1. Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.
    Dismiss Notice

led MATRIX CURRENT DRAW

Discussion in 'Microcontrollers' started by MrDEB, Oct 4, 2012.

  1. be80be

    be80be Well-Known Member

    Joined:
    Aug 23, 2008
    Messages:
    4,807
    Likes:
    135
    Location:
    morristown,tn
    I really said that wrong you want the resistors on the column's not the ROW but I did mean you maybe have that backward which looks like you did and what Jon is telling you is the more leds you light there no way to do that these LED Matrix where made to light one dot at a time.
    View attachment 67997
    Now if at the top left you set high portC.0 and low portb.0 the first led lights. Turn that one off and go for the next.

    I see now how you got all out of whack you looked at grams game then hooked resistors to the rows as he did then you tried to light a whole row at a time which places the leds parallel which could work but the resistor limits the power and divides it not evenly and if you change value when a led is off, You light 2 of them your power to them goes threw the roof. Like 5 you see 20mA and 2 you see 100mA. Not a good idea
     
    Last edited: Oct 14, 2012
  2. MrDEB

    MrDEB Active Member

    Joined:
    Apr 16, 2007
    Messages:
    4,364
    Likes:
    22
    I have tried turning on ONE LED at a time as was suggested but its very dim. Tried at 10ms. BUT with the LEDs not getting full current of 20ma they will be DIM. Perhaps thas why the 10ms delay looks like crap.
    Going to wait from the parts from Digikey and see what I get. I turned on two segments in one row and yes the current was almost 60ma but I was running with 8v using a transistor. I do not plan to light up more than two leds per row. My original design had 220 ohm resistors on the rows. That is no longer what I am using. The schematic I posted is where I am at this point.
    I told Jon on several occasions I AM ONLY GOING TO USE THE LED MATRIX for durability issues. He didn't understand english or ?? he kept bring it up as usual.
     
  3. JonSea

    JonSea Well-Known Member

    Joined:
    Oct 1, 2012
    Messages:
    1,118
    Likes:
    90
    Location:
    Seattle, WA
    MrDeeb, perhaps I have a better understanding of your (lack of) skills than you do. You certainly don't let a complete lack of understanding influence your thought process.

    If you had read my last post (you do know how, yes?), you would have seen that the entire thing was directed at using your (free samples from Kingbright) LED matrix. Oh, since you failed to notice that, let me put it here for you:


    The really important thing you refuse to believe is the LEDs cannot be operated in PARALLEL. Each LED must have its very own current limiting resistor. One LED, one resistor. No LEDs connected in parallel with a single resistor, ever.

    Understanding that is essential to using a matrix display. Let me say it again since you've ignored it numerous times already.

    No LEDs connected in parallel with a single resistor, ever.

    One LED. One resistor.

    One LED. One resistor.

    One LED. One resistor.

    Why is this? LEDs will not share current nicely. Period.

    Here's some help:

    LEDs for Dummies

    So you've ignored Brad's unbelievable efforts to help.

    You've even failed to acknowledge Mike's comments, questions and offer to create a circuit for you.

    You've ignored Burt's offers to help.

    And you've ignored all of my comments, including two complete discussions of the required steps to use a multiplexed display.

    Do you think none of the four of us understand what you're trying to do? Really?????
     
  4. dave

    Dave New Member

    Joined:
    Jan 12, 1997
    Messages:
    -
    Likes:
    0


     
  5. be80be

    be80be Well-Known Member

    Joined:
    Aug 23, 2008
    Messages:
    4,807
    Likes:
    135
    Location:
    morristown,tn

    If you look at this picture in the post you could light the Red leds at the same time If you can sink 166.5 mA

    You can't light all the blue you'll run out of power the resistor limits it to 33 mA so one led may light but the rest will not if two lights then your leds are not droping 4 volts I f you raises the voltage to light all leds then If you light just one or two maybe three you'll more the likly blow a led or two or three.
    Don't hook it to a port use 7 NPN's on portB too sink not shown in picture
    View attachment 68013
     
    Last edited: Oct 14, 2012
  6. JonSea

    JonSea Well-Known Member

    Joined:
    Oct 1, 2012
    Messages:
    1,118
    Likes:
    90
    Location:
    Seattle, WA
    Burt's red LEDs each have a current-limiting resistor. These can all be illuminated at the same time if each port pin is configured correctly.

