Wemos ESP2866 Burnt When Connecting WS2812b LED Strip

Some time ago I purchased a Wemos ESP2866 module to use as a wireless controller for a WS2812b led strip.
.

After soldering the goldpins I connected it to the computer. Smoke went up. I quickly unplugged the unit, but noticed no damage (apart from a stain near the orange capacitor). I changed the cable, reconnected it to the computer.

I managed to upload the WLED software. That is, by this point the board was relatively functional. I proceeded with the next part. I tried to connect to the WLED but was unable to, the board was going out. The hub has an auxiliary power supply, so I decided to switch to the charger from the phone. Connection diagram below inspired by what I found on the WLED website (https://github.com/Aircoookie/WLED/wiki)

Shortly after connecting, the board died on the amen. Heavily heated, the miniusb port pulled out of the board when trying to disconnect. I am unable to verify if it is still bootable.

Photo of the damaged board


What went wrong? Is a manufacturing defect to blame? Or did I make a mistake later and "hit" the board when connecting to the power supply and led strip? If I did something wrong, I would like to avoid the same mistake in the future

How is the ESP board powered? I am referring to the positive pole of the voltage. According to your drawing, it is fed from the level converter board, not from the power supply.
Also, if I'm not mistaken, you didn't connect +5V from the converter to the Vin of the ESP board only to, depending on whether the board is shown from the side as in the picture or the other side, to D0 or A0. The Vin is near the USB port.

Something more about this level converter and the power supply, will you write ?

PaintedMan wrote:

Connection diagram below inspired by what I found on the WLED website (https://github.com/Aircoookie/WLED/wiki)

.
Not exactly a successful inspiration. The VIN in Wemos is the power input before the stabiliser on the board and certainly should not be connected to the "5V" pin of the level converter - in such a configuration this converter will not work. You should use the "5V" output from the Wemos board for this, but not all models of this board have this pin out. In this situation, the "5V" pin of the level converter is connected directly to the +5V supply.

If you are simultaneously connecting power to the Wemos via USB, it is safest to forget about using the VIN pin. It is possible that you have stoned the AMS1117 stabiliser.

Grzegorz Markowski wrote:

How is the ESP board powered? I am referring to the positive pole of the voltage. According to your drawing, it is fed from the level converter board, not from the power supply.
Also, if I'm not mistaken, you didn't connect +5V from the converter to the Vin of the ESP board only to, depending on whether the board is shown from the side as in the picture or from the other side, to D0 or A0. The Vin is near the USB port.

.

The board was powered by a microusb. Although I also suspect some problem with the cable because on the first one (the one I used when the smoke first went) the module was not detected in the device manager. Although I find it hard to believe that a faulty cable would cause such damage.

The vin was supposed to be a 5v source for the logic converter, as I read that it can be either a power source for the board (but you can't use it as a source at the same time as connecting it via usb) or as a source of voltage for the components connected to the board (provided the current requirement is within esp capabilities)

The damage I noticed first: the square element marked AB1742 where the indentation in the pictures is visible in my case is an exposed leg. You could feel the components heating up strongly near the 3.3v outputs Then the orange capacitor was burnt.

emarcus wrote:

The 3.3V high state from the ESP is sufficiently recognised by the WS2812B in its 5-Volt logic

.
I guess it also depends on the particular unit. According to the WS2812B note:

.

The minimum V IH is 0.7V DD , which on a 5V supply gives 3.5V, so it's not so sure. I'm not saying it won't work without a converter, but I'd rather suggest using one - the cost is small.

I have read that without a logic converter there are voltage fluctuations which cause flickering so I figured since I have a converter there is no reason not to use it. However, going back to my initial question, could the cable have damaged the board when first connected? For future reference I will avoid using the VIN as this may have caused additional problems.

What would be a better alternative to such a board? Are there perhaps some based on the ESP8266 but having 5v logic on the pins, like the arduino nano?

PaintedMan wrote:

could the cable have damaged the board when first plugged in?

.
If the cable was previously used and in good working order, it is more likely that the USB socket on the board was already "tref".

PaintedMan wrote:

Are there perhaps some based on ESP2866 but having 5v logic on the pins, like the arduino nano?

.
I haven't come across these types of modules.

emarcus wrote:

I apparently came across 'super' copies (all 60 in the strip) !.

.
It was enough that the first WS2812B on the MCU side was working correctly i.e. its DIN input was working correctly with 3V3 logic. DIN and DINOUT are not directly connected to each other in a given WS2812B circuit.

Thank you for your help and comments, topic altogether to be closed.

khoam wrote:

It was sufficient that the first WS2812B on the MCU side worked correctly i.e. its DIN input worked correctly with 3V3 logic. DIN and DINOUT are not directly connected to each other in a given WS2812B circuit.

.
Exactly!!!
The trick described https://hackaday.com/2017/01/20/cheating-at-5v-ws2812-control-to-use-a-3-3v-data-line/ has worked reliably for me for several years now