Got stuck on a stubborn LRU fault that came down to two test methods

Had a nav unit on a Citation that kept throwing a 429.1 code, intermittent but enough to ground it. The book said to start with the built-in BITE test, which came back clean every time. After three hours of that, I switched to a full signal injection test with the bench rig, feeding known good waveforms into each pin. Found the issue in under twenty minutes, a tiny drop in voltage on the ARINC 429 output that the BITE just wasn't sensitive enough to catch. The BITE is fast and easy, but it only checks what the box thinks it sees. The signal test is slower and needs more gear, but it shows you the real picture. When the fault is sneaky, which method do you guys trust more, the quick internal check or the full external signal path test?

4 comments

4 Comments

aaron30520d ago

We had a similar thing on a King Air's autopilot last year. The cockpit indicator would just blank out for a few seconds in cruise. The system self-test passed every single time, no codes stored. It took us forever to finally hook up a scope to the actual pitch servo command line. Turned out to be a cracked solder joint on a board connector that only lost contact during certain vibrations. The box's own check was completely blind to it, just like your nav unit.

nancyj1120d ago

Wait, the BITE test passed every single time on a fault that was grounding the plane? That's terrifying. It's one thing for it to miss a tiny drop, but to give a clean bill of health while the thing is completely broken is a whole different level of bad. Makes you wonder what else it's happily ignoring while you're flying.

william_harris20d ago

Oh man, that's the classic story right there. The box lies to itself and then lies to you. I've wasted so many hours chasing ghosts because the built-in test gave me a happy green light. It's like asking a kid if they cleaned their room, they say yes, but you still have to go look under the bed. For anything intermittent or sneaky, I'm pulling out the gear and checking the actual signals every single time. The extra setup is worth not having to do the job twice. What's the point of a quick check if it's wrong?

henryp4020d ago

Remember when we all thought the built-in test was the final word? I used to trust those green lights completely until a nav box had me swapping a perfectly good antenna. The unit swore it was fine, but the actual signal was full of noise. Now I treat the self-check like a first guess, not a fact. If something feels off, I go straight to the scope or meter to see what's really happening.