Autopilot not flying courseline on approach

I have been having a problem with my IFD540, Aspen, STEC-30 combination during an approach.  Before now, I was mostly hand flying approaches and apparently never completed an approach on autopilot down to the minimums.  Hand flying the approach was fine.  The problem came when I let the autopilot do it.  The autopilot was flying the course set by the IFD just fine until it got to about a one mile final and a couple hundred feet above minimums.  At that point the autopilot would direct a turn to the right and I would have to disengage the autopilot to get back to the runway.  If I hand flew the approach, all was good and I could fly right to minimums on course.

I had s string of emails and a phone call with T.J. Sutton at Avidyne this week and he nailed the issue right on the head. I can't tell how many people I have talked to about this and he is the only one that has had the correct answers.  Thanks T.J.

For anyone else out there that has an STEC-30 - Aspen - IFD combination, here is the low down on it.

While en-route, with a flight plan in IFD540, the autopilot stays on HD mode and the Aspen has GPSS engaged.

When you cross the IAF inbound on the approach, the IFD is in VLOC, LPV, LNAV +V, or whatever approach you're on and your LDI and VDI become active, switch the autopilot to HI TRK mode and then turn off GPSS on the Aspen.

In that configuration, the autopilot flies on the line, all the way down to MDA or DA.

This may be obvious to everyone else, but it was not to me.  I hope this may help anyone else with similar handicaps to mine.

Thanks again to T.J. Sutton.

HenryM wrote: Hi KC! Why does the problem occur if you don’t change to mode? I have an IFD-540 and an STec 30 with GPSS. The few GOS approaches I’ve tried, the AP keeps the plane laterally aligned aall the way down to the MAP and then I guides me through the missed approach. Is this an Aspen issue? What is actually happening behind the scenes to elicit the turn of you don’t change modes?

This was not just with a GPS but also with an ILS approach.  If I leave the STEC-30 on HD mode and the Aspen on GPSS, the autopilot takes a right turn at a one mile final and a couple hundred feet above minimums.

My guess is that it is specific to the Aspen.  Something to do with the ACU and multiple connections between them.  I don't pretend to understand it all.  T.J could probably answer this a lot better than I can.

Here is a portion of the email from T.J. to me about this.  It might help, but if not, maybe T.J. can chime in here:

GPSS should be fine to the IAF, and probably ok to the FAF, but inside the FAF the autopilot should be driven from the main left/right input, rather than the HDG/GPSS input.  A change in the Nav Mode results in a change in the CDI scaling as shown below from the IFD pilot guide.  However, when being driven by the GPSS roll steering label 131, the scaling doesn't change (at least not in the IFD).  It will stay at the "Enroute" value.  The closer you get to the runway, the tighter the scaling should be, but GPSS will not give you that result. 

 

 

Below is from the Aspen Pilot Guide.  There may be some confusion, because they reference a Garmin GNS unit in several places and the GNS transmits ARINC429 Roll Steering Label 131 that actually does change scaling.  The IFD only changes scaling in VLOC/ILS mode, or in GPS mode if the GPS is selected as the Nav Source (not GPSS).

 

 

On the STEC-30 you'll want to have HI TRK mode engaged for either a GPS or ILS approach.  This will give the autopilot maximum authority and try to maintain +/- 1 needle width on the CDI.  and since you'll be using GPS or VLOC as the selected Nav source on the Aspen side, the deviations will be driven from the Main Left/Right outputs of the ACU.

 

 

 

A GPS/NAV combo box like an IFD440 or Garmin 430W can navigate using either GPS derived location or VOR/LOC derived location.  As pilot, you get to choose which, just as in the days before combo boxes you could choose whether to have your panel's course indicator accept signals from your KX155 or your King KLN94, depending upon whether you wanted to navigate via VORs or via satellites.  Combo boxes are all the rage now, of course, since you only have to buy one box and they take up less panel space.  Bad news for the King boxes that back in the day ruled GA.

