GPS CDI guidance in holds and procedure turns

[Vance Harral]

Not really a Mooney-specific question, but no luck on another board, and we've got lots of IFR experts here.  Note that while the certified navigator in our airplane happens to be a Garmin GTN-650, this is a general question, possibly with different answers depending on different equipment.

For various reasons - often "historical" - certain instrument flight procedures are not based on a specific course over the ground. The obvious examples are timed holds and "remain within XX NM" procedure turns. I learned to fly IFR without an IFR-certified GPS, so I'm comfortable flying these segments old-school: by reference to headings and timed legs. I confess I generally ignore the CDI on the outbound leg and through the turns of such holds/PTs, as I'd do navigating purely by ground-based systems. I just verify when intercepting the inbound course that the GPS sequences to the appropriate flight plan segment, and that the CDI guidance looks reasonable. But I'm vaguely aware the navigator paints a magenta line on the moving map for the turns and outbound leg of holds/PT, and that the CDI is indicating "something" while you're flying them.

My original understanding was that a CDI in GPS mode would simply depict lateral displacement from the inbound course throughout timed holds/PTs, just as VLOC guidance with a ground-based transmitter would depict angular deviation, and I'm pretty sure I've observed that behavior in at least some cases. But complicating the matter is that holds and HILPTs on GPS approaches are defined by a specific course over the ground, and the navigator could theoretically provide guidance all the way around the hold, much as it does when flying a DME arc in GPS mode. Things get even more interesting in light of stories I hear/read about autopilots with GPS roll steering automatically flying holds and procedure turns. We don't have roll steering, but in an aircraft that does, I think this can only be possible if the GPS is providing guidance throughout the entire hold/procedure turn.

So... I'm sitting down with the GTN trainer this morning to fiddle with this, but my initial attempt at understanding was to read the GTN manual, and I'm not actually having any luck there. There are mentions of solid vs. segmented flight plan segments on the moving map in the Flight Plan section of the manual, but it doesn't actually seem to discuss CDI or other guidance behavior. Hoping someone more knowledgeable than me can comment here, ideally with reference to an "official" (manufacturer or FAA) document that describes how this works. Any takers?

[midlifeflyer]

I'll give it a shot, Vance.  The concept to keep in mind is that what the navigator does and what the autopilot does is not always the same. They usually are but to understand certain functions you have to keep them conceptually separate.

A dotted magenta line in the GTN indicates the navigator is not providing "positive" course guidance.  That just means the course displayed is not required. You are seeing it in a HILPT, like the GTN graphic below (which should be familiar), but they appear elsewhere too. For example, load a SID with an initial Radar Vector segment. Some navigators will paint a doted magenta line to the transition fix.  That doesn't mean yo fly it - you fly the assigned vectors until that instruction to proceed direct to the transition, hit the D → and get a solid magenta. In both cases they mean same thing. They are primarily for reference purposes and are not instructions to fly that course.

That's clear in the radar SID example.  It gets mentally jumbled in the HILPT because it sure looks like you are supposed to fly the line. But you are not. That 1 minute of the holding pattern (or 4 nm for a distance hold) is a maximum. ou don't have to fly that dotted magenta line. You can turn in early if you want. Turning the hold around, the GTN might announce "parallel" but you are free to choose teardrop if that's what you want to do. 

Overly simplistic, but it comes down to solid magenta = required; doted magenta = optional. So, the solid gets displayed on the CDI; the doted does not.

Stop here until that makes sense. If it doesn't I'll try a different, um, approach.

Now that we have established that the dotted magenta for the hold entry (as opposed to in inbound course) is optional, the CDI in the hold acting like a VOR - pointed to the inbound course and showing the offset, makes sense. The only required course in the inbound course, the type of entry and size of the outbound are optional. The CDI only shows the real course. You are required to fly the solid magenta to be on the required course. If you hand fly and follow the dotted magenta it's because  you choose to fly the doted magenta.

Move to the autopilot. A compatible autopilot - essentially one with GPSS  engaged - will follow the doted magenta. But that's because you are substituting the autopilot for yourself. Just like hand flying, by choosing GPSS , you chose to tell it to fly the dotted magenta. It may not be reflected in the CDI, but the data is there and you have chosen to tell the autopilot to use it. 

