Minnesota Crash

[aviatoreb]

3 hours ago, PT20J said:

That's easy during a training exercise, but more difficult under stress. The GFC 500 LVL will save you if the instrumentation is good but you just got distracted and lost control, but it may not help if the instrumentation is bad. In my new, about to be installed G3X/G5 there are three kinds of failures: Gross failures (like an accelerometer with no output) that are immediately obvious by a red X on the display, Recoverable failures that cause temporary loss of attitude information while realigning, and miscompare errors where the G3X and G5 don't agree and the software cannot tell which is correct. For this reason, I'm including a non-Garmin (so as to have technology diversity) AV-20-S. With three attitude indicators displayed, it will always be immediately obvious which is the problem and if that's the one that is driving the autopilot, I can revert the autopilot to the correct one.

Skip

I kept my already installed electric gyro lifesaver in place when I removed my ki256 vacuum instrument and put two gi275… even though gi275 do have revisionary mode if one fails.

[bradp]

22 minutes ago, toto said:

I may be oversimplifying, but in this particular case, a quick cross-check with a visible horizon before descending into IMC would identify the misbehaving AI.

Caution that unless you are truly in VMC.  I once got a case of the leans from visual cues / illusion due to layers creating a false horizon.  Cure was to go head down and stay there. 
 

I also recall one time turning onto final for an approach at the old IGX in IMC, I inadvertently kicked off the AP (CIIb without any alarm/indication of having done so).  Was hands off and looked down for maybe 2 seconds to reference my chart, simultaneously turbulence, and I’m in a 30+ degree nose down turn on resuming my scan.   Of course it felt like 1G the entire time. Practicing upset recovery in foggles is probably the most valuable part of any BFR/IPC. 
 

Things happen fast and often together.  IMC - even “gentleman’s IMC” - can be just as unforgiving. 

I agree with Don.   Remember that bo crash over Long Island?  That prompted me to ditch the TC and do multiple redundant sources of attitude information.  So now I’ve got an RC Allen electric, two G5’s and FWIW an AHRS in my GDL52R hardwired to an Aera 660.  Blue button + ENV protection  + redundant attitude info is a powerful cure to this stuff.

 

 

[whiskytango]

11 minutes ago, toto said:

Especially when they cost less than 0.8amu with a battery and require no installation. 

https://dynonavionics.com/pocket-panel.php

+1 on the Dynon Pocket Panel.  I have a backup vacuum source and a backup electric source, but I wanted an Attitude Indicator that was completely independent of the aircraft vacuum and electrical systems.  In a failure of the primary Attitude Indicator in IMC, your survival depends on making the correct responses within a matter of seconds.  I did not want to spend precious time diagnosing the problem and then energizing the appropriate backup system when I could simply look at my backup AI and know immediately what the proper response was. 

[A64Pilot]

5 hours ago, steingar said:

What I really don't get is the layer he was in was only 1000 feet thick.  How could he get those violent oscillations and not pop out the top into clear air?

I’m not understanding this accident either, the field was VFR, not matter what he did, he should have broken out pretty quick either up or down.

‘I’m going with a stroke or something, not a simple instrument failure. 

‘Of course it’s all just guessing with no real data to base it on.

People love to beat on Dr’s as poor pilots. but my experience is the opposite.

[tigers2007]

1 hour ago, toto said:

Especially when they cost less than 0.8amu with a battery and require no installation. 

https://dynonavionics.com/pocket-panel.php

Many have lost way too many of our fellow aviators due to not having one of these in plain view and operating when disorientation happens. I'm still blown away on the folks who seriously believe that they'll have plenty of time to pull out their cellphone/tablet and fumble fart with it until the AI displays, and only then attempt to regain positive control. Anyone who has had a close call with spatial disorientation knows that you really don't have time to burn like that. A backup AI like the pocket panel should be on someones to-do list along with shoulder harnesses, etc.