Edited by Catani - 14 May 2018 at 9:10am

Catani wrote: .......But with the new whiz-bang gizmos that changed GA, like an Apollo GPS or later Garmin 430, it became possible to preload all your flight plan waypoints before takeoff, just like new airliners!  ........If it's not done, the autopilot may not fly the glideslope, and will also fly final with less precision.

Thanks for the insight.  Two things there struck a chord with me.  Number 2 first.  My STEC-30 has altitude hold but no VNAV.  I might have picked up on the problem earlier if it did, since it wouldn't have captured the glide slop, but maybe not.  I can be thick headed at times.

The first point reminded me of a question from yesterday.  In all the hub-bub of getting ready for the testing, yet again, I inadvertantly not only entered a course in the IFD but I also entered the approach too.  I did not activate it but after taking off, it must have activated itself because that option was no longer available on the FMS page.  In addition, I took off southbound, intending to fly straight out, hit the IAF, do a procedure turn and shoot the LNAV+V approach back to the airport.  The course that the IFD had only went as far as the the IAF outbound.  No PT or approach.  I think I broke it.  I had to delete the approach and re-enter it to get a good setup.  I've never entered the approach while on the ground before and wondered if anyone has seen that behavior?

mfb wrote: Entering an approach while you're on the ground is not a problem. In fact, it's a good idea because it will reduce your workload in the air. But the IFD suffers from the same curse as every electronic gadget - it will do exactly what you tell it to do (whether it makes sense or not). What approach transition did you put in the flight plan? If you entered "vectors to final" the IFD would not do a procedure turn because you didn't tell it to do one. It's expecting to intercept the final approach course yourself. It will fly forever until it intercepts the course. And if you intercept going the wrong way the IFD will be very confused. But, if you tell the IFD to go to an initial approach fix that has a procedure turn, then it will do the procedure turn. The trick is to load the approach and transition you want, then look at the flight plan and the map representation of the flight plan. Make sure that it shows what you expect it to show. It will clearly indicate if there's a procedure turn or a holding pattern included in the approach. And look at the colors of the course lines on the map. The magenta line is what the IFD thinks it's flying now. The barber pole line is what it wants to do next. Stay aware of these and you won't have surprises. And there's really no "activating" an approach on the IFD, except for vectors to final. You just load the approach transition you want and fly to a waypoint on the transition. The "activate" thing is a leftover G* concept. Hope this helps. Mike

Well now I'm confused. I entered the KBRY RNAV 3 into the IFD with the initial fix being MAKRZ while I was on the ground at BRY.  I took off of RWY 21 and headed straight out towards MAKRZ (It's Bardstown KY, every fix is either Bourbon or UPS related).  The intent was to go direct MAKRZ, do a teardrop entry and fly the LNAV+V approach back in.

I usually preview my approach in the FMS before switching to the MAP view but was in a hurry and didn't.  After I was airborne and the map started scrolling towards the IAF.  The course line suddenly ended somewhere short of the IAF and didn't show any of the rest of the approach.  I thought that was very strange but dumped the approach and reloaded it and it came up just fine then.  Unfortunately I have enough expeirence with getting very good at doing that quickly.

If I'm not misstaken, any time I enter an approach, I get the ACTIVATE APPROACH bubble on the left side of the screen in the FMS section.  Most of the approaches that I do are direct to the IAF and for GPS approaches so I don't use Vectors To FInal much.

Here is the GPS Rwy 3 approach plate.

Kentucky Captain wrote: ...I entered the KBRY RNAV 3 into the IFD with the initial fix being MAKRZ while I was on the ground at BRY.  I took off of RWY 21 and headed straight out towards MAKRZ (It's Bardstown KY, every fix is either Bourbon or UPS related).  The intent was to go direct MAKRZ, do a teardrop entry and fly the LNAV+V approach back in....

Catani wrote: What was the first waypoint or procedure you entered on the FMS page after the IFD preloaded your departure point KBRY? If you entered anything else, please list them all in order.  What was the last waypoint or procedure you entered into the FMS?  Did you "activate" the flight plan before taking off?  Trying to re-create what your FMS would have looked like before takeoff.

I selected -  PROC  -  RWY3 (LNAV+V)  -  MAKRZ. 

