OK to decrease LED series resistor on mux/switching design?

Discussion in 'General Electronics Chat' started by hspalm, Dec 13, 2011.

  1. hspalm

    Thread Starter Active Member

    Feb 17, 2010
    201
    8
    Hi,
    I want to drive a 13x13 LED matrix. Since one LED will only be lit at a maximum of 1/13th of the time, is it okay to decrease the series resistor so it's brighter? I think I read this in an application note regarding IR LEDs in remote controls.

    Thanks!
     
  2. sheldons

    Active Member

    Oct 26, 2011
    616
    101
    post your schematic........
     
  3. KJ6EAD

    Senior Member

    Apr 30, 2011
    1,425
    363
    The LED data sheet will show the limits you need to know.
     
  4. hspalm

    Thread Starter Active Member

    Feb 17, 2010
    201
    8
    The schematic is attached as a pdf in this post (designspark pcb sucks on exporting). I'm am not yet sure of power supply solution but in the schematic now the LEDs are powered from 7805 regulated 5v which I will not do in the final design, because of unnecessary current through regulator. I will also find out if I can power ULN2003 with unregulated battery voltage, to save power waste in 7805 regulator.

    I'm not sure how this is gonna be powered yet, battery or wall warth or if maybe input voltage will be so low I don't need any regulator for the AVR. So I added decoupling caps on all supply inputs. Although ULN2003 won't work below 5v, hmmm.

    Transistor BC307 are rated Ice 100mA, but current will never exceed the current going through one LED. Worst case scenario the ULN2003 must handle current from all 13 LED in one column, so maximum current per LED is limited to 500mA/13=38mA which is more than enough. Please tell me of any other design errors you spot?

    edit: one more thing. I know it's bad but I don't have the data on the LEDs, they're all scavenged off of a PCB from a PLC controller with many I/O lines. They were used as indicators for active I/O, I suspect them being 20mA and voltage drop doesn't really matter do it, since I will always light only one in series on a 5v or higher supply? A picture of the LEDs re-soldered on prototyping board, all tested http://yfrog.com/kkmgwpaj.

    I think it's this type https://www.elfaelektronikk.no/elfa3~no_no/elfa/init.do?item=75-021-90&toc=20266
     
    Last edited: Dec 14, 2011
  5. hspalm

    Thread Starter Active Member

    Feb 17, 2010
    201
    8
    Just for fun, let's say it's this datasheet
    https://www1.elfa.se/data1/wwwroot/assets/datasheets/yx253184-253187_e.pdf

    Forward current 25mA, peak forward current 60mA. This means that average current can not exceed 25mA when pulsed with peaks no higher than 60mA?
    It doesn't say anything about peak current pulsewidth, what is common (is anything common)?
     
  6. hspalm

    Thread Starter Active Member

    Feb 17, 2010
    201
    8
    Okay, I have another question but the last ones are still open for replies :) (sheldons, I drew that schematic just for you! Series resistors of 1K have later been added on the ULN2003 inputs)

    I am using M54522P which is similar to ULN2008 only with 400mA per darlington pair. But what I'm not getting is this:
    [​IMG]
    The COM pin. What do I need it for? I tried searching for schematics on ULN2008 to get a hint, but sometimes it's excluded, sometimes it's not. If I understand the functional pinout correctly, the inverting buffer stage acts as a current sink when it's input reaches a certain level, pulling current from the load and into ground. But what use does the common(?) COM pin do?

    The datasheet says: A spike-killer clamping diode is provided between each out-
    put pin (collector) and COM pin.
    So is it just for connecting the spike-killer diodes in case of inductive loads? So if my load is a row of LEDs I leave it unconnected?
     
Loading...