Autopilot behavior not linear using HDG or NAV

After the 11/24 update, when setting the heading bug or capturing a flightpath, the turn rate is not linear. Try this test:

  1. From straight and level with autopilot on make a 180 degree turn using the heading bug.

  2. Start a timer.

You should reach 180 degrees from your start point in 60 seconds (standard rate turn.) Initially, the plane (in this case the C172 with steam gauges) banks as it should and turns at three degrees per second. As it gets within 30-40 degrees of the requested heading, the rate of turn slows way down and takes another 15-20 seconds to complete the assigned turn to the selected vector.

This non-linear behavior is not correct, and did not exist prior to this latest update.

Are you sure about the AP and a standard rate turn?

This would mean that the AP has to vary the bank angle with speed and it would be impossible to use the AP at higher speeds than ~200kts. (assumed 30Ā° bank angle)

Most autopilots bank with a fixed 25Ā° bank angle, regardless of the speed.

Standard rate turns are normal when flying an approach plate. In this day and age of RNAV and GPS they might be of less significance, but thereā€™s a reason you practice turns-about-a-point in flight training. Flying a missed approach requires standard rate turns as well, and you need to do the same when flying ā€œthe racetrackā€ when holding for release into the sequence for landing at a busy airport.

The new AP behavior is not correct.

None of this matters at cruise speed or altitude, but try to fly an approach with this sluggish new behaviorā€¦

Iā€™ve seen this as well after the update on most of the planes Iā€™ve flown. The rate of turn slows right down to an unrealistically slow crawl as you get close to the heading. I sometimes have to override the autopilot so I donā€™t miss the turn.

3 Likes

As for the realism of the AP using a standard rate turn: Autopilots in many light GA aircraft, such as the KAP140 in the steam gauge C172 or the S-Tec Fifty Five X, have no choice but to fly some percentage of a standard rate turn. They donā€™t know the bank angle. They just use the turn coordinator to control bank.

Whereas on the other hand I guess a more sophisticated autopilot like the KFC 225 that uses an attitude gyro for bank has no way of flying a standard rate turn since it has no idea what the turn rate isā€¦

Just looked in the 55X manual, in GPSS mode its limit is 130% of a standard rate turn. That seems a bit enthusiastic at 200 knots?!

Back on topic, the behavior you describe sounds very weird.

Thatā€™s exactly itā€¦ the autopilot should mimic the turn coordinator. I donā€™t know that Iā€™d care much about a two minute turn at 200 knots, but at pattern altitude and speed, to help with flying a procedure itā€™s important. At least thatā€™s how I learned, in the days before RNAV and GPS.

1 Like

I havenā€™t flown with RNAV and GPS in the first 20 years of my airline career and standard rate turns have never been used.

Wouldnā€™t make much sense, since most AP are using 25Ā° bank angle.

When flying manually you are also simply using between 25 and 30Ā° bank angle.

Most airliners donā€™t even have a turn coordinator.

If the SID or missed approach etc. is critical, the procedure mentions a speed limit and/or minimum bank angle.

Thatā€™s what I remember from my days of learning to fly and going for my IFR rating. I am strictly a CAVOK VFR pilot these days; havenā€™t flown IFR as PIC in a long, long while.

For me, the biggest issue is the lack of linear behavior turning to a heading after this upgrade. Thatā€™s simply not right or correct.

We can argue about standard rate turn or not all day long, but the current behavior is absolutely maddening.

The AP should roll at a reasonable roll rate to some reasonable bank angle, turn to the desired heading at a constant turn, and then roll out at the same reasonable roll rate. This isnā€™t rocket science and it was perfectly acceptable before.

If this is hard for Asobo to figure out, they should ask their partners at Garmin whose name they put in the glass cockpits of all these aircraft for some assistance.

7 Likes

I agree completely.

I tried to do a circling approach today, with timed legs & turns, on the autopilot due low cloud and vis.

It was completely impossible due to the uneven turn rate of the autopilot. I had to fly it manually.

2 Likes

Worth watching the Q&A from last night where they explain why theyā€™ve done what theyā€™ve done and talk about further tweaks to come to it

1 Like

100% agree, and that was my intention when starting this thread. The behavior was fine before, and now it is not.

Three times now for me: disengage autopilot and hand fly the approach. The latest into KLMI with a 600 foot ceilingā€¦ good practice I guess. :slight_smile:

1 Like

Oh excellentā€¦ Iā€™ll have to watch that. Maybe some answers are already there.

I donā€™t know if it makes sense though!

Thereā€™s no way to fly a timed hold or procedure turn with this now. Itā€™s totally useless for IFR procedures. It worked fine before, P. Gatcomb did a whole series of videos (none of which can be followed now)

If you look at the development roadmap you will notice that there will be a new feature discovery series about aerodynamics in January.

Looks like they are definitely not finished trying to considerable improve their aircraft. :slight_smile:

1 Like

Iā€™m surprised that more people have not mentioned this issue or upvoted it. I am not one to grumble on the forums but this is a bewildering step backwards, especially when trying to fly IFR and especially in an update where they said that they were working on improving the autopilot in particular. In a 180 degree turn while in heading mode the aircraft rolls on angle of bank to 25-27 degrees at an acceptable, normal rate (achieved by about 10-12 degrees heading change). As soon as it gets to 50 degrees before the desired heading, the angle of bank decreases at an ever decreasing rate. I have watched and re watched what Seb has said in the Q and A about the improvements to the autopilot. All I could glean that related to this in any way was that the AP should pick up a track more accurately now but I really have no idea why it cant just roll off the angle of bank at the same rate at which it rolls it on like before. I always try to be patient while people much smarter than I sort out the various issues in the sim but I canā€™t lie, this one is such a glaringly obvious issue and itā€™s bothering me.

4 Likes

Very well said. I feel exactly the same wayā€¦ itā€™s the first time Iā€™ve really felt the need to point out something that got broken, which was working, prior to the update.

Iā€™ll continue to be patient though, and work around this new wrinkle. Iā€™m feeling Thankful today that we have a good new Flight Simulator and a Forum to discuss such things.

:turkey: Have a wonderful Thanksgiving!

3 Likes

I had to hand fly my last approach in the TBM 930 because of the autopilot ā€˜fixā€™. I hope that they correct this soon.

I started a 180degree turn in the C172, went downstairs, made dinner, ate dinner, washed up and returned to the monitor to see the AP finish the turn.

Ok exaggerating a little bit, but seriously yet again a Patch has completely ruined something in the Sim, for the love of God please please please thoroughly test your patches prior to release, every patch this is happening now, you must do it better.

5 Likes