1. Home
  2. Electronics Design
  3. RGB LEDs off PCB -- ideas?

RGB LEDs off PCB -- ideas?

All,

1.) I have 3 RGB LEDs (Red, Green and Blue) that I have on my printed circu it board.

2.) Each LED is getting 3.3V, has a current limiting resistor and get's con nected to my microcontroller to sink the current. There are three separate connections to the microcontroller for each RGB LED.

3.) I have a requirement that now I need to move the LEDs off the board to a different location in the product. If I just wire things up directly I w ill have 11 wires that I will need to brought up from the PCB to where the LEDs are. (3.3V, Ground, Red1, Green1, Blue1, Red2, Green2, Blue2, Red3, Gr een3, Blue3).

4.) We looked at some "smart LEDs" but they have a non-standard digital int erface that will require bit-banging from the microcontroller that will che w up our development time.

5.) We looked at some light pipe options but we have some tight mechanical constraints and these seem they won't work for us.

6.) Do you have any ideas on how I could minimize the wiring and still use a standard interface?

Thank you.

-Bob

Reply to
electronicsman2016
Loading thread data ...

SPI, I2C and Dallas 1-wire parts will all do things like this, as well as just plain shift register type drivers.

Mostly depends on what your micro will support.

Reply to
Adrian Jansen

Half-pitch ribbon cable?

Multiplex?

Are you willing to go serial, namely SPI?

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com
Reply to
John Larkin

On Thursday, June 30, 2016 at 6:41:38 PM UTC-4, snipped-for-privacy@gmail.com wrot e:

o a different location in the product. If I just wire things up directly I will have 11 wires that I will need to brought up from the PCB to where th e LEDs are. (3.3V, Ground, Red1, Green1, Blue1, Red2, Green2, Blue2, Red3, Green3, Blue3).

e a standard interface?

Not quite enough information so this may be way off, but could you move the microcontroler

Reply to
dcaster

On Thursday, June 30, 2016 at 6:41:38 PM UTC-4, snipped-for-privacy@gmail.com wrot e:

o a different location in the product. If I just wire things up directly I will have 11 wires that I will need to brought up from the PCB to where th e LEDs are. (3.3V, Ground, Red1, Green1, Blue1, Red2, Green2, Blue2, Red3, Green3, Blue3).

Could you move the microcontroller off the board ? You might end up with two microcontollers. One on board and one with the LED's. You might end u p with two pcb's.

Dan

Reply to
dcaster

Yes, there are tons of parts designed just for driving LEDs with PWM brightness control, blinking, etc. A single SPI or I2C interface will drive a number of chips that each drive a number of LEDs.

--

Rick C
Reply to
rickman

So, these are four-terminal devices (common anode)?

They *share* the ballast resistor? I.e., one resistor feeds three (R+G+B) LED's? So, turning on more than one emitter in a package is dubious?

Why do you need to supply ground? I thought you were sinking current through the microcontroller (for each of the {Red,Green,Blue}{1,2,3} leads)

Are you able to put some electronics AT the LED's? E.g., a shift register could cut your wiring needs to +V, GND, Data, Clock. (if you can clock it quickly, you don't even need a "latch" or "enable" signal to present the data to the actual lamps)

Or, you could put a small microcontroller there and offload the responsibility for driving them, entirely (i.e., send the information that you normally use to drive the LED's to that microcontroller and let it do the flashing, dimming or you need).

Reply to
Don Y

A matrix arrangement would only need 6 wires (anodes 1/2/3, cathodes R/G/B).

If you can use both common-anode and common-cathode types, you could connect them in anti-parallel pairs to halve the number of matrix rows (5 wires for 3 or 4 RGB LEDs).

With tri-state outputs, you can use "charlieplexing" to drive up to 4 RGB LEDs with 4 outputs. For each LED, connect the common anode to one output, one cathode to each of the other three. Cycle through the LEDs in turn, driving the anode high, the cathodes of any lit elements low, with any unlit elements at high-Z. This requires that the drive voltage is less than twice the lowest forward voltage to avoid spurious (albeit dim) illumination of other LEDs.

Reply to
Nobody

On Thursday, June 30, 2016 at 6:41:38 PM UTC-4, snipped-for-privacy@gmail.com wrot e:

cuit board.

onnected to my microcontroller to sink the current. There are three separa te connections to the microcontroller for each RGB LED.

o a different location in the product. If I just wire things up directly I will have 11 wires that I will need to brought up from the PCB to where th e LEDs are. (3.3V, Ground, Red1, Green1, Blue1, Red2, Green2, Blue2, Red3, Green3, Blue3).

nterface that will require bit-banging from the microcontroller that will c hew up our development time.

l constraints and these seem they won't work for us.

e a standard interface?

You can multiplex 3x3 with six wires, double that if you make the wires bi-directional, etc., but any of it requires bit-banging.

Might as well use a shift register or your i2c LEDs.

Cheers, James Arthur

Reply to
dagmargoodboat

formatting link

Cute.

Cheers, James Arthur

Reply to
dagmargoodboat

Let the LEDs where they are and use an optical plasic fiber. Cheers, Reinhold

Reply to
rfa

I conjecture that it's likely possible to connect up two 7 segment LED drivers, like a 74HC4511, to the anodes and cathodes of the three RGB leds, and drive them both with the same 4 BCD control lines connected in the proper fashion, such that all 27 permutations of LED colors can be addressed using 5 lines total...

Reply to
bitrex

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.