[DXB]

1 hour ago, tigers2007 said:

Many have lost way too many of our fellow aviators due to not having one of these in plain view and operating when disorientation happens. I'm still blown away on the folks who seriously believe that they'll have plenty of time to pull out their cellphone/tablet and fumble fart with it until the AI displays, and only then attempt to regain positive control. Anyone who has had a close call with spatial disorientation knows that you really don't have time to burn like that. A backup AI like the pocket panel should be on someones to-do list along with shoulder harnesses, etc.

Agree pulling out your phone during the stress of primary instrument failure in IMC isn't going to go well. But a newer, faster ipad that's already mounted and running Foreflight and talking to a Stratus AHRS, or similar setup, isn't too bad really.  I'd never suggest that as your main backup. But I found flying under the hood for practice with my Stratus + Foreflight AI and synthetic vision to be really pretty workable - particularly with the ipad mini 5's processor.  And at the end of the day, the best backup is the one you're most practiced in using. 

[carusoam]

Gents,

Seriously consider…

What it is like to be cognitively overloaded…

Then start thinking about assembling plan C…

Your brain is going to tell you to focus on what you have in front of you…

And put everything else on hold…

If your spare AI is deep in a menu, or a seat pocket…. You might ignore it too… expecting that it will only tell you what you know already…

Yet, you don’t know what has failed…

You will be yelling out loud… stay focussed… 

TCs have officially left the preferred list of equipment to have on an MSers IFR panel…

Two AIs, minimum for flight in IMC…

TCs, Yes you can, but you don’t have to…

 

This post is is more of a public service message… not much to do with this accident, yet…

Best regards,

-a-

[Hank]

I've had that self-talk during stressful moments in the cockpit! Fortunately they were due to failures between the left seat and the yoke and not to anything in the panel.

The first was a bounced night landing coming back from Thanksgiving weekend and into a college football TFR, where Approach vectored me right over the field. Did I say I had about 100 total hours, and had finished my transition & insurance dual on Labor Day? Quit flying in the flare, bounce, bounce, go around hoping to miss the trees that I couldn't see in the dark that caused a displaced threshold when landing the other way on the 3000' runway. I still remember chanting "make a normal landing, figure it out later" over and over, all the way around the pattern.

And when you get stupid in the clouds, you must trust your instruments, because you may really be wings level and not in a steep bank after leveling off in the clouds. "Trust the instruments, trust the instruments" while fighting to not turn the yoke, sweating hard and feeling the leans. Ignore those visions of flaming wreckage that prowl around the edges of your awareness.

If you're doing well in the clouds and look too long at your tablet, approach plate or even panel-mounted GPS (or have been studying all of them), if you hear a whistling noise immediately check the ASI then the AI, reduce throttle and level off. Once speed is reasonable again, increase throttle and climb back to where you're supposed to be, and correct course if needed. When you're too fast, all you want to do is slow down and gently level off.

Those are my lessons so far. But if the instruments are bad, better have a plan ready to go. If the AI shows a climbing turn,  but ASI is increasing and VSI is pointing down, you don't have a lot of time to figure it out . . . .

[jlunseth]

This is starting to make a little sense to me in light of information provided by the NTSB. According to the NTSB the aircraft was on an instrument flight plan and was on the ILS 10R into flying cloud. If you look up this thread to about page ten, 1980Mooney puts up an ADSB graphic with two waypoints, ZAMUD and ZIGCO. His post was considering whether there was some confusion between the two RNAV’s into 10R and 10L, but we know now that the aircraft was on the ILS, not either RNAV. What is signficant about his graphic is the it shows the location of ZAMUD on the pilot’s actual route of flight. ZAMUD is the intermediate fix for the ILS 10R. The accident site is after ZAMUD. At ZAMUD the pilot is supposed to descend from 2700 to make 2600 at the final approach fix, which is STUBR. At ZAMUD the aircraft was at 3000-3100 or about 400 feet higher than the 2700 intercept altitude. This would not be unusual if the pilot is being vectored onto the approach, he might have been cleared to maintain 3000 until established, for example. For whatever reason, the aircraft deviated slightly left of course after ZAMUD, so he would have found himself around 3 miles from STUBR, around 400 feet too high and left of course, and he would have needed to cure the altitude problem, in particular, in order to get on the ILS downslope. It was at that point he turned right (south) and descended. He got to 2600 at one point, which is the altitude he needed to be flying at STUBR but was slightly higher at 2700 at the end of the southbound turn. Altitude was still relatively stable at that point.

