FPL was to North Weald EGSX, the ILS was to Southend EGMC, the latter was not in the FPL at all, so I am sure the AP did not know where I am going (e.g. DME or GPS distance to Southend) but I am sure it has an idea of wind corrections from GPS and these may feed the ILS tracking algo?
But wondering if auto-ident of VHF aids in NAV1/2 boxes does help something? not that I wrote the code, I am just guessing here…
I am sure if you constrain to rate1 turns there is some cases of tailwinds where an analog auto-pilot will not be able to do say a 90deg intercept of the localizer using 1 scale deviation signal only, so you may have to help him with GPS or rely on ATC to give you a max 30 deg vector
NCYankee wrote:
The winds generally will change between the FAF and the MAP, typically by 90 degrees in direction and the velocity reduces.
In my experience (and according to MET theory) 90° change is not at all typical – far from it. Typically the ground wind differs by 30° from the gradient wind (over land). If there is an inversion below the FAF then you could get arbitrary change in wind direction but that’s not typical.
I am sure it has an idea of wind corrections from GPS and these may feed the ILS tracking algo?
That’s a fair point; the AP has the heading data (X/Y/400Hz or ARINC429 heading) and the GPS mag track, so that can be used. And it knows the final approach track (from the course pointer) so it can work out the effect of the wind.
I’m in the middle of fitting a GFC500 so will be able to test once installed and I’ve got the hang of it if no one else with one can answer beforehand.
That will be great info. IIRC (from some video) GFC500 will disconnect on ILS approach if losing GPS signal but then it can be re-engaged and it will continue to fly based on signal from NAV only. Maybe I’m wrong or I don’t recollect correctly.
I have flown with the KFC225 in a fair number of Bonanzas and don’t see any particular improvement over its brothers the KFC200 and KFC150. They all drift off and S turn on an ILS in challenging wind conditions.
I asked one “industry source” who used to design this stuff for his input and this is what he came back with. Nothing concrete but still very interesting:
The KFC230 (and anything before) never uses any GPS inputs for the ILS
control loops. (The only thing that can theoretically be supplied from a
GPS unit is a “synthetic”/fake radar altitude). An RA input, if available,
is used for gain scheduling of both the LOC and GS loops). If an RA
input is not available, the control loops use just a primitive counter to
smoothly decrease a regulator gain (because the deviations are angular…).
Of course that the counter-based gain is limited (never zero) and is
derived to be for an “average” intercept distance/altitude. Absolutely
normally (if you don’t have a RA) you can experience a slightly different
performance every time… Having a RA input is of course more robust.
So to fly an ILS approach the only thing the AP needs to know is FAC (FAC
minus MagHdg, so that it estimates the time needed to turn into a FAC with
a given max roll = cert value. e.g. 22deg), and then LOC/GS deviations.
Then the AP can compensate for a crosswind (more dynamically than in a VOR
case), maximally ~30deg crab. This compensation is “secondary” – when the
airplane is unable to stay on the beam and the loop has a permanent
one-sided error. A “primary” compensation would only be in the case of GPS
roll steering – the navigator “knows” (has a permanent estimation) a
crosswind component (and the complete geometry as well) so that the
reaction can be faster, and both the intercept and tracking more precise -
like on rails.
Altogether no real need to know any distances, TAS, x-wind,… a trade
between simplicity and dynamics.
Re GPS enhancements (speculations for the Garmin case):
It’s not impossible that Garmin uses some GPS data to make the ILS loops
work better – G/S, TAS, distance, altitude, … and any of these values
don’t have to be super precise (as should be in an LPV case). They only
help the loops to be better tuned for various states of an approach. But I
strongly believe that the GPS data is only to enhance the performance, but
is not essential. I don’t believe that losing a GPS will disengage the
LOC/GS modes. It wouldn’t make sense. The performance loss would vary from
“some” to negligible.
In your particular case it seems like your max roll angle is set too low -
quite clearly the loop (and the intercept point estimator) calculates with
higher roll angle (22deg…?) and (if that wasn’t a strong head/cross wind
case) you might experience overshoots quite often. The larger the intercept
angle is the larger the overshoot will be. The LOC loop is tighter (than
say the VOR loop) and any overshoot or crosswind is compensated relatively
quickly.
Yes exactly my roll angle is too low; I need to set it for the proper 23 deg or so.
Just from own experience (of busting an ILS-approach): Max WCA for KFC-150 is 30 deg. So even if wind Dows not change, there are wind related limitations of “traditional” APs.
On the original post: The really interesting question is how that fact could surprise anyone (who is flying a G5/GFC-500). P. 53 of the POH clearly states “Glideslope mode requires a valid GPS position”. (Graphically very prominently marked in the POH).
So whoever is surprised that ILS can’t be flown w/o GPS either doesn’t fly a GFC-500 plane (and then it doesn’t matter) or significantly failed on ones homework…
Clearly mostly the latter…
An autopilot which needs GPS to fly an ILS is a cynical play on the fact that most pilots never read the POH or the avionics manuals.
I wonder if the Avidyne DFC90 has a similar limitation? Perhaps @avijake or @steveavidyne might know.
The other thing I find interesting is who is not posting on this topic 
Glideslope mode requires a valid GPS position
Is a ‘valid GPS position ‘ properly defined in the manual? IFR GPS should suppress loss of integrity after the FAF and revert to deduced reckoning. How does the AP know that the input is now based on deduced reckoning? But this may only occur if the approach has been loaded and activated?
Does the manual explain what loss of integrity messaging is sent, eg CAS message, or does the AP disconnect? Presumably the manual describes the required actions if you have an LOI once established on a coupled ILS?
Malibuflyer wrote:
So whoever is surprised that ILS can’t be flown w/o GPS either doesn’t fly a GFC-500 plane (and then it doesn’t matter) or significantly failed on ones homework…
What if I hand fly then turn on autopilot in NAV mode choosing ILS as source without loading the approach from database? Do I still need valid GPS position for GFC500 to fly the approach?
