RGB 16x2 LED backlight problem

[32768]

I reckon these 16x2 RGB negative backlight displays are a popular item and I like mine.

However I do have a set of issues with these things:

1.) Where's the rest of the datasheet? Mouser stocks these but theirs are CC and they have a proper datasheet.
2.) Why are the LEDs wired in parallel?

Here's how it works. The LED is a common anode (common V+ line). The line to this on the bottom of the PCB has a 100Ω resistor. Then each color has it's own 10Ω resistor on its ground lead.

The result is that when you want to make an RGB color mixer, as I have (with some diligent support), and red, blue and green are driven at full PWM 255 values from an Arduino, then red over-steps all the other colors and thus the output is not white, but pure red. Or if you desire violet, then red at 255 and blue at 255 means you get...Red!

I actually went far enough to duplicate this setup on a breadboard, and the result is identical!

Here is red and blue full-on, no µC or anything to interfere (which you can duplicate on the LCD itself):


Doesn't look like there's much hoot coming from blue. What if we turn all pots full on?



Green is about 60% and red is full on with blue...nothing.

Any particular reason for doing this? See I was told to never run LEDs in parallel for...many reasons. Even if you ran 12 red LEDs in parallel, they have minutely different vF values so that some will be slightly brighter than others.

[adafruit_support_bill]

Why are the LEDs wired in parallel?

They are not. Each has it's own cathode pin. If they were wired in parallel (or even in series for that matter) you would not be able to control them individually.

Red leds are considerably more efficient than green & blue. This is easily balanced by adjusting the PWM levels. Or, if you prefer, you can add about 100 ohms in series on the red cathode pin.

[32768]

There has to be a reason for the behavior where if red and green are full-on, that red will step over the other colors.

Here is a video of where I rewire my own ground around Pin 16 (Red). The output is yellow. Perfectly normal. Then I connect the ground lead to Pin 16 as if normally and the output is no green, but all red.

https://www.youtube.com/watch?v=k9-AxCObTAI

I also manually wired a ground around a setup of full blue and full green. With red full on, the output was white, perfectly normal. But going through Pin 16 for a ground for red meant the output was all red and no other color output.

Whereas if I rig up a red, blue, and green LED (Or a RGB LED) where the voltage goes straight into the positive pin and control the negative leads with three different pots, then I can get red, green and blue all showing up as either white or full on. None of this silly red-overstepping problem. After all, if that WAS the case, wouldn't we have a hard time with RGB LEDs in the first place?

[adafruit_support_bill]

The reasons are as noted above. That is the way these displays work. The red imbalance can be controlled via PWM or some extra series resistance.

[tldr]

if the leds have typical voltage drops of. say, 2v for red and 3.3v for green and blue, isn't your red led passing about twice as much current as the green and blue? that ought to make it a tad brighter.

[32768]

If it's not a case of parallel LEDs then it's a case of the 100Ω resistor forming a total current draw limiter for all the LEDs.

At any rate I desoldered the 100Ω resistor and replaced it with a solder bridge. After putting the proper resistors on the negative (cathode) side it works great. No silly red over stepping all the other colors.

AFAIK the variant at Mouser equivalent doesn't have a CA (common +) but is instead common ground and I don't know if it has proper LEDs.