WS2812b work with NodeMCU, but not with Wemos D1 mini

I have an addresable LED strip which works perfectly with my ESP8266 NodeMCU (it is running a rainbow sketch), but the same code does not work with a Wemos D1 mini. I have adjusted the PIN number of course, and when I keep pressing the reset button the strip changes color, but it seems to stuck in the first milliseconds. It does not work with any other pins or sketches. I am using the FastLED library.

Any idea?

tosoki_tibor:
Any idea?

Yep.

To reliably run a WS2812 or similar strip from a 3.3 V system you may need to use a level converter to provide sufficient drive to the strip. If the 3.3 V levels are too low or the 5 V supply too high, they may not reach the threshold for the first WS2812. The two 3.3 V devices likely differ in the exact 3.3 V from their respective regulators.

The most practical level converter is a 74HCT14 with two inverters in cascade and the other inverters' inputs grounded. A 74HC04 is second choice.

I have tried a logic level converter like this: https://5.imimg.com/data5/UL/XU/MY-9380557/5v-to-3-3v-logic-level-converter-500x500.jpg

But did not work, or am I wrong to think it should have worked?

Simply not appropriate. It does not answer whether the problem is the logic levels or not, you still are none the wiser. :astonished: I have carefully indicated a suitable approach to a level converter.

I would also double-check the matter of the pin number.

I juist wanted to know if a logic level converted would/should work. I dont have the IC you have mentioned, that’s why I tried this.

Well, the logic/ voltage levels may be the problem you have, or it may be the code. Clearly the NodeMCU and WeMOS D1 mini are essentially, functionally identical so there is no reason the correct code would not work on either and there are no other explanations that I can see.

But the logic level converter that you cite has certain limitations whereas what I have suggested performs the task reliably. The 74HCT14 in particular is specifically designed to work with a 3.3 V input logic level and 5 V supply at full speed, so if it did not work, any possibility of the problem being the logic levels, would be completely excluded.