737NG VNAV anomalies
 

Several times now when descending in VNAV Path, the aircraft has suddenly defaulted to the Flap limit speed no matter what the altitude or other restrictions inserted (xxx/FLAP) in the speed limit window.

It seems to happen mainly when going direct to the centre fix from some distance and at high altitude. All the decel points disappear and the bug goes to the 'Up' position. No changing of vertical or lateral path seems to make any difference. I have a suspicion it might be associated with changing to VNAV path when descending in Level Change in a Mach Number with IAS higher than the conversion IAS entered into the DES page but I would welcome any enlightenment on the subject.

Sounds like Geometric Path Descent profile management to me. Do you know if your FMC update has this option enabled? I'm not a GPD wizz but vaguely remember the FMC commanded speed bug moving to the 'Up Speed' temporarily to control the descent profile in VNAV with this option.

I think previous poster is correct! Flew 737's a few years ago and went through several FMS update cycles. Started out with a descent logic which was based on a level and dive profile to a continues descent profile.

Meaning: the old software, if you had a descent profile which was below an idle descent profile the FMS would program your descent so as to start descent towards the first constraint with an idle descent, then VS 1000fpm until the next idle descent.

A software upgrade of the FMS came and the logic was changed. The FMS would start an idle descent to the first alt constraint and from there on a geometric descent to any subsequent alt constraints. However, the FMS logic did not correctly take into account decelerating requirements!! Leading to some weird speed indications. Later software updates went to the old level and dive logic.

They way around it was, if memory serves me right, just before you hit an altitude constraint, delete the altitude constraint (before capture! You may miss an altitude constraint by 2 or 3 hundred feet) you will force the FMS into the 1000FPM until the next idle path descent and the idle path descent will take into account the decelerating part.

Hope this make sense and helps!

Pretty much "get on with it and discover"!

Some weird stuff happened,people wrote trip reports and within a short time a bulletin came out explaining why it was doing things and the solution I mentioned. I think, if memory is correct, that after a few months Boeing came out with a new software update going back to the old style.

I've seen the same and various other oddities with the VNAV logic.

The new geometric path logic will give you an idle descent to the first altitude restriction but it will then fly level until the TOD to the next altitude restriction. Not ideal for a continuous descent profile. Pressing ALT intervent (if fitted) and cancelling the restriction just before levelling off gives a 1000' ROD until reaching the path and does go someway to providing a solution, if a little inelegantly.

This wasn't quite the problem I've seen - the use of Alt Intv I'm fairly comfortable with. A clearer idea of the scenario:

Approx FL280 descending direct to a 10nm final at our destination, with say .76/280 in the DES page, with a restriction of 250/100, when without warning the speed restriction goes to 207/FLAPS, the bug goes to 207 knots (no UP speed at this FL!) and wants to start decelerating the aircraft. Thereafter the path keeps talking rubbish as it starts to tell us we are high, where if it were calculating for 207 kts, we are clearly low.

We just ignored VNAV for the rest of the descent, but my colleague was trying to see what he could do with the FMC to make it work. Every time he made a new entry, the target speed changed to something sensible, then straight back to the FLAPS restriction. The only altitude constraint initially in was 180B/3000A at the centrefix.

I'm mainly interested in how to get out of it, as doing an ILS it's no issue, but if one were to have the same problem when planning to fly a VNAV NPA, we would have zero confidence that the VNAV Path info was any good.

Putting in a new cruise altitude may sometimes help.

CHILD OF MAGENTA?
 

Classified as "TOO DIFFICULT!"

Engage brain, crunch some SIMPLE MATHS, engage MCP vertical modes and enjoy being in control?

Life is too short to start second guessing what computer programmers have been up to, especially when the green/brown bits out the window, or the green bits on the terrain display, are getting more obvious?

Yes it might be nice to know what the logic was, but I try to keep most of such curiosity about the FMC for the perusal of and education by these pages. :)

The point is not whether you can use VS or Level Change and the 3 times table - it's that there appears to be a bug somewhere in the software and I'd like to know both what might cause it and solutions to it. As I said, I would rather fly an NPA on VNAV than VS and this doesn't do much for confidence in the system. Does this bug only affect idle descent paths? Or does it also affect the FMC "On Approach" logic?

Can you confirm if your VNAV is set up for geometric approaches, approach geometric only, or old-style logic?

Did this happen when routing directly to the CF or IF for an airport?
Is there a further point programmed after the point you were routing to other than the RW__ point?

These may seem odd questions, but are fundamental in determining what went on inside the 'logic' of your aircraft's VNAV.

Routing direct to the CI point on the two occasions I recall, with at least the FI between there and the RW point. In both cases also from long distances (40NM+). I am not sure about the exact FMC logic - is it obvious from the software version?

I will check what the FMC IDENT page says when I fly next.

Are you sure you did not start your descent after the FMC calculated descent point, and entered Energy Compensation?

Secondly, in approach, the path will meet al the constraints regarding minimum altitudes, so you have no reason to worry about hitting something.

What you are describing JW is energy compensation mode. The jet is burning energy by going for the high speed but it drives the bug to flap speed. If you leave it without interference as the energy is burned in descent the CAB will return to the descent speed as long as the speed remains 5 knots below Vmo or 255 knots below 10,000 feet.

The jet has two descent modes, performance and geometric. Performance will hold height until a thrust closed descent profile is obtained. Geometric draws a line between two points an will average the descent rate between the two points.

Not all B737 have geometric path enabled.

C100 driver: in energy compensation mode, the bug speed will indeed go to F up speed, and the airplane, on autopilot, will do exactly fly that speed.

I'm still struggeling to get used to all this '737 black magic' going on. Why does it have to be over so 'complicated'? On my previous type (E-Jet) all one needed was a average descent wind component (HW/TW) and a/c weight. That generated a FLight Patch Angle (from QRH) and after some experimenting with it you could easily do a 120nm idle descent from TOD to the IF incl alt/speed constraints and nail it. The descents were done in FPA mode. Never once did I have anything unexpected. None of this 'what is it doing now/why is it doing this' crap.

I seriously hope that with the partnership between Embraer and Boeing the 737 Max gets a massive dose of 'Embraer sense and simplicity'. It's truely a much nicer aircraft to operate. And yes, I am working on improving my understanding of this aircraft. These are just my observations/frustrations so far.

Rant over, back to work...:ok:

It happens to me as well after passing the 250/100 restrictions inbound the CF, but i was creating a point on a right base 4 mile before with a spd rest of 200B, because there was no decelaration point.
And then the engineers told me that this restriction was the cause..

Having had a look, we have 107APP-P in the PERF DEFAULTS section. SOFTWARE OPTIONS are BCG-005-AB

This is very firmly NOT energy compensation - in energy compensation the speed bug moves, but the speed restriction at 4L on the DES page doesn't change. In this case, the restriction itself changes to 207/FLAPS for no good reason I can see.

The restriction was 180B/3000A as in Below 180KIAS, not a 15000' window, apologies for not making that clear.

In VNAV Path and on descent with the IAS bugging inappropriately to the Up speed just select LVL CHG and have your mate enter a Cruise Altitude of several Hundred feet lower than your current Altitude and execute.

This does solve the problem but I have had to repeat the operation a second time on a couple of occasions.

I will leave the whys and how's to the experts.

I should also add that this will change the descent speed back to a figure that is more appropriate. It may not be exactly 280/250 but you will now be able to ammend this descent speed and execute it.

Hardest part of the problem is convincing other people of the issue without it being discounted as being in an "energy effeciency mode".

I share your pain.