It was at that point that things went wrong very quickly. at 17:39:39 the aircraft turned north and after a couple of more seconds, began to descend further. It was now northbound, away from the approach course and toward the impact site. From there until impact was in the vicinity of 7-11 seconds

There are several possibilities. There could have been a vacuum or gyro failure. I have had both, and neither one happens in 10 seconds in a vacuum operated gyro. The gyro would not have been working well, but if it tumbled at that exact point there would have been symptoms and altitude variations before then.

Another possibility is a failure in the autopilot control panel. As an example, my AP is a KFC200. It has an Up/Down button. If the AP is in Alt Hold mode, pressing the down button will cause a descent at about 700 fpm until the pilot takes his finger off the button and then the AP holds the new altitude. I use this in just this situation, with a thousand feet or less to lose and a reason to hold altitude thereafter. This spring, this function failed. The Up/Dn button acted as a trim control, pressing the Dn button would continuously run the trim down and then hold the down trim rather than a new altitude, when the button was released. I discovered it during VFR practice approaches and got it fixed. 

The pilot also may have, either intentionally or inadvertently, disengaged the AP. The AP can be inadvertently disengaged by simply using the trim button on the yoke, which the pilot might have done in order to change trim and get down to the desired altitude. 

If the AP remains engaged and in Alt Hold mode but for some reason is losing or gaining altitude, and the pilot tries to correct with pressure on the yoke, the effort will work for awhile, but the AP will run the trim to counteract that pressure and the pilot finds himself working directly against the AP and needing to disengage to fix it.

At about the point where the southbound turn began, the pilot would need to dial the inbound course in with the course needle and switch to APCH mode on the autopilot. If he switched but had inadvertently failed to dial in the correct inbound course, the aircraft would immediately turn to find that incorrect course, which could explain the sudden northbound turn. The pilot would respond by disengaging the AP and hand flying, and making that transition with no notice is difficult.

These are all possibilities and there are certainly others. What is significant is that the plane was stable and on the approach until just prior to the FAF when the pilot would have needed to correct course and lose altitude quickly. Then something happened.

[Scott Dennstaedt, PhD]

16 hours ago, steingar said:

What I really don't get is the layer he was in was only 1000 feet thick.  How could he get those violent oscillations and not pop out the top into clear air?

Agreed.  Even with a shallow descent rate it should have taken less than 30 seconds to descend through a 600 ft layer of relatively smooth air without some kind of extenuating circumstances (e.g. control surface failure, incapacitation, etc.). 

[Scott Dennstaedt, PhD]

10 hours ago, carusoam said:

Gents,

Seriously consider…

What it is like to be cognitively overloaded…

 

Also an emergency checklist may also help with some of that overload in some circumstances...as well as practicing emergency procedures. 

Over the last 25 years of flying, I've had my share of weirdness happen in an airplane in IMC and I'd have to say, the checklist was invaluable in all those situations.   

[Raymond J1]

Rather, the question should be: Who has not experienced spatial disorientation in IMC condition ?

We are not all equal when symptoms present themselves and our different recent experiences do not make us react in the same way.

It is almost certain, however, that the symptoms are similar and that attention should be paid to the first sign rather than waiting for an alert from the plane. For the symptoms, some assume that the horizon veers to the left, others imagine that the bottom is on top, there are finally those whose first symptom is to imagine the failure of the flight, and more...

