DFC 90 GS Activation for ILS

mvitalis wrote: I am having difficulty getting ILS GS activation in my recently installed DFC90 in an SR22 G2. However, if I am using HDG (I also tried HDG+NAV) up to this point, and then decide it is now time to arm the DFC90 to capture the localizer, if I press NAV, it turns outbound (!) to the IAF and will not capture the active leg. If I press GPSS, it will capture the active leg, but will not capture the glideslope (although it will arm the GS, however it remains cyan and never turns green). Pressing GS or APPR does not do anything. I am careful to be at the correct altitude and therefore am below the GS.

I'm in Maine-- So I guess going flying together is ruled out!

mvitalis wrote:    I will check this out, but the manual does say (even recommends) setting the CRS which, in any case, is ignored as the CDI is driven by the localizer  regardless what the CRS is set to (PFD Pilot's Guide page 29).

I flew the ILS 14 MFR approach many times today and got a better handle on what is happening.

You were right, the CRS was key. My testing wasn't exhaustive, but it seems that (most of the time), the CRS pointed to the current BRG to the target/initial waypoint when the approach was activated or loaded. If I switched the 430W CDI back to GPS then VLOC it would change to the then current bearing to SAMIE. Once I got "close enough" to the localizer, it would then switch to 143, the correct course for the ILS. Once I had the correct CRS, I could then activate the appropriate leg in the 430W then go from HDG or HDG+NAV to NAV and everything would work normally (i.e. it armed then activated the GS).

I don't have a complete handle on which combination of knobs, or where in the approach cycle, the CRS would correct. I would be flying down the approach, then break off, typically before AMASE (the "FAF"), cancel the approach by removing it from the 430, sometimes load a GPS approach just to make sure everything was cleared out, fly using HDG and ALT for a while while I got everything setup and then activated another ILS/SAMIE.

I did try using GPSS to fly to SAMIE rather than HDG, then activated the leg from SAMIE to AMASE and hit HDG, but it picked a CRS of 309 out of the air (that wasn't the current bearing to SAMIE), and so I had to start over.

I will typically be flying GPSS, then once I get vectored, will be using HDG until the approach will be loaded or activated, then switch to HDG+NAV then NAV. But frankly, until I can repeatedly get the correct CRS corresponding to the ILS, I'm not going to trust this system.

Maynard

Hi guys,

We should be able to wade into this thread when I return on Monday and double check a few things in the lab.

Steve

Hi Maynard,

I had a chance to look more closely at your report and verify a few actions in our integration lab.  

Before I get into that, I’d first like to thank you for your feedback and question set.  We do prefer for the professional IPs and trainers to provide instructional guidance and training.  I am however, very interesting in ensuring we adequately explain any behaviors you are observing with our equipment and determining if, and where, we can improve our Pilot Guides.

While I don’t recall any other field reports of “turning the wrong way” behavior, I will offer up two salient points/observations that I personally experienced during development flight test for the DFC90.

The first one is a carry-over from PFD-STec55X-Garmin430 configurations and has the same behavior in DFC90 equipped aircraft.  The most common scenario is that you are flying multiple radar patterns during practice instrument work and taking ATC vectors to instrument final.  If the pilot is in HDG mode of the autopilot during radar downwind while on ATC vectors and then arms NAV (so in the DFC90, you see HDG in green and NAV in cyan) before the aircraft heading is within 90 degrees of the final approach  course value, the Garmin 430 can pick up the localizer signal, and if close enough to the localizer beam, the system will say, “Ahah, I see a strong localizer beam so I will transition from HDG mode to NAV mode and track the localizer beam.”.   The issue here is that the PFD/DFC90 logic is such that aircraft heading is taken into account to determine if the system should be tracking the inbound localizer beam or the outbound localizer beam.   In this scenario, the approach has been properly loaded and activated in the Garmin 430 but the pilot has elected to arm the approach before being on some 45 deg like dogleg to final or when ATC has cleared you for the approach.   For example, I personally experienced a few cases where I tried to “get ahead” of the approach and while still on downwind but getting close to where I thought ATC would turn me to some kind of base leg, I activated the approach in the Garmin and then armed NAV in the autopilot.    Then, the Garmin would pick up the localizer beam, I would see the transition from HDG to NAV and the autopilot would turn outbound and track the localizer until I took a manual action. (One other point to consider here is that I intentionally wrote in the Pilot Guide that the recommended time to arm NAV is when ATC has cleared you for the approach.  That was intended to deal with the case where ATC either intentionally or unintentionally vectored you across final without explicitly clearing you for the approach so you wouldn’t want the HDG to NAV capture to occur.)