    If you make the resistor small enough to get enough current to light each blue LED at the desired level...you'll let the smoke out. LEDs aren't exactly the same. The one with the lowest forward voltage will grab most of the current.

    This effect is easy to see with an RGB LED. if a constant current source is connected in parallel across the red, green and blue elements, all you're likely to see is a bright red LED since its forward voltage is the least. You may see a dim light from the green LED but the voltage is clamped to the forward voltage of the red LED and the blue will never see enough voltage to turn on.

    Sorry, I know MRdeeb doesn't like explanations, but some may find it useful.
     
  7. MrDEB

    MrDEB Active Member

    Joined:
    Apr 16, 2007
    Messages:
    4,364
    Likes:
    22
    I stated that I have one resistor per colum(cathodes) and no resistor on the row (anodes) AND ONLY enabling one row at a time. So in fact I have say two LEDs on with two resistors (one per LED). YES IF I enabled all 5 LEDs in one row I have 125ma current draw but all the 5 LEDs will be at full brightness running at 25ma each. BUT only for 10-20ms before the next row is enabled and turn off the previous row.
    I don't plan on running 5LED per row at one time. That was a demo test.
    My first design HAD 220 ohm resistors on each row(anodes) but on the last schematic the resistors are on the colums BUT only one LED per colum on at a time.
     
  8. JonSea

    JonSea Well-Known Member

    Joined:
    Oct 1, 2012
    Messages:
    1,118
    Likes:
    90
    Location:
    Seattle, WA

    Ok then. You know everything. So why isn't it working?

    I am truly sorry I have tried to help you. Sorry I pointed out that whatever you described was wrong and wouldn't work. Sorry to have troubled you with all those explanations in the hopes that you might learn something. Sorry to have upset your fantasy world where things work because you want them to. I am sorry to have wasted all your time.

    It won't happen again. You have yourself a nice day.
     
  9. MrDEB

    MrDEB Active Member

    Joined:
    Apr 16, 2007
    Messages:
    4,364
    Likes:
    22
    I am not saying your right or wrong.
    Here is an example schematic which is pretty much exact what I am doing.

    http://tobiscorner.floery.net/led/big-7-segment-display-clock/
    using a UDN2982 which is the same as the TD62783APG
    this guy is using 18-20v where as I am only going to use 7v. He is scanning the displays using 80ma dividing by 4 displays = 20ma.
    An his project works.
     
  10. JonSea

    JonSea Well-Known Member

    Joined:
    Oct 1, 2012
    Messages:
    1,118
    Likes:
    90
    Location:
    Seattle, WA
    Perfect. Perhaps he'll be your next victim? I think you've just about bled everyone here dry.

    By the way, the link you posted is exactly the circuit I described again in post #103. Sorry all the explanation kept you from understanding.

    Watch your necks guys. See you around.
     
    Last edited by a moderator: Oct 16, 2012
  11. be80be

    be80be Well-Known Member

    Joined:
    Aug 23, 2008
    Messages:
    4,807
    Likes:
    135
    Location:
    morristown,tn
    He not doing what your doing Doug 7 segment leds are like having 7 leds by there self for a total of 28 leds
    you have 70 leds and you can't fire 7 at time he can. Only one so you give that one up to 40mA flash and go to the next one and see how it goes.
     
