Hitting altitude constraints in VNAV PTH for 737NG

[Donstim 109]

My experience has been that I'm always a bit high over waypoints containing altitude constraints on descent. The path appears to be predicting arrival at the altitude constraint correctly, but then when the airplane begins to level off, and the rate of descent is reduced for a nice level-off, I end up crossing the waypoint about 100-200 feet high or more.This is all under autopilot control on a VNAV path descent. Is this the intended behavior?Don S.

[downscc 1,327]

I can't answer, but I can add: I assume you're talking about a hard constraint rather than crossing a at-or-below altitude. I have seen it get pretty close to nailing the altitude, but in any event I don't think it matters to ATC as long as you're close and trending closer. I've been off at least that much (won't admitt to more) in the real world and ATC doesn't seem to mind as long as I am at the fix laterally. I do know it's real uncomfortable for pax to bottom out a descent too quickly... they really don't like even light G's.My complaint is with the soft constraints, say an at-or-below, where the FMC hangs up sometimes and most times fails to follow the path itself has set up after the fix to the next constraint.

[Donstim 109]

Thanks, Dan. Yes, it's at a hard constraint, and one where ATC has not yet given clearance to descend further. In other words, the MCP is set to that altitude as well, and a level off is required.I agree with your comment regarding the soft constraints. You really have to watch it sometimes.Don S.

[downscc 1,327]

<> ... and yet it settles down to the desired altitude at barely 200fpm.. this is also something I've seen. I agree, it is kinda sloppy.I've gotten kinda frustrated with the VNAV and rarely use it for descent anymore, I use VertSpeed with FMC providing guidance. I've read in this forum that most RW aviators do the same, and it makes sense to me because it keeps me in the control loop. I also gave up on VNAV near the top of climb, it just takes forever to meander up that last two or three thousand. Now I routinely use a constant speed climb once I hit machs 0.74-0.76... and it still is sluggish the last thousand but it's better. I always thought it was because the air is thin up there but now you've help me put the two together and I guess it's the control laws.