Originally Posted by
Dani
I repeat my post again gladly, where I wrote that it's not the MCAS which was the main problem but the data gathering unit (I'm not a B guy) from the AOA probe to the flight control computer.
I would have to look at the system architecture to be sure, but I'm almost positive that the AOA output does NOT go direct to the FCC. AOA is an input to both the Stall Management Yaw Damper (SMYD) computers and the Air Data Inertial Reference Units (ADIRU's). AOA is just one input into the SMYD's and ADIRU's. A bad pitot tube input could also have generated a false stall signal.
The primary responsibility of generating an approach to stall signal belongs to the SMYD which then activates various other aircraft systems (including the FCC) in response to the impending stall. I believe MCAS is a subroutine within the FCC. As I posted earlier, in many respects the two SMYD's act independently requiring only one "vote" to activate various systems, though there will be some other alerts to indicate the disagreement.
People insist that it was a probe failure, but this is highly inprobable. 2 probe failure on a new plane within months
Not sure if you are talking about the two AOA failures at Lion Air, or the two failures at two different airlines. At Lion Air, the accident aircraft had a defective AOA on a previous flight which was then replaced. We don't yet know why the first AOA was defective. Perhaps a ground worker bumped a piece of equipment into it. We do have evidence that the replacement AOA was 20 degrees out of calibration. Its DFDR readout exactly paralleled the good AOA, just 20 degrees higher. How this came to be is one of the subjects of the investigation. At Ethiopian, the DFDR data suggests that the AOA was working during the takeoff run, but was disabled shortly after liftoff, possibly by a bird strike.
I have never heard of a AOA probe failure, it happens very rarely. It must have been the AOA data that was corrupted, not the AOA probe itself. Only then MCAS made the mixup with the data (rubish in, rubish out). MCAS reacted as programmed, it received the wrong data.
AOA failures do happen, but until recently they did not result in a major accident. The DFDR records the AOA output, so we really do know that it was a faulty AOA, though the reason for the failure was different in the two accidents. Again, it wasn't MCAS's job to determine the stall condition. That responsibility belongs to the SMYD's. The problem was that it only took one SMYD to activate MCAS.