@LagaffeVFR5476 Thanks for your feedback, but in the meantime, I found my “glitch”. In the formula, the “10” should be “l0” (character “L”). And now it makes more sense. sp0 stores value 1 and pops it from the stack, and l0 loads the value from register 0 to the top of the stack. With other words, 1 sp0 l0 does the same as just 1. These kind of weird constructions are the result of the RNP compiler that is used by FBW (at least, that is what I understood from some explanation I got on Discord). Now it makes sense, because I don’t compare the EFIS value with 10, but with 1. And this is the CSTR button.