The distraction is not the only cause. There are also concerns about blood pressure, blood sugar. The control checklist is a good alternative yes, but the eye on the AI remains the best parade with a good mastery of its machine without AP.

Edited August 12, 2021 by Raymond J

[Marauder]

It is always difficult to put yourself into the accident pilot's place. I think all of us want to know what caused the accident and more importantly what learnings we can take away from it.

I will mention one topic related to pilot proficiency. And PLEASE don't read into this that I think this was a pilot proficiency issue that caused this accident. We just don't know at this point. Pre-COVID, I did a fair amount of safety pilot flying. I've seen it all. Some of the pilots I flew with were heavily dependent on the autopilot. It usually starts with "I know how to hand fly an approach but I find it safer to use the autopilot." Sitting in the dummy seat, on some planes it is pretty easy to flick the AP breaker discretely and watch how "safer" evolves. 

I also see a fair number of pilots who fly well until something goes wrong or gets changed unexpectedly. 

One thing I have seen that bears mentioning is fixation. It is a pretty common problem. I flew with one pilot who wouldn't divide his attention between flying the plane and looking at his iPad. I timed him multiple times looking at his lap based iPad for an average of 9 seconds each time he looked down. A lot of bad things can happen in 9 seconds (and it did).

Hopefully a clear cause will be revealed. 

[mike_elliott]

9 hours ago, jlunseth said:

This is starting to make a little sense to me in light of information provided by the NTSB. According to the NTSB the aircraft was on an instrument flight plan and was on the ILS 10R into flying cloud. If you look up this thread to about page ten, 1980Mooney puts up an ADSB graphic with two waypoints, ZAMUD and ZIGCO. His post was considering whether there was some confusion between the two RNAV’s into 10R and 10L, but we know now that the aircraft was on the ILS, not either RNAV. What is signficant about his graphic is the it shows the location of ZAMUD on the pilot’s actual route of flight. ZAMUD is the intermediate fix for the ILS 10R. The accident site is after ZAMUD. At ZAMUD the pilot is supposed to descend from 2700 to make 2600 at the final approach fix, which is STUBR. At ZAMUD the aircraft was at 3000-3100 or about 400 feet higher than the 2700 intercept altitude. This would not be unusual if the pilot is being vectored onto the approach, he might have been cleared to maintain 3000 until established, for example. For whatever reason, the aircraft deviated slightly left of course after ZAMUD, so he would have found himself around 3 miles from STUBR, around 400 feet too high and left of course, and he would have needed to cure the altitude problem, in particular, in order to get on the ILS downslope. It was at that point he turned right (south) and descended. He got to 2600 at one point, which is the altitude he needed to be flying at STUBR but was slightly higher at 2700 at the end of the southbound turn. Altitude was still relatively stable at that point.

It was at that point that things went wrong very quickly. at 17:39:39 the aircraft turned north and after a couple of more seconds, began to descend further. It was now northbound, away from the approach course and toward the impact site. From there until impact was in the vicinity of 7-11 seconds

There are several possibilities. There could have been a vacuum or gyro failure. I have had both, and neither one happens in 10 seconds in a vacuum operated gyro. The gyro would not have been working well, but if it tumbled at that exact point there would have been symptoms and altitude variations before then.

Another possibility is a failure in the autopilot control panel. As an example, my AP is a KFC200. It has an Up/Down button. If the AP is in Alt Hold mode, pressing the down button will cause a descent at about 700 fpm until the pilot takes his finger off the button and then the AP holds the new altitude. I use this in just this situation, with a thousand feet or less to lose and a reason to hold altitude thereafter. This spring, this function failed. The Up/Dn button acted as a trim control, pressing the Dn button would continuously run the trim down and then hold the down trim rather than a new altitude, when the button was released. I discovered it during VFR practice approaches and got it fixed. 

