm wrongly thinking of LEDs as "resistive loads").

    If you check this site here and enter:
    Source Voltage: 12
    LED Vf: 2 (let's assume red/green only to start off)
    LED current: 20
    # of LEDs: 30

    The resulting schematic produces 5 1 ohm resistors in series with the LEDs. But 1 ohm, isn't that so miniscule that it could be ignored? I would think that the "resistance"(?) of the LEDs in this arragement would keep the overall current under control so they don't blow out. Is this wrong?

    Edit: Actually, after thinking about Ohm's law a little, maybe I take back the 1 ohm is "miniscule" comment... but I'll wait and see if someone can explain it a little clearer here.
     
    Last edited: May 16, 2012
  4. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,346
    Hello,

    The current regulation of that calculator is not good.
    A 1 Ohms resistor is to low.
    Any voltage change on the power supply will give a huge change of the current.

    The following led calculator will give a better result:
    http://ledcalc.com/

    It gives 6 strings of 5 leds with a 120 Ohms resistor for each string.

    Bertus
     
  5. Wendy

    Moderator

    Mar 24, 2008
    20,766
    2,536
    You will find most people with experience tend to not favor LED wizards, because they have no common sense. Like I said, you must have some voltage headroom to work correctly.

    A good work around for circuits that don't have the headroom is the following schematic, it is a constant current source with a very low dropping voltage.

    [​IMG]

    http://forum.allaboutcircuits.com/album.php?albumid=152&pictureid=1670

    I am one of the local draftsmen, and I like writing. I don't mind answering questions, just send me a PM with the link to your post (we don't do private tuturials here).

    Bill's Index
     
  6. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,172
    397
    My openion: the 2N6027 makes a good clock or high Z sawtooth generator, but a poor LED driver. Would be nice to know color of LEDs, prob. red & green for holidays? Measured Vf averaged over 25 ea., red- 1.98V, green- 2.17V.
    Using 6 strings of 5 [ assuming 2V LEDs] 8 D cells should give about 160 hours.
     
    Last edited: May 17, 2012
  7. kuza

    Member

    May 11, 2012
    41
    0
  8. Garoad

    Thread Starter New Member

    Apr 24, 2012
    26
    0
    Ok, the new calculator result makes more sense. I guess I'll have to check the results before I trust those. My first string will probably be red/green for the reason Bernard suggests, and the obvious of Vf similarity. (Later on I'll do a blue/white maybe.) And 160 hours sounds exactly on spec with the battery life I want, so that's good.

    Why would you say 2N6027 is a poor LED driver? I'm also interested in 555 solutions, but I figure the simplicity (starting out) of one or a few transistors would probably help me get a first project completed.
     
Loading...