led 7 segment driver not working as expected

Discussion in 'The Projects Forum' started by davidhoff, Oct 20, 2013.

  1. davidhoff

    Thread Starter Active Member

    Sep 19, 2008
    66
    0
    I'm building a circuit that receives 12v pulses from an external circuit and counts them using a microcontroller. It then controls large led 7 segment displays through 2 of these MIC2982 (datasheet: http://www.mouser.com/ds/2/495/mic2981-242019.pdf) to display the count. Each led segment is 15 red leds (3 parallel strings of 5 leds + resistor each) running off 12v (about 60 mA per segment). The problem I'm seeing is that while the lights do turn on when they are supposed to, they sometimes turn on when they aren't supposed to. Sometimes they will be completely on when they should be off. Other times they are flickery. Other times they are off or very dim. I'm not using pwm, and the pulses I'm counting come roughly once per second. I'm attaching a schematic and picture of the board layout. Pads 22-25 are all jumpered on the top sided of the board to distribute 5v power. R1 is 10k & R2 is 4.7k. They drop the 12v pulses coming in at pad 1 to a uc friendly 4ish volts. Pad 2 is 12v power in. Pad 20 is gnd. Both caps are 0.1 uF. If I unplug the uc from it's socket so all the MIC2982 inputs are floating, all the lights turn on full brightness. I can turn a segment almost all the way off by jumpering from one of the inputs on the MIC2982 directly to ground, but it's still on very dimly. When I put a meter on the floating inputs, they read anywhere from 1 to 5v and aren't very stable. It seems to me there is some kind of noise screwing things up, but I don't know where it could be coming from. Power is coming from a 12v switching supply, so that could be something I guess. So, my questions are... Why wouldn't grounding the inputs while the uc is out of the socket force the light completely off? Why are the inputs so unstable? Where might this (assumed) noise be coming from? Am I doing anything glaringly wrong here? Any help is appreciated.
     
  2. davidhoff

    Thread Starter Active Member

    Sep 19, 2008
    66
    0
    I may have found something. It looks like there may be some sketchy connections. At several places on the board where there should be no connection, I've discovered resistances in the kOhm to 10s of kOhm range. I'll try to track them all down and figure out what's going on. It may just be a case of a sneaky solder bridge or something like that.
     
  3. crutschow

    Expert

    Mar 14, 2008
    13,056
    3,245
    What is the source of the pulses?
     
  4. tracecom

    AAC Fanatic!

    Apr 16, 2010
    3,871
    1,394
    Sounds like noise on the power.

    I have trouble reading the schematic as drawn, there are no component values shown, and I have never used an Atmega, so my comments may be inappropriate. C1 should be .33μF and C2 should be .1μF; in addition, there should be a 10μF electrolytic across the output of the 7805. There should be a .1μF across the power pins of the μC.

    Why are you powering the display drivers with unregulated power? What is its source and voltage?
     
  5. RichardO

    Well-Known Member

    May 4, 2013
    1,238
    385
    Did you accidentally use acid core solder? I have a friend that knows someone who did that. :-(
     
  6. tubeguy

    Well-Known Member

    Nov 3, 2012
    1,157
    197
    Can you post some photos of your board?
     
  7. davidhoff

    Thread Starter Active Member

    Sep 19, 2008
    66
    0
    RichardO, you are very close. I discovered that the flux I had used was not for electronics, but plumbing. It is acidic and also slightly conductive. I got some rosin type flux, and everything works great now.
    Thanks for the suggestions everybody. Just for the benefit of others who come across this, I'll answer the questions:

    About the caps-
    I used .1uF caps everywhere because... 1) I found the same basic circuit with the 7805 many places online with varying cap values 2) I already had them around, and in the past the 7805 has worked for me even without the caps
    I didn't use the cap next to the micro pins because it's only about an inch from the caps on the regulator. Maybe I should have used it anyway, but it's working ok now. I didn't use the 10uF across the output because I didn't realize I should.

    About unregulated power to the display drivers-
    1) The display drivers are rated 5-50v and I'm using 12v, so I didn't think regulation was necessary. 2) The led segments should be within current limits up to 14v. 3) The 12v supply is actually a regulated supply, it just happens upstream of this part of the project.

    About source of the pulses-
    This project was to improve on the design of an existing system. These led number displays are used outside and the original driver multiplexed them, so they were dim and hard to read in daylight. This design gets rid of the multiplexing to make them brighter. There was already some sort of 555 based pulse generator circuit board. That board is the source of the pulses. It outputs a 12v pulse for about 1/4 second every second.

    About the solder-
    The solder wasn't acid core, but the flux was acidic and somewhat conductive. This turned out to be the problem. Also, I noticed the next day that the flux had corroded the copper on the board a little. Some rosin based flux fixed the next board.

    About pictures of the board-
    I'll post some soon.

    Thanks everybody!
     
  8. davidhoff

    Thread Starter Active Member

    Sep 19, 2008
    66
    0
    Pictures attached. The copper side of the board looks shiny because I sprayed it with polyurethane after everything was tested and working. This is to help protect the exposed copper from corroding over time. The pcb was cut out at home on my mill.
     
Loading...