The pilot also may have, either intentionally or inadvertently, disengaged the AP. The AP can be inadvertently disengaged by simply using the trim button on the yoke, which the pilot might have done in order to change trim and get down to the desired altitude. 

If the AP remains engaged and in Alt Hold mode but for some reason is losing or gaining altitude, and the pilot tries to correct with pressure on the yoke, the effort will work for awhile, but the AP will run the trim to counteract that pressure and the pilot finds himself working directly against the AP and needing to disengage to fix it.

At about the point where the southbound turn began, the pilot would need to dial the inbound course in with the course needle and switch to APCH mode on the autopilot. If he switched but had inadvertently failed to dial in the correct inbound course, the aircraft would immediately turn to find that incorrect course, which could explain the sudden northbound turn. The pilot would respond by disengaging the AP and hand flying, and making that transition with no notice is difficult.

These are all possibilities and there are certainly others. What is significant is that the plane was stable and on the approach until just prior to the FAF when the pilot would have needed to correct course and lose altitude quickly. Then something happened.

I have had a 256 roll over in the soup on AP with me before, thankfully an Aspen provided conflicting info and verification said the aspen was correct, so I disengaged immediately before a really really bad thing happened. Train often for fine swiss watch failures. These are not new devices that are being relied upon. This particular plane did not have an Aspen digital to analog e100 driving the K(r)AP 150, but instead still relied on attitude info from the fine swiss watch KI256

[jlunseth]

I feel that flying an approach on autopilot actually is safer. However, I have had enough quirky things happen with the GPS + GPSS + analog AP system to know the importance of constantly checking that what you think you have commanded is what the system is actually doing. The biggest challenge is when something happens that requires flying by hand on a moment’s notice, or resetting a procedure on the GPS. So I practice both. A major issue, I think, is that these are not cohesive systems, they are separate components made by separate manufacturers and added in over time. No one, to my knowledge, has written a manual that says how the overall system is going to work. The pilot needs to intimately understand the quirks of each of them.

[chriscalandro]

An autopilot should not be a crutch. In a scenario where an airplane goes from level (ish) flight to a multi thousand  FPM descent in a matter of seconds, especially due to weather, no autopilot is going to correct for that and many, including the GFC will disconnect. 
 

if you think pressing the level button in this scenario would have been useful, I think you need to do a bit more research on capabilities. Autopilot is not MSFS slew mode. 

Edited August 12, 2021 by chriscalandro

[DXB]

28 minutes ago, chriscalandro said:

An autopilot should not be a crutch. In a scenario where an airplane goes from level (ish) flight to a multi thousand  FPM descent in a matter of seconds, especially due to weather, no autopilot is going to correct for that and many, including the GFC will disconnect. 
 

if you think pressing the level button in this scenario would have been useful, I think you need to do a bit more research on capabilities. Autopilot is not MSFS slew mode. 

Though no autopilot would be helpful once there is full loss of control and rapid descent, this is still a pretty nuanced issue.

If he was hand flying, the LVL button on the GFC500 would have been very helpful if he was primed to use it at the first sensation of disorientation. Even if he didn't use it, the ESP would also have kicked in to prevent extreme attitudes and speeds and gone to wings level automatically if lack of further inputs indicated pilot incapacitation.

But more likely, he was not hand  flying but rather using an older autopilot, and an avionics failure precipitated loss of control. If he had a digital AI controlling a GFC500, that type of failure would have been less likely to happen. But if that modern autopilot did indeed suffer loss of attitude info, loss of ship's power, or an internal failure, then he would have been in the exact same pickle.

Since most of us fly in IMC and on approaches much of the time using he autopilot , the greatest benefit of modern hardware seems to be increased overall reliability, not necessarily automated envelope protections and LVL buttons (even my lowly STEC-30 has a form of the latter).  If modern hardware does fail for any reason, the need to identify reliable attitude info and hand fly accurately while still in the normal flight envelope remains just as pressing.  

