I have two LED light bars from an old (manufactured January of 2000) Unisys check scanning machine. They apparently (based on the PCB's labels) have 72 LEDs. Most of the LEDs are green, but a few red ones are scattered among them for reasons unknown. The boards have 16-pin ribbon cables and metal pieces glued to the backs that seem to be heatsinks. Breaking the glue, there seems to be a line of thermal paste between the heatsink and the metal surface of the back of the PCB. The individual LEDs seem to be isolated from each other, not even sharing a common ground or power. There don't seem to be any flip-flops or microcontrollers on the board. I found only one pin on the ribbon cable that seemed to have any continuity to the LEDs. Below are pictures of these boards. Any ideas of how these could be driven with e.g. an Arduino, besides soldering to each LED? Thanks.

Front of the board Label on the board

  • \$\begingroup\$ no two LED contacts have continuity? impossible i say, keep trying. they could be in series... \$\endgroup\$
    – dandavis
    Apr 3, 2018 at 1:40
  • \$\begingroup\$ Well, they all have the same polarity and if I hold one pin on one contact and slide the other one along, nothing lights until I get to the point of the two pins being on opposite sides of the LED. \$\endgroup\$
    – user48147
    Apr 3, 2018 at 1:52
  • \$\begingroup\$ could be in series, which means both sources would be on the same side; check from - on B to + on A/C. also, those two junctions might be more than a continuity or dioide feature supports, and i suspect there's UV leds in there to read the security features, so you might not see anything. but 16->72 with no little boogie man in the chain means they must share, right? \$\endgroup\$
    – dandavis
    Apr 3, 2018 at 1:59
  • \$\begingroup\$ They've got to! But I've already tested like you suggested and found nothing. I've hit nearly every LED with power from my Arduino and every one of them seemed to be either green or red. And if they were in series, wouldn't they light up as I moved the power jumpers from the Arduino along the LEDs? I was really hoping someone might recognize these as some sort of standard equipment but it seems to be all custom from what I can Google. \$\endgroup\$
    – user48147
    Apr 3, 2018 at 2:07
  • 1
    \$\begingroup\$ 72 LEDs and 16 bus lines, so many are connected in series/parallel arrays. Get a DMM and make a wiring list \$\endgroup\$ Apr 3, 2018 at 2:11

1 Answer 1


My probing is beginning to bear fruit. Here is what I have discovered so far:

  • My multimeter's leads are too blunt to make good contact; jumpers from my Arduino kit work better.
  • The LEDs seem to be divided into sets; within each set the LEDs are wired in series.
  • Pins 1 through 7 on the ribbon cable are electrically connected inside the PCB and form the positive power supply (approx. +12V)
  • All the red LEDs are a set (I can get up to 3 to light up at a time with 5V from my Arduino)
  • Every ninth green LED is part of the same set.
  • Pin 8 is the ground for the red LEDs.
  • Even with 12V, only 7 out of the 8 red LEDs will light (I have to touch one of my power leads to the second-to-last LED rather than the ribbon cable). With the green LEDs, only 6 out of each set of 8 will light with any reasonable brightness. I assume that with higher voltages I could get them all to light at once.

The two light bars are slightly different (the ground pin details above refer to the one labelled "R/R") but they are very similar otherwise. The only major difference seems to be that the "L/F" one has two sets of red LEDs and the "R/R" one has only one. Thanks for your suggestions; I think the mystery is solved.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.