I’m new to this kind of project and I’m hoping you can help me out. Over the Christmas holiday I received a customized RGB LED lit headphone stand; it’s a piece of laser cut/etched acrylic that sits in a base that houses the LEDs that illuminate the acrylic design.
This LED base is nothing special just an IR receiver and a capacitive touch button with a microcontroller to control the LEDs. But as I was looking at it I got to wondering, could I mod it to make it smart? So as one does, I opened it up to look at the circuit board.
Here’s what I’m seeing on the board, it seems pretty simple: it’s got a micro USB for power, an unlabeled microcontroller that’s connected to the IR receiver and touch button and driving each of the three color channels on the 8 RGB LEDs:
I appreciate any insights or help you folks can provide on this; in my head it seems like it should be as straightforward as pulling the microcontroller off and soldering some jumper wires between the pads on the board and the respective pins on an ESP32 loaded with WLED firmware, but I’ve never done anything like this before so I wanted to see if there was something obvious I’m overlooking
From what little I can trace out in your photo, I would guess the 8 RGB LEDs are analog.
That means they all come on with the same colour/brightness, you can’t have a Red LED besides a Blue LED at the same time.
That implies the 3 small SMD transistors labelled “J3Y” are channel drivers (probably MOSFETs) with three 102 (1K) resistors driving their gates. A simple test would be to set up the light at full brightness an measure the voltage on the Red, Green, Blue pins.
Probably won’t be definitive (as it’s likely still PWM), but it should give you an idean of the voltage level on those gates. Even better would be 100% of a single colour so you can compare against another channel that’s off.
Once you verify the levels on the gates, you can move to trying to remove the CPU (if you’re gentle it should survive). You could then try some brief static tests by raising the R, G, B pads to 3.3V (or 5V) and seeing if that lights the corresponding colours.
After that, you can try to connect a WLED loaded ESP, make sure you also connect the ground line.
You might or might not need levelshifter depending on the MOSFETs employed.
Thanks for the info! To follow up on your comments:
The LEDs are indeed simple analogue ones, as you suspected.
The components you identified as MOSFETs does seem to be a transistor (at least according to a data sheet I found for it).
The voltages on each of the lines is all around 5v when I measure on my multimeter - the power is all coming from USB and there’s nothing on the board to do any voltage conversion so that makes sense.
The transistors you identified are actually BJT PNP transistors.
It’s likely they are connected on the “high” side of the LEDs and will activate a channels when the respective resistor input is pulled down to 0V.
That actually makes it easier to connect an ESP32.
You just need to check the “Inverted output” option when configuring the PWM RGB protocol for those LEDs.
As a follow up question, if you will indulge me, I’m seeing tons of different ESP32 boards available while researching. From what I can tell the WLED firmware just needs a GPIO pins per each color channel and pins for the IR data line and the touch button (ie 5 GPIO pins in this case). Looking at the boards that are listed as compatible it seems like the QuinLED-ESP32 would be a good choice given its small size and numerous GPIO available. Am I on the right track here? Or is there another board that would be better/faster/stronger/etc. that I should look into?
The KB is a reasonable synopsis of what has been shown to work with WLED.
IMO the Quin boards are all well designed and a solid choice for ready to go out of the box.
They have a proven record of reliability and functionality.
Others who may be more intent on the DIY route, like to stock up on the more bare-bones board versions and add all the needed bits and pieces themselves to suit their designs.
I fall into the 2nd category as I have some 40Yrs practice in blowing up parts until I learned something.
The advantage you can get today is:
Parts are waaaaaaaaay cheaper and more readily available.
The learning (and passed on experience) resources like this forum are a huge leg up for playing with electronics.
The performance issues with the ESP line are basically ESP8266 vs ESP32, and the general rule of thumb is you might as well go the ESP32 route unless you have a bunch of ESP8266’s lying around, that you really need to use up.
In the end - Welcome to the WLED Rabbit Hole.
Watch that 1st step, but after that it’s lots of fun all the way down