Dunno if that helps. This is s much easier to explain and demonstrate live.   

And yes, it's all document in what you read. It's what it means which is missing.

 

[gsxrpilot]

Hi Vance,

I don't have a GTN650 but I do have an IFD540 coupled to a KFC150 with GPSS provided by Aspen. I'd be happy to take you up in my Mooney and you can see how it works in a live environment. I'm sure it would be a great opportunity for me to learn from an experienced IFR pilot like yourself as well. My Mooney is just down the road at BJC.

[Vance Harral]

OK, based on these replies, some time with the iPad trainer app, and a conversation with a buddy, I think I've got this sorted out.  I'll lay it out @carusoam style:

1. In GPS mode, the CDI always indicates the difference between the present position and the desired track (DTK).
2. DTK isn't set directly, it's derived from the flight plan.  Holds and PTs are just another segment in the flight plan.
3. For hold segments (including HILPT), the GPS sets DTK to the inbound course and keeps it there for the duration of the hold.
4. For traditional 45/180/45 procedure turn segments, the GPS sets DTK to the outbound course on crossing the fix outbound, then autosequences DTK to the inbound course halfway around the PT.
5. The data stream that drives the depicted flight path on the moving map, the "Turn to XXX in YY seconds" messages, and autopilot roll steering data, is independent of the CDI.
6. Some depicted flight path segments are "variable", or what @midlifeflyercalls "optional" above.  These are depicted as dashed lines when they are not the active segment.  When they become the active segment, they turn magenta, but also change from plain dashes to arrows, to help emphasize turn direction.

As I dig into this, one thing that interests me is that the depicted flight path for "variable" flight plan segments change as speed changes.  Not hard to understand why, as the turns in these segments are based on standard rate turns, and the radius of a standard rate turn changes with speed.  But in a baseline installation, the only speed data the GPS has is ground speed.  If the algorithm were really going to be perfect, it would take heading data from a magnetometer and airspeed information from an air data computer, use that to determine winds aloft, and produce flight paths that take into account winds aloft.  Do "high end" avionics installations couple this data to the GPS via a input port, such that the GPS can use that information in computing variable flight paths?

Further comments welcome, thanks for the replies so far.

[Vance Harral]

7 hours ago, gsxrpilot said:

I don't have a GTN650 but I do have an IFD540 coupled to a KFC150 with GPSS provided by Aspen. I'd be happy to take you up in my Mooney and you can see how it works in a live environment. I'm sure it would be a great opportunity for me to learn from an experienced IFR pilot like yourself as well. My Mooney is just down the road at BJC.

I'd love to go flying with you in the immediate future, I'll PM you for details.

[midlifeflyer]

I like the way you put that summary, @Vance Harral. I'm going to use "variable" for the dotted line. More descriptive than the "optional" I came up with in the moment.

BTW, on #2, DTK can be set directly. That's what OBS mode does.

[Vance Harral]

6 hours ago, midlifeflyer said:

BTW, on #2, DTK can be set directly. That's what OBS mode does.

Absolutely correct, yes.  When I say "DTK is not set directly", that assumes you are in flight plan mode as opposed to OBS mode.

These boxes are wonderful, and I think not actually that hard to understand; but they sure can be difficult to accurately describe with the written word.

[midlifeflyer]

2 hours ago, Vance Harral said:

Absolutely correct, yes.  When I say "DTK is not set directly", that assumes you are in flight plan mode as opposed to OBS mode.

These boxes are wonderful, and I think not actually that hard to understand; but they sure can be difficult to accurately describe with the written word.

The comment about DTK was a nit-pick

Definitely true about descriptions of functions.  Which is why I often answer questions with a short video of the simulator performing it.

What I have also found with all of them is that there are practical capabilities which are roughly based on information contained in the manual but are not really obvious. One example is using the flight plan catalog to set up an alternate airport (or a series of airports in a practice session), including the desired approach, in advance. I was explaining that function to my instructor on my last IPC. He didn't really get it until he saw me pull up the next approach on a 3-approach plan, confirm, brief and load it  while the GPS was continuing to give me course guidance on the missed for the prior one.