  12. Mike - K8LH

    Mike - K8LH Well-Known Member

    Joined:
    Jan 22, 2005
    Messages:
    3,637
    Likes:
    109
    Location:
    Michigan, USA
    Hey Doug (MrDEB), the guy who designed that project knows what he's doing. You should pay particular attention when he says;

    This brings up a fundamental design consideration that I suspect you may have overlooked. Please consider that there are a number of ways you might drive your multiplexed 7x10 matrix;

    (A) one row by five columns with fourteen cycles to refresh the entire display (7.1% duty cycle)
    (B) one row by ten columns with seven cycles to refresh the entire display (14.2% duty cycle)
    (C) one column by seven rows with ten cycles to refresh the entire display (10% duty cycle)

    If you're designing a full brightness display with 20-mA "average" current per segment you'll need to provide ~280-mA "peak" current per segment at a 7% duty cycle, or ~140-mA "peak" current per segment at a 14% duty cycle, or ~200-mA "peak" current per segment at a 10% duty cycle. I haven't seen any displays that can handle more than 100-mA "peak" current per segment. And so, duty cycle becomes a fundamental consideration when designing any matrix display because it directly translates into your ability to meet LED "peak" current specifications. Which BTW is why I recommended multiplexing configuration 'B' with its much higher duty cycle back in post #29.

    Now consider current requirements for driving configurations A, B, and C above. For configuration 'A' each pin on your sourcing row driver needs to provide up to 1400-mA "peak" current for five (280-mA "peak") column LEDs. Configuration 'B' also requires up to 1400-mA "peak" current for ten (140-mA "peak") column LEDs. Unfortunately, none of the sourcing driver ICs you've looked at can handle 1.4 Amps "peak" current. Configuration 'C' would require each output on the sinking row driver IC to handle up to 1400-mA "peak" current for seven (200-mA "peak") row LEDs. Again, the sinking driver ICs you've found cannot handle that much current.

    Fortunately, the prognosis is not as gloomy as it sounds. Since you're using a very high luminous display, I suspect you'll find that ~14-mA "average" current per segment is nearly as bright as 20-mA "average" current per segment. This can be achieved by using method 'B' with its 14% duty cycle, seven NFET/PFET pairs on the anode rows capable of sourcing 1000-mA "peak" current to ten column LED segments, and a pair of TLC5916's connected to the ten cathode columns. Bias the matrix for ~100-mA "peak" current per LED segment (~14-mA "average" current), and you're in business.

    I'd also like to point out that Jon's suggestion for using thirty five discrete red/green LEDs has merit. All it would take is five TLC5916, or similar, sinking driver ICs to drive all of the LEDs at 100% duty cycle with 20-mA "average" current.

    Food for thought.

    Regards, Mike
     
    Last edited: Oct 15, 2012
  13. MrDEB

    MrDEB Active Member

    Joined:
    Apr 16, 2007
    Messages:
    4,364
    Likes:
    22
    I fully understand what your saying Mike. I hopefully am only going to enable two LEDs at a time via the anodes. Only utilizing the red colums separate from the green colums.
    So selection B 14.2 duty cycle x 2 = 28.4 duty cycle if I am interpreting it right?
    BUT I am not going to enable all the 35 leds. Each routine enables 5 LEDs per routine but only 2 max at any one time.
    I admit I am the last person to converse about math but just curious if possible about the 280ma max?
    If I have two leds ON in one row at 20ma does that not equal 40ma. Or am I missing a crucial math equation?
    LOST as usual
     
  14. MrDEB

    MrDEB Active Member

    Joined:
    Apr 16, 2007
    Messages:
    4,364
    Likes:
    22
    Now I am really LOST
    I have 5 segments per row so 5 x 20ma = 100ma max.
    This is getting confusing.
    The link I posted was exactly what I have been describing. My schematic I posted is the same but without the High Side driver you suggested which I have ordered.
     
  15. Mike - K8LH

    Mike - K8LH Well-Known Member

    Joined:
    Jan 22, 2005
    Messages:
    3,637
    Likes:
    109
    Location:
    Michigan, USA
    Hi Doug,

    Are you saying that you're only lighting patterns of five LEDs at any one time? If so, you could use that to great advantage to increase duty cycle and brightness. Could you elaborate, please? What patterns are you going to generate? I'm really very interested and very puzzled about your project. For example, I imagine you're building the display to use for right and left turn signals and brakes, but since those are always red I can't imagine why you would need green or yellow colors.

    As for your other questions, I'm a bit lost. Please let me study to see if I can figure out what you're asking. Thanks.