The second one is also a carry-over from the PFD-STec55X-Garmin 430 configuration and has the same behavior in DFC90 equipped aircraft.   Even if the autoslew (automatic ILS CDI output switch) capability of the Garmin has been turned on, if the pilot elects to manually select VLOC as the Primary Nav source either via the PFD Primary Nav line select key or the “CDI” button on the Garmin 430, prior to the autoslew action taking place, the angle of the last leg in the Garmin prior to the manual VLOC selection will be the value set in the PFD course window.  This would have the affect of the autopilot tracking a wrong (but dialed in) course.  In this case, the pilot is improperly intervening in the automation sequencing and not setting the desired course in the course window.

From the Avidyne perspective:

1.     It feels like the tracking outbound localizer signal experience described above may be what you observed on your first sortie;
2.    We do want to continue leaving the training up to the professional trainers;
3.    We think that if you take advantage of the autoslewing method and automatic transitions of the PFD and autopilot, you should see nominal operation.  In other words, the DFC90 did not insert a system-level “bug”, nor do we think your particular unit is somehow misbehaving;
4.    I will make some edits to the DFC90 Pilot Guide on its next iteration to include addressing the technique of starting an approach from the IAF and not just taking vectors to final, providing a user tip that covers the premature arming behavior described above, and trying to reinforce the explanation of course behavior if manually intervening in the automation sequencing.
5.    If you have other specific pilot guide improvement recommendations, I’d be thrilled to hear them and try to incorporate into the next revision as well.


BTW, one note/question from your original 20 April AvidyneLive posting.  In it, you wrote, “If I press GPSS, it will capture the active leg, but it will not capture the glideslope (although it will arm the GS, however it remains cyan never turns green).”  Per the top of page 4-3 of the DFC90 Pilot Guide, you should never see the GS arm (go cyan) while GPSS is the selected mode.  So, are you sure that really happened?   We designed the code such that this can’t happen.  The reason was principally that we can’t vouch for the same vertical protections in GPSS mode that are guaranteed in non-GPSS modes from the Garmin.)


Do you think I’ve correctly understood the problem you were seeing in the airplane?

Steve

Steve,

Thanks very much for your response.

I've posted this question here and on the forums at COPA where there are several instructors who responded, but none of whom could answer the question satisfactorily. In fact, there was enough variety in responses that suggested to me that there is considerable confusion about what to do concerning setting up for an approach (so there is an education opportunity!).

I also think there is sufficient confusion that there are several suggestion for clarifying your Pilot's Guide, which I'll make in a separate post.

I learned, from my testing, and as you confirmed above, that before the localizer is sensed, if VLOC gets activated (either manually or autoslew), then CRS is set to the current bearing to the IAF.

As Joe pointed out, I have typically LOADed the approach 30-50 miles out from the airport, depending on how soon I can get the weather, and will then ACTIVATE the approach before picking which leg I'm likely to get vectored to (this isn't always obvious depending on how far out you are and what direction you are approaching the airport from). This potentially causes the initial problem of autoslewing to the IAF rather than some intermediate point on the localizer.

But it does bring up the question that I somehow have not yet picked up: When does autoslew occur? When the approach is ACTIVATEd? Does anything happen when the approach is LOADed (presumably not)? If the approach is only LOADed, but not ACTIVATED, what happens (which happens when the pilot is not fully awake)? If out of localizer range and the approach is ACTIVATED and a leg has been selected (say SAMIE -> AMASE), is the bearing to AMASE then set as the CRS or the bearing to SAMIE (the IAF)?

The one trick I've learned from all this is that I can manually re-slew (new verb) once the localizer is alive by switching the CDI to GPS, then back to VLOC so as to get the correct CRS loaded.

But it still seems to me there is a hole in time where, depending on conditions, the autoslew, if not in localizer range, will set the WRONG CRS.

What am I missing?

Maynard

Steve,

Here are some thoughts on the Pilots Guide. Please note that it affects both the EXP5000 PFD guide as well as the DFC90.

Firstly, a few quotes which I feel sow confusion:

The Avidyne DFC90 Pilot's Guide section 4, page 39 says:

The CDI course is automatically set to the inbound localizer
course resulting in a hands-free transition if the navigator has
been set to autoslew (automatic ILS CDI output switch) and
VLOC was not manually selected by the pilot on the PFD Pri Nav
line select key. If the navigator hasn’t been set up to auto slew,
or VLOC was manually selected in the PFD Pri Nav field, it is
recommended that the inbound course be set using the PFD
course set knob to serve as a reference during the localizer
intercept and tracking.

