Defender Panel and wierd LEDs

I’ve been busy working on the Panel for Defender whilst I have the time, and I managed to find the impossible today. I had a fault with the LED matrix that meant all the LEDs lit up dimly when they shouldn’t have.. Most odd, so I disconnected the ‘column’ pin which supplies the 5V to the LEDs, and guess what? they *still* lit up, if anything, brighter than before.

How on earth was this happening? here’s a simplified part of the circuit, imagine the ‘nc’ pin is my ‘column’ and the the pins labeled 5v and 0v are attached to a buffer chip which when a pin is low (i.e. 0v) and the ‘column’ is high, the LED is lit.

WTF LED problem

So, give the schematic above, with nothing attached to the ‘nc’ pin what would you expect? would you expect LED A to be lit? LED B to be lit? both to be lit? or neither?

Personally, I’d expect neither to be lit as LED-A is blocking the 5v as it’s backwards. which is what I’d counted on for my matrix. But, there’s something wierd about these LEDs.

In the case above, LED B lights up, yep, that’s right, it lights up! It seems that these LEDs are very strange, I metered across LED-A and there’s 0.7v! I tried two other LEDs (not the ones in the switches) and they behave as I’d expect, i.e. neither lights up. I can only assume that the LED in these switches has a diode connected in reverse across the LED. Why this should be, I really don’t know or understand.

The resolution is, that the column pins now have diodes in series and the row pins now have diodes in series too. This way there’s never 5v going to the cathode of one of these LEDs, it’s either 0v or floating (care of the diode).

Anyway, after loosing an hour and half trying to figure out what on earth was happening things are back on track again.


I contacted the supplier of these switches, and they contacted the manufacturer, and here’s their response -

“For Static electricity resistance, the pure green LED, white LED and blue LED with zener diode is needed to discriminate the characteristics of LED by fixing current and measuring voltage. Normally, the value of reverse voltage of pure green LED, white LED and blue LED is 0.7V.”

This explains what I was seeing and hence the 0.7v reverse bias voltage. Interesting to know though :-)

Comments are closed.