[David Lloyd]

A question of how this happened in a cloud layer only 600' thick.  From the time headings and altitude control went to heck, altitude varied from 3300 to 2600.  When he popped up to 3300, he may have been on top by a little, but at that point he was going straight.  Off heading but straight.  May not have realized he had a problem until a few seconds later after descending back in the clouds.  Likely stayed in the cloud layer the whole time of the upset until it was too late.

Look at the track provided: Straight, on course at 3100' until a bump up to 3300' and a left turn, then down to 2800, then up to 3000.  Then a right turn of about 90 degrees in 9 seconds (standard rate would take 30 seconds) and down to 2700, then 2600, then 2900, then left almost 180 degrees in 15 seconds (standard rate would take 1 minute).  During that last turn, the nose drops and gets a huge rate of descent.

At this point, I am even more convinced of an attitude indicator failure.  He was following it as it as it wobbled back and forth, up and down.

Just for your own realization, go out with a friend an try a 180 degree turn in 15 seconds.  Can you hold altitude?  Bet not the first couple times if at all.  Put on your vision limiting device and try again.  Likely at the end of the turn a lot of ground will be showing in the windshield.  How 'bout without a attitude indicator.  Of course if you looked only at your turn coordinator, you would never bank that steeply.

[toto]

5 minutes ago, 1980Mooney said:

 

How do we know it was only 600 or 1000 feet thick where he was at the time? He was 10 miles from the nearest weather reporting at Flying Cloud.  How many times have you been on an approach to an airport and either ATC or Tower thinks you are either above or below a layer based upon their weather data/reporting yet you are in IMC? The layer he encountered could have been much thicker.

The 600' estimate I believe came from analysis by @Scott Dennstaedt, PhD

https://www.avwxtraining.com/post/weather-analysis-for-recent-mooney-accident-near-minneapolis

[David Lloyd]

21 minutes ago, 1980Mooney said:

 

How do we know it was only 600 or 1000 feet thick where he was at the time? He was 10 miles from the nearest weather reporting at Flying Cloud.  How many times have you been on an approach to an airport and either ATC or Tower thinks you are either above or below a layer based upon their weather data/reporting yet you are in IMC? The layer he encountered could have been much thicker.

Heck, I was always surprised on an approach when the clouds were as reported.  I figured 200 & 1/2 was a code for the airlines to sharpen up and pay attention.

[Bartman]

I was one of those who initially thought the images were an artifact. I was wrong. My experience from owning and flying a Mooney and from knowing there were only a couple of in flight breakups created bias in my thinking. 
 

I am not an engineer and I am certainly no expert in any way in evaluating the cause of this or any other accident. However I have made mistakes with the autopilot and with approach entry I both the GNS430w and with the IFD440. On more than one occasion the unit did not sequence as I thought it would so I opted for a missed approach and came back around for another try. I can only imagine trying to “fix” the entry error and resulting in unusual attitude. 
 

I cannot imagine the rate of turn as reported in the post above from @David Lloyd  That would no doubt induce excessive Gs and especially if doing it while turning left and right in an uncontrolled manner. I do remember being taught and stressed that if in a descending turn the mantra of reduce power, level the wings, then and only then level off and then add power. In fact, my primary instructor would not sign me off for my PP checkride the first time we went up for the sign off because I was trying to do them simultaneously and he was not happy. We went back the next day and did unusual attitude for a very long lesson. I remember it well.  
 

I can imaging the final seconds of frantically turning left and right with the yoke in the pilots stomach. 

[chriscalandro]

2 hours ago, David Lloyd said:

 

Just for your own realization, go out with a friend an try a 180 degree turn in 15 seconds.  Can you hold altitude?  Bet not the first couple times if at all.  Put on your vision limiting device and try again.  Likely at the end of the turn a lot of ground will be showing in the windshield.  How 'bout without a attitude indicator.  Of course if you looked only at your turn coordinator, you would never bank that steeply.

