Hey, everyone,

I have a project I need help with...

Let's say you have a bunch of LEDs on a simple power circuit, with your supply voltage and ground, two conductors. What I want to do is control each individual LED (brightness, dimming, on/off and color) instead of centrally controlling the whole line. I was thinking about multiplexing a digital control signal over the top of the supply voltage to a chip that is addressable and able to control its companion LED. Is there some small chip out there that would work for this?

Thanks!

 

Let's say you have a bunch of LEDs

How many is a "bunch"?

 

How many is a "bunch"?

Oh, let's say 16... Think "Christmas lights"...

 

WS2812B. You can control 100’s of them from a single signal, full color and brightness.

 

Here is my crazy version of that idea

 

Here is my crazy version of that idea

That's really cool!

Why was it necessary to program with light?

Why did the LED units need to be light-sensing?

Wasn't there another way to address programming the array?

Finally, couldn't your PCBs have been mounted on edge, perpendicular to the foam board, allowing tighter spacing? (Sure, it would require a PCB redesign...) If the cost of the prototype PCBs was low enough, this would have allowed production without the single-chip integration hurdle. (And still allowed the optical programming method.)

Really cool idea, sir!

 

WS2812B. You can control 100’s of them from a single signal, full color and brightness.

Hey, thanks!
Looking at the datasheet got the WS2812B, I see four terminals: power, ground, data in and data out. That tells my pea brain, "Hey buddy, four wires!"... Is there a way they're able to multiplex data on two wires?

 

It is 3 wires between each. The out of one goes to the in of the next in the chain. They cannot be connected with only two wires. There may be others that can, using a one-wire type protocol, but I am not aware of them.

Why is that a major hurdle? Three wires coming from your controller can control 100s of lights.

 

That's really cool!

Why was it necessary to program with light?

Why did the LED units need to be light-sensing?

Wasn't there another way to address programming the array?

Finally, couldn't your PCBs have been mounted on edge, perpendicular to the foam board, allowing tighter spacing? (Sure, it would require a PCB redesign...) If the cost of the prototype PCBs was low enough, this would have allowed production without the single-chip integration hurdle. (And still allowed the optical programming method.)

Really cool idea, sir!

There are lots of non trivial problems to be solved here, if there was a simple solution, this would already exist as a product.

The WS2811 comes close, but still suffers from limitations that prevent it from being used in more applications