    Regards, Mike
     
    Last edited: Oct 15, 2012
  16. MrDEB

    MrDEB Active Member

    Joined:
    Apr 16, 2007
    Messages:
    4,364
    Likes:
    22
    The red leds for signaling a progressive arrow pointing either left or right.
    The green will just be blinking as a n indicator to vehicles that HERE I AM.
    I have a video of what I am refering to. Let me locate.
     
  17. Mike - K8LH

    Mike - K8LH Well-Known Member

    Joined:
    Jan 22, 2005
    Messages:
    3,637
    Likes:
    109
    Location:
    Michigan, USA
    Basically, only one digit in that four digit display is on at any time. The first digit is lighted, then the second digit, then the third digit, and finally the fourth digit. Each digit is on for 1/4th of the time that it takes to refresh the entire display. That's a 1/4th or 20% duty cycle. The designer wants to drive each LED in the display with 20-mA "average" current for full brightness but since each LED is only on 1/4 of the time he needs to bias the LED so that it's lighted with X = 20-mA / (1/4) = 80-mA current. Does that help?

    Regards, Mike
     
    Last edited: Oct 15, 2012
  18. MrDEB

    MrDEB Active Member

    Joined:
    Apr 16, 2007
    Messages:
    4,364
    Likes:
    22
    rough idea I am doing

    here is a video of my basic idea.
    The POV using LEDs is kinda out the widow dut to the LEDs being dim.
    http://s992.beta.photobucket.com/user/MrDEB/media/Video2.mp4.html
    Really a rough idea but it shows the progressive idea I am trying to achieve.
    Only a maxium two LEDs per row on at any one time. This may change ??
     
  19. MrDEB

    MrDEB Active Member

    Joined:
    Apr 16, 2007
    Messages:
    4,364
    Likes:
    22
    Houston we have LIFT OFF!!

    Well I connected the LED High Side driver and the LEDs are brighter BUT need to change some resistors and as Mike pointed out duty cycle(presently set at 10MS)
    The high side driver TD62783APG controls the ROWS
    at present I have the ULN2803 on the cathodes w/ a 220 resistor.
    one row enabled: 1 colum = 12.4ma
    2 colums = 22.9ma
    3 colums = 32.8ma
    4 colums = 42.3ma
    5 colums = 51.2 ma
    1 ROW, 2 COLUMS alternating back and forth w/ DELAYMS(10) = 12.3ma
    Now I get what Mike was referring to about duty cycle

    Conclusion dim LEDs screw up the POV idea. Been there done that
    contemplating finish up the code THEN experiment with the TLC5917 Constant Current Sink. This might eliminate all the resistors down to one. Goes in place of the ULN2803.
    Looking over the matrix cube description of how the current sink is set up I might just take a whack at it.
     
  20. be80be

    be80be Well-Known Member

    Joined:
    Aug 23, 2008
    Messages:
    4,807
    Likes:
    135
    Location:
    morristown,tn
    Why
    I told you how it was wired and I made this one and It doesn't have any chip diving it just npn's sinking
    [video=youtube_share;OsUJwhyjJJc]http://youtu.be/OsUJwhyjJJc[/video]

    And the room has enough lights on to find a nat you no what in. It was bright

    And believe it or not there only one led on at a time just scanned fast in loops so you can see them
     
    Last edited: Oct 16, 2012
  21. be80be

    be80be Well-Known Member

    Joined:
    Aug 23, 2008
    Messages:
    4,807
    Likes:
    135
    Location:
    morristown,tn
    Here a 5x7 led matrix and yes it's POV only one led on for 500uS then it's totally shut off for 100uS next led is loaded and switch on then the same for the next till all five have flashed for 500uS each in a loop.

    Yes it's POV one led at time and there getting 25 mA each



    And the same leds in the dark[video=youtube_share;54cIqr-d0d8]http://youtu.be/54cIqr-d0d8[/video]
    Dang there all the same brightness

    And Doug i reassure you this is POV there one led flashing at a time in a loop and it looks like there all on to your eyes.

    P.S. one more thing these leds are powered by the port pins only.
    7 pins are sinking and 5 pins sourcing
     
    Last edited by a moderator: Sep 10, 2013

Share This Page