The original project is designed for Arduino Pro Mini, ATmega328P, external quartz 16 MHz, 12 LED strips WS2812B each of which contains 20 LEDs.
I took an ATmega2560 - 12 tapes are too small for me, and in mege2560 you can hang up to 50 tapes and even more. But this is not the most glavnoe - the main thing is that you will have to compile the Arduino project in the same IDE version as the author, as well as with the same version of the FastLED library. I've tried compiling in the latest version of the environment to date (1.8.13) and the latest FastLED version (3.3.3) - there were errors. I changed the version of the Arduino AVR Boards medger to 1.8.3 - it also didn't work.
It compiles without any problems in IDE version 1.8.12 and Arduino AVR Boards manager version 1.8.2 under i5 Windows7 x64. Moreover, 1.8.12 IDE is not compiled on i7 in Windows10 x64, but it is compiled under i5 Windows7 x64-it may be related to the version of Windows, maybe with hardware.
The current firmware version provides for connecting 50 WS2812B strips of 30 LEDs each.
LOW - 0xFF
HIGH - 0xD9
EXTENDED - 0xFC
You can use a quartz resonator from 12 MHz to 20 MHz (I now work with quartz 20 MHz. It doesn't start at 10 MHz). At a higher frequency (even at 20 MHz), glitches and freezes are possible, so it is better to use the maximum frequency offered by the datasheet-quartz 16 MHz.
One note: if you use the current firmware (designed for 50 tapes) with 20 or 30 tapes connected, you will get pauses in the meteor lights, because all 50 are processed in the code, and you expect to use only 20 or 30 pieces. For a more beautiful picture of falling meteors, you should make changes to the code and comment out "/ / " with such a double slash on unused feeds.
P.S.: Two features were noticed:
1. If the cable length is more than 2m, the device will not start. Guaranteed start with a cable length of up to 2 m
. 2. Connected 20 pieces of 30 LEDs each. The cable length of any tape is not more than 2m. The device started, worked for 30-40 minutes, and the MK burned down. I re-soldered the MK, ran it, and got the same result. The point is in the current-limiting resistor, apparently. And Mega2560 is kind of gentle too. We solder the current-limiting resistors to 1k and test them.
I replaced the MK with an ATmega1284P. We use 24 feeds. In fact, I took 20. Current limiting resistor 470 ohms. It worked for two months without any problems.