Originally Posted by
pitz
No, the 737 does not have a full authority flight computer. The A320, 787, 777, etc., does. The 737's pilots can turn off the computer(s), and fly the airplane. A320/787/777/etc. pilots who pull the circuit breakers for the computers are left with a plane with extremely limited function. The 737's systems are best thought of as an overlay onto the existing analogue system of pulleys and cables, while the systems used in true fly-by-wire airplanes fully replace mechanical linkages with computer messages.
MCAS is flawed for multiple reasons, including, but not limited to, the use of inherently non-fault tolerant instrumentation (ie: only 1 AoA sensor, instead of 3), excessive authority required for MCAS to meet certification requirements, and of course, a computer control architecture that was not fault tolerant.
They might have to given that the design basis of the autopilot (its an autopilot -- not a 'flight control computer') did not anticipate that it would be required to operate in such a way that it could not be cut out/disabled and flown manually.
https://fortune.com/2019/08/02/boein...trol-computer/
"
Boeing is working on new software for the 737 Max that will use a second flight control computer to make the system more reliable".
(in contrast, the A320 has, I believe, 5 flight control computers running in parallel, with different CPU architectures and codebases to avoid the possibility of common-mode faults...).
"All 737s"? Really? I guess if you consider pilots to be 'computers'.....
Here are references to the architecture of the 737 flight control system:
Automatics,
Stabilizer Trim and
737 MCAS - Failure is an Option. Okay, I misspoke when I said "all 737s", but the references explain the "dual-dual" flight control computer architecture, starting in the NG (if not also the Classic) and continuing into the MAX. The references don't describe the differences that were introduced on the NG, relative to the Classic. In any case, there are two flight control computers, each comprising a pair of dissimilar processors. Is it the same as a modern architecture? Obviously not, but that's not the issue. The issue is that the MCAS function was not originally implemented to use the full redundant FCC and AoA sensors.