I went back to read my STEC 30 autopilot and ST 901 GPSS converter manuals, and neither mentions needing to go to Hi Trk mode to shoot an approach. In fact, both show examples of various approaches (LOC and VOR) and use the HDG mode (or GPSS + HDG in the case of the ST 901 manual). They explain the greater authority of Hi Trk vs Low Trk, but when they show examples of using approaches, they don't use those modes.

HenryM wrote: ...Thus as long as the needle is centered, it is on the correct course, regardless of what signal the AP is using to stay there. As long as I monitor the CDI, I'm following the right path and the sensitivity of the needle changes as needed at various points.  I understand I shouldn't rely blindly on the GPS position, but if the CDI is correct, I can keep going on my merry way. No?

If it's keeping the needles centered on an ILS (or even an LPV, but an ILS would make the following true for sure), I doubt the autopilot is remaining in GPSS/HDG mode in order to follow the GPS's heading guidance for GPSS capable autopilots.  Somehow, the autopilot is automatically switching to approach mode (my STec 55X will NOT automatically switch to approach mode from GPSS - it stays in GPSS and does a sloppy job of staying on the final course and can't follow any glidepath).  That must be so since it is following the glideslope, and because of its increased sensitivity to course deviations.  That doesn't happen in GPSS mode. I'd call tech support at the autopilot factory if the manual does not explain that behavior.  In any case, it sounds like you are legal.  Would be good to to understand why and confirm it though.

HenryM wrote: I went back to read my STEC 30 autopilot and ST 901 GPSS converter manuals, and neither mentions needing to go to Hi Trk mode to shoot an approach. In fact, both show examples of various approaches (LOC and VOR) and use the HDG mode (or GPSS + HDG in the case of the ST 901 manual). They explain the greater authority of Hi Trk vs Low Trk, but when they show examples of using approaches, they don't use those modes. From what I can tell, the autopilot can use inputs from the HDG indicator (which the GPSS emulates) or the CDI/LDI/VDI. The latter are used when you go to one of the Trk modes. However, the approach examples seem to always rely on the HDG mode of the autopilot. In my experience, the autopilot keeps the CDI centered very well. The IFD-540 switches to the right mode (say VLOC for an ILS) and shows deviations from the glide path and course. I assume it is indicating information from the NAV radio, not the GPS, when in that mode. Thus as long as the needle is centered, it is on the correct course, regardless of what signal the AP is using to stay there. As long as I monitor the CDI, I'm following the right path and the sensitivity of the needle changes as needed at various points.  I understand I shouldn't rely blindly on the GPS position, but if the CDI is correct, I can keep going on my merry way. No?  Is there something I'm missing? Would the GPSS switch that changes the input to the AP from the HDG system to the GPS navigator also affect the IFD-540 and cause it to drive the CDI differently?

Thus, my confusion.  Everything I saw or heard from the folks at STEC said to leave it in HD mode.  I was told that Lo TRK and HI TRK were a throw back to an era of VOR approaches and ILS and GPS didn't need anything other than HD mode.  The reality that I experienced was somewhat different witih it toeing the line all the way down to around one mile and then rolling off to the right.

Turning off GPSS steering on the Aspen while the STEC is still in HD mode results in the autopilot following whatever was set in the heading bug of the Aspen.

Setting the STEC to HI TRK and then turning off GPSS steering gives you a straight ride all the way down the final.  All of this after passing the IAF, of course.

All I know is that I'm a happy camper after nearly a year of battling this autopilot issue.  I thought all along that this final issue that I was having, was something left over from when the moron tech fried my panel by plugging a 28-volt GPU into my 14-volt C172, while it was powered up.  It started a cascade of failures in the panel.  For some stupid reason, I allowed the same shop to try to repair their screw-up.  Everything didn't quit at once.  It was a cascade of things over time.  The shop decided that it wasn't their fault and refused to do any more unless I was paying shop rates, so I turned it over to my insurance and went to a reputable shop.

