Multiplexing - digits too dim

Discussion in 'General Electronics Chat' started by kgstewar, Apr 22, 2014.

  1. kgstewar

    Thread Starter Member

    Apr 5, 2012
    151
    6
    I have a 4-digit display that I am multiplexing (using a MAX7219, which has a scan rate of 800Hz). These are large digits (2.3"), and when I drive a stand-alone single digit with a 12V supply, I can use a 510 ohm current-limiting resistor per segment and it's just as bright as I want it (and near the maximum allowable current). When I multiplex using the same value of current-limiting resistor, of course it's dimmer, so here is my question:

    Can I just reduce the value of the current-limiting resistor to make my multiplexed digits brighter?

    If I use a lower-value resistor, I'm concerned that during the "on" time, the current will exceed the rating of the LED segments. If this is a valid concern, is there a standard solution to increasing the brightness when multiplexing? Or are dim digits the price one pays for multiplexing?


    Thanks in advance.
     
  2. tshuck

    Well-Known Member

    Oct 18, 2012
    3,531
    675
    A current-limiting resistor? You should have one per segment - only one means the LEDs will not share current equally and will result in, you guessed it, dimmer segments.

    Put a current-limiting resistor on each segment.

    See here for more information about driving LEDs.
     
  3. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    One way is to use latches. The data is multiplexed to the input side of the latches, each latch remembers its own data and sends it to the display until new data is latched in.
     
  4. kgstewar

    Thread Starter Member

    Apr 5, 2012
    151
    6
    Hmmm, need to look into this. I do have some TLC59213 drivers which have a latch.
     
  5. kgstewar

    Thread Starter Member

    Apr 5, 2012
    151
    6
    I think I said a current-limiting resistor per segment in my original post :). But my original question remains: Can I simply reduce the value of the current-limiting resistor per segment? Even if at non-multiplexed, constant 12V the 510 ohm resistor per segment is about as low as I can go before exceeding the per-segment max current?
     
  6. tshuck

    Well-Known Member

    Oct 18, 2012
    3,531
    675
    Typically, yes, you can lower the resistor. You are only on for a short time so the LED doesn't have much time to heat up and can withstand over the rated current.

    Before you do this, however, lowering the multiplexing frequency should help. At 800Hz, each digit is only on for 5ms each cycle, try something like ~200Hz.
     
  7. ericgibbs

    AAC Fanatic!

    Jan 29, 2010
    2,503
    380
    Do you have a datasheet to post for the LED.??
     
  8. tshuck

    Well-Known Member

    Oct 18, 2012
    3,531
    675
    Sorry, I guess I misunderstood what you said. You said that you used one per segment as a basis for the brightness, but referred to it as a single resistor once you mentioned multiplexing.
     
  9. kgstewar

    Thread Starter Member

    Apr 5, 2012
    151
    6
    Voltage = 12V

    Peak current per segment = 20 ma

    Forward voltage per segment = 6 Volts


    By my calculation, this yields a current-limiting resistor per segment of 330 ohms. So I could conceivably go this low, but I guess I'm wondering if in a multiplexed situation, could I go even lower? Say 220 ohms?
     
  10. tshuck

    Well-Known Member

    Oct 18, 2012
    3,531
    675
    We'd need more information to say definitively. Do you have a datasheet for these LEDs?
     
  11. kgstewar

    Thread Starter Member

    Apr 5, 2012
    151
    6
    I can upload one tonight, at work right now. That said, I may have written the parameters down that you are looking for. Can you tell me which values from the datasheet you need besides the ones I listed? I may have jotted them down. Thanks!
     
  12. ericgibbs

    AAC Fanatic!

    Jan 29, 2010
    2,503
    380
    I make 6v/0.02 =300R which is a preferred value, this will almost double the present current when using a 560R.

    Why not just modify one digit to 300R and see if that is bright enough.
     
  13. tshuck

    Well-Known Member

    Oct 18, 2012
    3,531
    675
    Ultimately, it'll be absolute maximum ratings and how they were measured that will dictate what it's allowed by the LEDs, however, there are other things to try first...
     
  14. kgstewar

    Thread Starter Member

    Apr 5, 2012
    151
    6
    That may do the trick, thanks. But let's just say for this LED I wanted to go to 200R because I wanted it even brighter. Hypothetically speaking, does multiplexing permit "overdriving" the segment? If not, that's fine, but just wondering if the fact that the LED is off for much of the time does this permit a bit more current than if the LED were constantly on. My instinct says overdriving is probably a bad idea, even if it's blinking off a lot.
     
  15. kgstewar

    Thread Starter Member

    Apr 5, 2012
    151
    6
    Yes, those values were the max ratings. If there are other things to try, I'm all ears! Thanks.
     
  16. ericgibbs

    AAC Fanatic!

    Jan 29, 2010
    2,503
    380
    The datasheet may give a current rating for a specified pulse period.
    The 20mA I suspect is a continuous rating.
     
  17. kgstewar

    Thread Starter Member

    Apr 5, 2012
    151
    6
    Excellent, I will check on this.

    EDIT: I just did a quick google search and found a datasheet for 2.3" LED and it looks like the max current per segment is commonly LOWER for pulsed digits. (Footnote 1 of this datasheet recommends max of 20 mA for pulsed operation. Max continuous current is variable depending on color but is commonly greater than 20 mA).


    So, if I do need brighter, I may need a different trick other than lowering the resistors below the minimum recommended value.
     
    Last edited: Apr 22, 2014
  18. tshuck

    Well-Known Member

    Oct 18, 2012
    3,531
    675
    I think you've got that mixed up.

    Peak Current Per Segment is the pulsed current - note 1 explains what the waveform looks like that measured that value. It also goes on to say that the lifetime will be lessened, which is true of any over-driven component...

    The DC Forward Current Per Segment is for constant current.
     
  19. BillB3857

    Senior Member

    Feb 28, 2009
    2,400
    348
    When you look at raw multiplexing without a latch, the effect is the same as using PWM for brightness control The more digits, the lower the light output since each digit will be waiting for all the other digits to be scanned before it gets its turn to light up again.
     
  20. kgstewar

    Thread Starter Member

    Apr 5, 2012
    151
    6
    Thanks, I AM mixed up, I see now. Pulsed is higher current, typically, than constant. Except for blue....strange...

    In any case, it seems I need to calculate the average current when pulsing, and ideally this should not exceed 20 mA (or whatever this value is for my displays).
     
    Last edited: Apr 22, 2014
Loading...