FAO: Gabriel; Why don't [i]we[/i] just unplug MCAS?

[Gabriel]

I suspect that it is similar in the MAX as it was in the -200, which must be an electormechanicalhydraulicgravityspringpitotptressure analog thing, into which incorporating a feature that works with "logic" may be complex. The trim on the other hand is already controlled by the FCC (which is an electronic computer) on behalf of the autopilot and the speed trim. Adding the MCAS is a matter of adding some lines of code.

It's a lot cheaper to tweak an electromechanicalhydrograviticelastopitot thing than add some lines of code, though...

While I tend to agree with that in general, introducing LOGIC (of the IF ==> THEN type) into an electormechanicalhydraulicgravityspringpitotptressure analog thing (especially an already existing one) may be of a complexity well beyond "tweaking", and hence more costly than introducing lines of codes in an already existing program that runs on an already existing computer, where this new extra code will take already existing inputs that are already going into that already existing computer and, as output, activate an already existing motor switch that is already being controlled by said computer. It is 100% pure code. Not a single new processor, memory, sensor, interface, wire, or even connection is needed (apparently).

[ocelot]

It's a lot cheaper to tweak an electromechanicalhydrograviticelastopitot thing than add some lines of code, though...

While I tend to agree with that in general, introducing LOGIC (of the IF ==> THEN type) into an electormechanicalhydraulicgravityspringpitotptressure analog thing (especially an already existing one) may be of a complexity well beyond "tweaking", and hence more costly than introducing lines of codes in an already existing program that runs on an already existing computer, where this new extra code will take already existing inputs that are already going into that already existing computer and, as output, activate an already existing motor switch that is already being controlled by said computer. It is 100% pure code. Not a single new processor, memory, sensor, interface, wire, or even connection is needed (apparently).

That may be. But the history of computer-controlled systems is full of lessons where people thought adding some stuff in software was going to be easier and cheaper. And in some of these, people got in the way afterwards and they did died. In the absence of more detailed information (that probably nobody outside Boeing has) it seems like a questionable decision.

Also, if what you want is a feel adjustment, it's best _not_ logic but a continuous analog function computed from the angle of attack and a few other things, and it's probably more readily done electromechanicalhydrograviticelastopitotically. Though since it probably isn't linear it may not be particularly easy that way either.

BUT if what you want is a stick pusher, you want specific activation criteria and it becomes much simpler as logic. So, I wonder...

[Gabriel]

Also, if what you want is a feel adjustment, it's best _not_ logic but a continuous analog function computed from the angle of attack and a few other things, and it's probably more readily done electromechanicalhydrograviticelastopitotically. Though since it probably isn't linear it may not be particularly easy that way either.

BUT if what you want is a stick pusher, you want specific activation criteria and it becomes much simpler as logic. So, I wonder...

To be fair, things like artificial feel (continuous function) and even stick pushers (simple IF ==> THEN decisions) have been relatively simple non-digital "mechanisms".

I imagine that the logic of the MCAS was possibly more complicated. But I don't know.