They ended up replacing the Aspen, it's ACU, the IFD540, AXP340, and Garmin GNC255A .  The PSE PMA450 and STEC-30 were sent back to the factory and recertified.  The JPI EDM900 MAP sensor blew up and had to be replaced also.  Not only that but they completely screwed up the installation of the synthetic vision, AOA, and ADS-B unlocks on the Aspen.  It's been almost a year now and I just now feel that the airplane is 100% again.  Thanks go out to the folks at Avidyne and Aspen who fielded so may phone calls and emails from me in the journey to get this bird back to where it was before.

It was easy to assume that this last issue I was having was part of the initial problem but in reality I just hadn't had enough experience with the autopilot thrown into the mix.  I was flying all approaches by hand.

Just to stop the noise, I've had several folks, including Aspen initially tell me that there should be no damage since all of the equipment in question is designed to work at either 14 or 28 volts.  I pointed out that while that may be true, I don't think that applying a 28-volt current to a 14-volt system while it was already powered up is what the manufacturers had in mind.  They all tended to agree.

We had just pushed the airplane into the hanger so that they could install the Skytrax100 and do the IFD 10.1 update.  I remembered that there was a good chance that I would lose all of my configuration settings on the IFD so I told the tech to wait a minute while I powered everything up and get some screen shots.  I was sitting in the airplane with the master switch on and the avionics bus powered up, just snapping pics of the panel, when the engine started turning over on its own.  Scared the crap out of me and I immediately turned the master off.   That's when I saw what the tech had done.  Nearly a year later and all is good....except there is this intermittent issue of the panel mounted power plugs that I had installed with the new panel not working.  Every time I get it to the shop, they work.  Hard to fix intermittent. Someday....

HenryM wrote: 1.  I still don’t understand what is happening with your Aspen commanding a right turn. Where would the next point in the missed approach be? To the right?  2.  It makes sense that turning off GPSS would make the AP follow the heading bug. The GPSS switch changes from following the GPS signal to following the DG bug. Changing the AP to Hi Trk makes it follow the CDI signal, so that behavior makes sense too. The IFD is driving the CDI in VLOC mode, following the ILS signal.  When the Aspen is controlling, it should be following the GPSS commands and shouldn’t just get off track like that. It might not be as accurate as the ILS, but it still shouldn’t just turn right like it does. With the GPSS turned on and the AP in HDG mode, it just thinks it has a DG bug telling it how far off the desired track it is. The Aspen is providing that error signal.  3.  Does the problem happen at one of the approach waypoints where maybe it is changing to a new segment after that point? My system is similar to yours, except that the GPSS is done by an ST-901 instead of the Aspen. The IFD should be sending the same commands to either unit. The Aspen is smarter than the ST-901 and perhaps it is getting ahead of the game too soon.  4.  Hopefully someone knowledgeable with the GPSS commands can come up with an explanation. I would like to understand the root cause, even if there is a workaround, like changing to follow the CDI instead of the Aspen.  5.  An experiment you could try is just changing the STEC to Hi Trk mode without messing with GPSS. It should start following the CDI regardless of what the DG says. Then again, in your case the CDI is in the Aspen too, so who knows what its logic is for the CDI output when it thinks it is in GPSS mode. The STEC-30 has no idea what is driving it. It just takes either a bearing ear or from the DG (in HDG mode) or a CDI signal from the NAV (in Lo or Hi Trk). The difference between Lo and Hi Trk is just how quickly it reacts to an error. Lo Trk tries to dampen the scalloping of a VOR signal while Hi Trk tries to follow the signal more quickly. It doesn’t know if the signal is a VOR radial or an ILS. Hi and Lo just determine how fast it reacts to deviations.

1. I was never able to stay with it that long, only to the threshold.  The view there was not what you want to see at the DA or MAP.  Runway way off to the left.

2. True.  Turn off GPSS mode and the autopilot follows the heading bug on the Aspen.  Real handy in getting vectors to final, around traffic, etc. without losing your flight plan in the IFD.   Once the approach is loaded, you are inside the IAF, and the autopilot is in HI TRK mode, turning GPSS off follows the approach all the way down to minimums like it never has before.

