View Full Version : B737 A/P pitch software

7th Mar 2010, 18:20
Ever since I began flying the 300 series in 1988 I have been convinced that there is a bizarre minor software bug in its pitch control in A/P and I wonder if this contributes to the much-stated abysmal speed control in descent.

Before I start, I must caution that you need to be one of those pilots who is blessed/cursed with the ability to detect small changes in a/c motion. I know from several discussions here on crosswind take-off aileron techniques with 'he who got himself canned' that there are some pilots who fly by numbers, and some, like me, who sense minute a/c motion from an amalgam of sensory inputs, visual, balance organs and audio, which probably accounts for the fact that my sim performance has for ever been far worse than my actual flying ability. :)

If you are one of the latter, try this experiment on a nice visual horizon.

In a steady, LVL CHG descent, wind the speed up or down a big chunk - it matters not which - and watch the nose pitch motion VERY CAREFULLY at the beginning. I'd be very interested to hear what you see.

Out of interest, we hear nothing here from the AB gang on this A/P mode - does that mean that the AB does not have the wild speed tendencies of the 737?

7th Mar 2010, 19:35
Just a thought, would you rule out possible small aerodynamic flexing of THS initially providing opposite effect until elevator authority takes over?

FD (the un-real)
Not rated on 737 anymore.

7th Mar 2010, 20:44
Impossible for me to answer, really. Do I take it you know of what I post?

7th Mar 2010, 21:08
Well actually, I'd put myself into the first group of pilots. I certainly do not remember observing or being pointed out at anything unusual when on 737. I would hazard a guess based on what you had written but I see you prefer not to describe more until somebody else comes up with an observation similar to yours. It was your reference to AB (you discussed with AB friends and they say there's no such behaviour, correct?) that had me thinking what is the difference between the two. Freeplay and tensile behaviour of control cables was the first idea and second being that THS could flex with elevator movement with a 'control reversion' effect that may or may not be recognisable. Although this would apply for AB as well.

FD (the un-real)

7th Mar 2010, 21:25
BOAC, unlikely to be a bug, just a limit of the design / implementation of the autoflight system. See http://www.pprune.org/tech-log/407999-mach-overspeed-737-800-a.html#post5556610

It’s possible / probable that as the aircraft type has evolved, various ‘short cuts’ have been used in the development and certification.
Thus, when a ‘new autopilot’ is installed, some aspects of the ‘old’ aircraft mathematical model may have been retained, or the 'old' auto system muscle / rate limits retained in the autos design spec. There may be similar ‘adjustments / assumptions’ when adapting an existing auto system to a new aircraft variant, also when adding new FMS derived modes.
FMS modes may actually provide autopilot output demands; alternatively, the FMS and autopilot have to be harmonized, and in many systems, the autothrottle has to be similarly matched, but often only as an after thought.