I’m having a strange issue with multiple Broadlink Mini 3 devices whereby learning the same button results in different codes being captured.
I first noticed this while using broadlink.learn to capture Hitachi AC remote commands which are long and complex so I repeated the test with simpler TV, receiver, and ceiling light remotes as well as using a different capture tool, Broadlink Manager, but the codes are always inconsistent. I’m not near any strong sunlight nor ambient IR signals AFAIK.
Here’s a sample of 2 captures of my LG TV on/off button.
The strangest thing is that so far every code learned, while different, still works. This rules out the possibility that these are “rolling codes” for security, as repeating a rolling code should not work.
If I were only controlling simple devices with immediate confirmation, like lights and TVs, I’d just ignore the differences and say “it just works”, but the AC codes carry many parameters which cannot be observed immediately (eg. temperature) so I’d prefer to understand why the codes change before putting in hours of capture efforts.
Is there a solution to solve this? I have the same problem. Some buttons of the same remote work and some give a differtent code each time. This is an example from my lg soundbar mute button from the /config/.storage/broadlink_remote_24dfa7d608c9_codes file (10 tries with 10 different outcomes);