3. No, it only happens around the one mile final mark and a couple hundred feet above DA.  My understanding from Avidyne is that there is more than one path from the IFD to the ACU and depending on what mode you are in, what path it takes.

4.  If T.J. Sutton from Avidyne would chime in here, he could explain it a lot better than I can. He is the one that figured out what was happening and its resolution.

5.  I asked that very question of T.J. and without trying to find the email to quote, he said something like it would probably produce similar results but the presentation that I would be seeing would be GPS derived instead of navaid based,  I think.  I do plan on testing that configuration out on my next flight.

It takes a little more planning when trying to get someone to fly with me to test this out.  Being retired is great but it seems that all the people around me are always working.  I'm not fond of having that much head down time in the cockpit while flying around VFR by myself.  I much prefer having someone with me to keep a lookout for other aircraft and make sure I don't do something stupid while wrapping my head around what's happening on the panel.

Kentucky Captain wrote: ...Everything I saw or heard from the folks at STEC said to leave it in HD mode.  I was told that Lo TRK and HI TRK were a throw back to an era of VOR approaches and ILS and GPS didn't need anything other than HD mode...

Ask an instrument DER or a local experienced CFII what they think about that claim.  Or just read AIM 1-2-3 to them.  If it's illegal for you to handfly inside the FAF following GPS course guidance, I don't see how it could be legal allowing the autopilot to do it.

Catani wrote: Kentucky Captain wrote: ...Everything I saw or heard from the folks at STEC said to leave it in HD mode.  I was told that Lo TRK and HI TRK were a throw back to an era of VOR approaches and ILS and GPS didn't need anything other than HD mode... Ask an instrument DER or a local experienced CFII what they think about that claim.  Or just read AIM 1-2-3 to them.  If it's illegal for you to handfly inside the FAF following GPS course guidance, I don't see how it could be legal allowing the autopilot to do it.

After Avidyne used small words to explain how it works, I totally agree.  I was just repeating what I was told when I first took delivery of the new panel.  No wonder I have been so screwed up.

A couple of things I would like to hear clarified.
1. As far as I know Aspen does not "command" anything. The GPSS signal processor is in the Aspen PFD but it needs an external signal for navigation data.
2. I use GPS inside the FAF all the time. I think what is meant above is that using the GPS as a heading source and flying the heading instead of the approach course is not approved.

HenryM wrote: I follow the CDI down, handflying the glide path. The CDI is being driven by the IFD-540 in VLOC mode, thus the ILS signal. The AP is connected in HDG mode but I make sure the CDI is centered.... Is there an FAR that is being broken? The AIM is not regulatory.

You have the option of coupling the autopilot to the localizer, but choose not to do so for reasons I don't get.  Leaving that aside, as I said previously, I don't see how if it's illegal to hand fly a GPS course, I don't see why it would be OK letting an autopilot do it. 

chflyer wrote: The obvious question is "What would you do if the ILS/LOC-driven CDI starts to show (lateral) deviations while the AP is engaged, probably because the AP is tracking GPS corrections rather than the ILS/LOC heading error signals?"

And that's why I started this whole mess.  The autopilot was driving the airplane to the right of course at one mile even though the Aspen's HSI and LDI were showing it moving from the centerline to right of course.  That was with the STEC in HD mode and the Aspen in GPSS mode.  Flying the approach by switching the STEC to HI TRK and the Aspen off of GPSS mode inside the IAF gives me a really nice approach all the way to the minimums.

You mentioned an STEC GPSS converter and I'm pretty sure that I don't have one because when I was having the panel installed, they mentioned that I didn't need the extra equipment because the Aspen had that already.

One more thing that caught my eye in this thread; all the talk has been about going to HI TRK and off of GPSS when doing an ILS approach.  My understanding from Avidyne was that I was to do the same procedure for a GPS approach too.  My testing bears that out because with GPSS in  and HD mode on, a GPS approach ended with the same dismal results as an ILS.  GPSS off and HI TRK mode on, the autopilot flown GPS is right on the money.

To summarize again, here is my understanding of the system:

Edited by HenryM - 17 May 2018 at 9:00am

Correct, Orest. It was a mistype. I meant to say Aspen. I corrected my post.

