0
\$\begingroup\$

My last post (Multi-Zone Addressable RGB+CCT LED Mode Switching) got closed as off-topic because I was “seeking recommendations for a specific product”... I found on my own what I believe to be the exact product I was seeking in that post and now I am looking for expert advice to validate whether I am on the right track with the overall design, if I am interpreting the data sheet correctly and wiring it up to produce the desired result.

My project, detailed extensively in the post listed above, calls for multiple “zones” of LED strips that need to be individually switched between RGB mode and WHITE mode. I am planning on using 12v WS2811 RGBCCT Addressable LEDs with separate RGB and WHITE controllers. There is one channel of data for RGB, and two channels of data for WHITE (cool + warm) - three channels in total that I need to switch. Each zone will have an ON/OFF/ON switch to control whether that zones LED strip is ON and receiving the RGB or WHITE data. All zones will be wired in series so that patterns can extend between them, and data needs to flow past zones that are off and continue through the rest of the zones - in other words, each input channel (RGB, WW, CW) has two output paths, it either goes into the current zones LED strip or it is carried down to the next zone.

Ultimately I determined that I needed one2-to-1 channel demultiplexer” for each of the three data channels - which led me to these CD4053B Triple 2x1 MUX/DEMUX chips apparently satisfying the “ideal solution” I outlined in my last post -> “small package, preferably a single chip that has three data inputs and three triggers that control three pairs of outputs.” The datasheet for the CD4053B chip is here. I've annotated a few of the key charts from it below.

"A" channel controls RGB data. "B" channel controls warm white data. "C" channel controls cool white data. Each channel has two possible outputs (X/Y) and its own selector pin to choose between the two. "X" is used when switch is OFF and data needs to flow to next zone. "Y" is used when switch is "ON" (activating selector pins) and data needs to go to the current strip. I noted in the description column how I plan to use each pin.

Pin Mapping for CD4053B

The diagram below is annotated with the pin numbers and functions of each, matching it to the table above. Functional Diagrams of CD4053B

So then the chip pin wiring would look something like this... again color coded and labeled to match above two diagrams. RGB+CCT Zone Switcher Diagram

  1. If the on/off/on switch is "OFF", pins 9/10/11 receive no power, so AX/BX/CX receive the three channels of LED data which flow down to the next zone**
  2. If the switch is in "RGB" mode, pin 11 (A channel) is powered switching RGB data to AY and it flows to the current zone. Meanwhile pins 9/10 are not powered (white is off) and the WW & CW channels flow out BX/CX to the next zone.
  3. If the switch is in "WHITE" mode, pins 9/10 (B/C channels) are powered switching WW & CW data to BY/CY and it flows to the current zone. Meanwhile pin 11 is not powered (RGB is off) and the RGB channel flow out AX to the next zone.

**Not shown above is another "Switch + CD4053B Combo" in front of Zone 2 to control its data in the same manner. Zone 1's data inputs (pins 4/14/15) probably come straight off the RGB & CCT controllers. Zone 2 (and subsequent zones) could receive data from two sources, for example:

  1. If Zone 1 RGB is ON, the strip will consume some of the RGB data and pass the rest down the line (depicted with purple lines).
  2. If Zone 1 RGB is OFF, the unmodified RGB data flows down the line (depicted with green lines).

In the diagram, this is where the purple data lines coming off the end of zone 1 strip meet the green data lines that skip down to zone 2. There shouldn't be a case where Zone 2 receives RGB data from both paths, so it seems safe to just combine the two source wires at the input of the CD4053B, knowing that only one would have data on it... although I would probably use diodes to ensure that the data doesn't "backflow" to a previous zone. Recommendations on the best way to protect against this are welcome if I'm overlooking something?

Also not shown above, I would probably use the +12v RGB & White triggers coming off the switch to actually power that zones strip as well. If a zones switch is off, its strip need not receive power. Additionally, when switching between modes, the strip would lose power, turning off its pixels, as to not retain the state of the previous mode (ie: white pixels staying lit when switching to RGB mode & vice versa). Each zone is only at most a meter long, so with +12v being injected at the beginning of each strip, they should be sufficiently powered.

My goal would most likely be to design a simple circuit board that accepts the necessary input and output wires and contains the CD4053B & any other necessary components. I will have 6+ zones in the van I'm building that each need one of these boards for ultimate control of each zone (as described in my last post).

Additional questions to the community:

I'm a novice "ic chip data sheet reader", and new to MUX/DEMUX so I hope I made accurate sense of how this device works! Looking for validation on a few things ->

  1. Does the CD4053B in fact work as I have described above and is it a suitable solution for switching 12v LEDs data streams in an automotive application (+12v to VDD power supply, channel selector, etc?)?
  2. What should VEE (“negative power”) be connected to? Is this just a second connection to ground?
  3. Do I need resistors or anything else anywhere in the circuit? Between switch and control pins maybe?
  4. Am I overlooking anything that would prevent this from working as intended?
  5. Are there "simpler" solutions or alternative chips that I should consider to achieve the desired effect?

Any and all feedback or questions are welcomed & appreciated. Thanks!

\$\endgroup\$

1 Answer 1

1
\$\begingroup\$

Answer to your Q4: The overlooked part is that the white LEDs are not driven by data but power so this mux won't work. Like it says on the link, it is ARGB+CCT. CCT is not data, it is power. This chip can't handle power.

The mux is intended for low current analog and digital signals.

It can switch very little current and it gets damaged after 10mA.

You are switching currents far above 10mA for the LEDs in the strip.

\$\endgroup\$
6
  • \$\begingroup\$ thanks for that feedback. The total current draw for several meters could certainly reach several amps, depending on brightness, etc. However, I cannot tell from looking at the specs for the strips or controllers how much current the data lines that I would be switching actually draw. Each pixel in these addressable strips shifts off a few bytes from the SPI data stream and then relays the rest of the instructions down the line. It would seem then that the data lines being switched would be relatively low draw, but I will do some additional research to see what I can find out. \$\endgroup\$
    – Matt
    May 5, 2022 at 3:08
  • \$\begingroup\$ You don't need further research. The white LEDs are not controllable by data lines but power so it won't work. \$\endgroup\$
    – Justme
    May 5, 2022 at 6:12
  • \$\begingroup\$ So then what would you recommend as the best path forward? Are there other ICs along the lines of this multiplexer that can handle the higher currents? Or do I need to look at replicating the multiplexer logic using some combination of multiple mosfets that I link together on my own to handle the higher currents? Something else? Thanks \$\endgroup\$
    – Matt
    May 6, 2022 at 15:25
  • \$\begingroup\$ The RGB LEDs likely keep the previous color sent to them. But to have multiple segments for the white LEDs, you need a separate CCT controller for each segment to keep their white LEDs on at the white tone you want. The CC pins are just the return current wires for the white LEDs to drive them with a constant current controller. \$\endgroup\$
    – Justme
    May 6, 2022 at 17:04
  • \$\begingroup\$ Grateful for your responses. I'm looking at the problem totally different now, realizing the WW/CW channels on the WS2811 strips don't carry data TO the leds, they are not addressable as I had thought. Instead all of the white pixels are simultaneously controlled by "low side" dimmers that blend warm + cool white to achieve desired color temp. In that case, I mocked up a new diagram @ i.sstatic.net/o5hZa.png that routes RGB SPI data TO strips similar to above and interrupts or connects the CW/WW return paths to the white controller to turn those on/off. Feedback welcome. \$\endgroup\$
    – Matt
    May 6, 2022 at 19:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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