Glideslope problems

Airbusjoerg

Unfortunately, I have to confirm that the autopilot glideslope following is not actually usable. I have the additional problem that it can't really cope with extending the flaps. This leads not only to oscillations but sometimes it just ends up leveling out or at around 200 FPM or so - in the middle of the final approach - so it loses the glideslope completely and then no longer corrects the path at all. Sometimes it works. All in all, not comprehensible. I have not yet observed this kind of behavior with other addons and I have to agree that this really dampens the enjoyment of an otherwise excellent add-on.

Laurreth

I finally sat down and tried to record a demonstration, and apart from some minor oscillations the AP of course nailed it, whereas it was a complete dummy the flights before.

Airbusjoerg

Must be a weird combination from weight, balance, weather, trim and perhaps timing (gear down, flaps, speed). Sometimes, at slow speeds the aircraft indeed feels a bit sensitive around the elevators but manageable.
Yesterday I had the worst case, where the autopilot corrected to around 0-200 FPM and stopped further corrections during approach. That was in EGLL, not much wind, no gusts, perfect configuration, turbulence even set to low… it was simply not able to hold the glideslope.

Highlander_821

What's truly bizarre to me is that the JF F-28 is perfect every time. Subject to the same variables as the 146, including the MSFS turbulence and weather settings. Never a problem. But the 146 is absolutely terrible on the glideslope with the autopilot. Exasperating!

Laurreth

My test flight was very light to the point of the plane behaving more like the Hawk pretty much jumping off the runway, maybe it breaks down more under "simbrief conditions" pushing the CG back? There are definitely conditions that the AP can't handle.

Airbusjoerg

That is a good idea. I use always Simbrief and import the weights. Perhaps I should try some flights with standard-weight-settings.

Airbusjoerg

I have now tested this and have not changed the loading of the 146-100 (apart from the fuel). I can therefore rule out the influence of the "simbrief conditions".
My tests were carried out exclusively with the BAE146-100.

I can definitely say that the autopilot has a problem with

a) the transition from climb/descent to level flight at slower speeds
b) holding the glideslope

regarding a): at slower speeds (I tested in the range between 220-180 knots) when reaching an armed altitude and transitioning to level flight, the autopilot regularly start oscillating up and down. After a while (a few minutes), it recovers. At higher speeds, for example climb to cruise alt and transitioning to level flight it worked well.

regarding b): when following the glideslope, the autopilot either oscillates regularly up / down or it deviates upwards or downwards from the path and then no longer corrects this. For me, in 9 out of 10 cases it is not usable.

I have now made many flights with the BAE146-100 because it is otherwise a very nice airplane. At least one of the above problems has occurred in every single one of these flights. These were flights in all kinds of weather, with or without wind. I am therefore of the opinion that this is a permanent and reproducible problem that does not have to be proven by any records.

Mark

We do have this logged on our internal bug tracker and it is under investigation. At the moment our plan is to make these improvements to the RJ which is currently in development and then back port those improvements to the 146.

Mark - Just Flight

Airbusjoerg

@Mark Thanks!

Laurreth

FWIW, I've been running into those issues some more, but going into SYNC and quickly trimming out the aircraft fairly reliably "fixes" autopilot behaviour down the road (and the glideslope). Basically it starts happening when slowing down during arrival or approach, I trim the plane once, and it'll keep working the rest of the way.

I hope the RJ received some better logic there and we see it ported back soon :-)