Steep turns on altitude are a main part of the private pilot ACS standards. Everyone should be able to do a steep turn without a major altitude deviation as it’s a basic skill. 

[kortopates]

1 hour ago, 1980Mooney said:

 

How do we know it was only 600 or 1000 feet thick where he was at the time? He was 10 miles from the nearest weather reporting at Flying Cloud.  How many times have you been on an approach to an airport and either ATC or Tower thinks you are either above or below a layer based upon their weather data/reporting yet you are in IMC? The layer he encountered could have been much thicker.

Read Scotts excellent right up, it was based on real data, including radiosonde data at 23Z that only abiut 5 mi away along the apporach to FCM. Secondly stratcu clouds are capped by an inversion layer which physically caps their altitude at the top of the inversion creating very uniform tops. But that said, it doesn't really matter if it was 600' or 1200' the point is it wasn't a thick layer and the pilot was VMC on top most of the way till starting down on the approach - and then starting having lateral and vertical deviations going IMC.

In my neck of the woods, in the busiest TRACON in the country, the controllers make it their business to track tops reports in the sectors their managing, They're virtually always giving vectors with knowledge of whether your on top on in it. Additionally, when they know your practicing, its common if they're not to busy to offer you a few hundred feet lower (assuming their MVA allows) so the pilot can get more actual. I see it all the time. 

Edited August 12, 2021 by kortopates

[Raymond J1]

I think I remember that in the US, on airports with parallel runways, it is quite customary for last-minute runway changes... And so last-minute approach reprogramming too...

But if he reprogrammed his GNS 530 after the Fix Approached, he must have had some surprises.

By doing "activate leg" after FAF we actually get a route, but with a resolution of 1 NM (TERM). As long as LNAV, LNAV/VNAV or LPV is not displayed on the GPS, the GPS is not in approach mode, we will not have the accuracy of the approach, and we will not have a slope (glide path).

To switch to approach mode, it is necessary to walk between IF and FAF in a limited angle. At some point the GPS will switch from TERM mode to approach mode. If it does not go into approach mode, it is necessary to abandon the approach. The PBN documentation insists quite a bit on these limitations, on not accepting radar guidance to the FAF or a DCT FAF, and on being established on the 2 NM axis before the FAF.

The ACTIVATE APPROACH function simply does DCT to the IAF of the procedure, but does not switch to approach mode until one arrives in the intended window.

[Marauder]

1 hour ago, Raymond J said:

I think I remember that in the US, on airports with parallel runways, it is quite customary for last-minute runway changes... And so last-minute approach reprogramming too...

But if he reprogrammed his GNS 530 after the Fix Approached, he must have had some surprises.

By doing "activate leg" after FAF we actually get a route, but with a resolution of 1 NM (TERM). As long as LNAV, LNAV/VNAV or LPV is not displayed on the GPS, the GPS is not in approach mode, we will not have the accuracy of the approach, and we will not have a slope (glide path).

To switch to approach mode, it is necessary to walk between IF and FAF in a limited angle. At some point the GPS will switch from TERM mode to approach mode. If it does not go into approach mode, it is necessary to abandon the approach. The PBN documentation insists quite a bit on these limitations, on not accepting radar guidance to the FAF or a DCT FAF, and on being established on the 2 NM axis before the FAF.

The ACTIVATE APPROACH function simply does DCT to the IAF of the procedure, but does not switch to approach mode until one arrives in the intended window.

You make a great point about needing to reprogram a GPS when you are expecting something else. I used to be one that would program "Vectors to final" when I heard the controller say "you can expect vectors to final". Only to be surprised to be assigned an IAF or IF part way through. If you aren't keystroke fluent, it can substantially increase your workload.