The STEC-30 doesn't have an APPR mode. It only knows about a heading source and a CDI. 

oskrypuch wrote: The ILS deviation signals by the IFD are a discrete output. There is an update to the spec for GPS units, where a different output is used to provide BOTH the ILS deviation, and an APV (precision GPS) approach, whichever is active. That is the output that is now normally wired to your autopilot, so you can fly both types of approaches. The autopilot has no idea whether it is flying an ILS or LPV approach, and it reacts identically. The increased sensitivity on an LPV approach is generated digitally by the IFD, to mimic the sensitivity change you get on an ILS, as you near the threshold. This output does NOT provide the magenta line GPS deviation signals that you use for enroute, and that latter signal is disabled once you pass the FAF. That is why you must switch to an APPR mode on your autopilot, once past the FAF on a precision approach. * Orest

chflyer wrote: That would explain and is consistent with Avidyne telling Kentucky Captain that he should switch his STEC-30 from HD (GPSS) mode to TRK HI even for APV approaches. Very useful information !! BTW, is there any reason NOT to switch to TRK HI as soon as established, rather than waiting for the FAF? I don't see one and would think the earlier the better to keep the workload after the FAF as low as possible.

Actually, Avidyne told me to switch to HI TRK after the IAF.  In test flying the approach, as soon as I hit the IAF, the IFD went from GPS->LPV to LPV and the LDI and VDI appeared on the Aspen.  At that point, I switched the STEC to HI TRK and turned GPSS off on the Aspen.  The depicted course didn't change and the autopilot stayed on course all the way to the DA.

I haven't had the opportunity to try it out on an ILS.  I like doing this testing in calm air.  With the air so rough lately that you have to put one hand on the overhead and a 20-knot quartering wind, it gets hard to tell what is the airplane and what is the weather.  By the time in the evening when the gale force winds calm down, UPS is cranking up and SDF is the only place close to do an ILS.  They don't appreciate my 90-knot airplane in a 170-knot final.  At least I didn't when I was working it.  I'll wait for the weekend.

I'm trying to wrap my head around this discussion and apply it to my (ancient) Autocontrol IIIb. Like the STEC30, it doesn't have an approach mode-- it has HDG, LOC, and HI modes though. Normally the AP mode selector stays on HDG and the AP thus flies whatever CDI deviation indications it gets from its nav source. In my particular plane, the AP source is a 2-position switch: NAV1 (which really means "GPSS" or "HDG" modes from the GPSS) or NAV2 (CDI deflection generated by NAV2). 

What could the Aspen be doing a mile from the MAP that would cause the AP to turn? The next waypoint is straight ahead in this case. The only thing I could think of is the heading bug is to the right of the heading, and for some reason the Aspen switches off GPSS. The AP is in HD mode when this happens. Why a change at this point? Is it trying to go to APPROACH mode while the STEC-30 doesn't know to switch to HiTrk automatically? I'd say at this point, Aspen tech support would be the only one that can provide answers.

I could be way of base here (easy pickoff) but the Aspen must use a source for the navigation data then apply the GPSS calculations to the output signal to the A/P. 

Cruiser wrote: I could be way of base here (easy pickoff) but the Aspen must use a source for the navigation data then apply the GPSS calculations to the output signal to the A/P.  If the IFD is supplying a route and the Aspen is in HDG mode applying GPSS corrections to that heading data, then there is an interruption in the signal from the IFD (or change of mode) from the flight plan laid in the IFD. Is the approach loaded and activated? is there a break in the sequence to the next waypoint?  It seems that anytime the NAV source data is interrupted the Aspen is not getting instruction and reverts to the raw heading bug on the HSI.  Try setting the heading bug to a specific number and see if the autopilot is commanding the turn to that point.

In the problem I was having with the deviation at a one mile final, the Aspen was in GPSS mode, the autopilot was in HD mode, and there was no fixes or transitions at that point.  With the Aspen off of GPSS mode, the autopilot in HI TRK mode inside the IAF, the approach was right on the numbers.