Okay, so we are talking about the surface mounted "NeoPixel" WS-2812B type LEDs.
I can't think of any way connecting a PIR could affect the LEDs. The PIR connection is not even very close to the LEDs so I don't think this is something you caused with soldering heat etc.
So all I can really suggest is to inspect the soldering of the LEDs very closely (with magnifier is possible) and see if any connections may have cracked or otherwise gotten broken, You could also just re-touch LED connections with your iron to reflow them but be careful and work fast - those LEDs do not like heat!
If that doesn't solve the problem then I guess you could try replacing the LEDs. Note that they are in a "daisy-chain" configuration so it's possible that only one of them is bad. If the LED under the 1-hour tube is bad or has a bad connection then the LED under the 10-hour tube also will not work.
Look into it later when the dust is clearing off the crater.
i'll do, i'm surprise as well, even because was fully working.
Just a question, because i did not see the diagram, the LEDs under the hours are driven by what? because may be the problem is not the led and before change (not easy) i want to be sure
and second question, there are a way to measure if there are voltage on the pin?
Those LEDs are not like normal LEDs where there is 1 pin for each color and a common. They are controlled by serial data in a "daisy chain" configuration. The get +5 power on Pin 1 and 0v on Pin 3. The first LED in the "daisy chain" receives color and brightness data on Pin 4 from the Wemos I2C pins. The first LED passes data to the second LED on Pin 2. And so on.
Here is the data sheet for the LEDs. It explains it better than I can. Look at the last page for a diagram showing what I mean by :daisy chain".