The nomenclature is confusing. The PFD uses "CRS" in the Primary Nav window. No where in either Pilot Guide is CRS referenced, although the PFD guide does refer to it as "course".

I would recommend explicitly referencing "CRS" in both guides since this is what the pilot sees on the PFD. CDI is only "seen" on the 430 so this sentence is confusing since you are presumably talking about the PFD course and not the 430 course.

Also, line select key is not described anywhere (in either guide). This has been a discussion point at COPA as several people have tried to divine what LSK actually means.

I would recommend that you be more consistent in the use of your terminology, particularly across Guides for different products, and have a Glossary of Terms (the same for all) in all Guides that define what these terms mean.

The situation when autoslew is not selected suggests that entering the PFD CRS would be useful, but not necessary. This inferred (to me) that CRS was like setting the HSI course for the ILS, useful as a reference, but not necessary. But, as you pointed out, if this is slewed before its time, it could be wrong and therefore cause problems.

Question: If I change the PFD CRS to the correct inbound ILS course, does the AP then use that when I press NAV, or does the AP get that internally and the CRS is just a screen reference???

I recommend you clarify this statement to make more clear what really happens if the pilot manually presses the 430 CDI to select VLOC 1) before the localizer is alive, 2) after the localizer is alive. Also discuss if the pilot manually sets the CRS directly on the PFD, if that actually gets used or is for "reference only".

The Avidyne EXP5000 Pilot's Guide section 3.5 page 32, says:

PFD Nav Source = VLOC 1 or VLOC 2 and the current
frequency is an ILS or localizer. In this case, you can set a
course for reference. The CDI is driven by the received localizer
signal, regardless of the course set.

This suggests (to me) that you could set the PFD course to what ever you wanted and the DFC90 would still do the right thing, but it doesn't unless you cycle VLOC back to GPS and back to VLOC in which case the correct CRS will show up. I found this to be quicker than fiddling with the PFD knob to change the CRS directly.

I recommend that you coordinate the changes in the DFC90 PG with this statement here so its clear. But note that the PFD PG statement clearly suggests that it IGNORES what is manually entered in the CRS.

Finally, the DFC90 Pilot's Guide, Section 4-6 page 42 further clouds the issue by "recommending" that pilots use VTF, which we all know is a bad idea:

When flying a Vectors-To-Final approach, use the Heading knob
and the “HDG” button on the autopilot control panel when ATC is
issuing vectors.
NAV APPR can be armed prior to capturing the localizer beam
but it is highly recommended to wait until ATC clears you for the
approach before trying to arm NAV and APPR. To arm NAV
APPR while still in Heading mode, press the “HDG” and “NAV” (or
“HDG” and “APPR”) buttons at the same time. Once armed, the
system will automatically capture the localizer signal and
transition out of Heading/Vectors mode on its own.
□ Ensure the Primary Nav LSK is set to a VHF source and
tuned to a localizer
□ Ensure the front course is set in the Selected Course
window

This suggests several things which may result in the wrong conclusions. First, VTF is NOT considered appropriate by any instructor I've ever talked with or any student that has a reasonable amount of experience, much less an experienced IFR capable pilot.

Second, NAV doesn't capture the localizer signal, it uses what in CRS (which is exactly my problem!). If the localizer is alive when I hit NAV it will follow what is in CRS not the localizer signal!

Third, it's not clear what you mean by the "Selected Course window". I now know it is the PFD CRS. Also note, this is the first time (that I can recall) that "front course" has been used in this section - it would be useful to underline this is because you need to set the front course always including the case when a back course approach is flown (OK, this is a very minor point!).

Fourth, what happens if the front course in the Selected Course window is not correct? Re-slew? Manually change it? Is there a 3rd option?

I would recommend that you remove all mention of VTF and discuss all approaches in the context of real-world scenarios: selecting the most appropriate transition into the approach and optionally selecting and intermediate leg or waypoint along the approach, particularly when/if ACTIVATing versus LOADing the approach may cause confusion with regard to when autoslew happens.

Lastly, if the CRS is wrong for any reason, what should the pilot do?

Thanks for reading this far. I hope this is useful - it certainly has been useful for me to think through all this stuff on the ground since this process does not work in the air!

Maynard

Joe,

I think I've seen this also, but would have to confirm given all the other things I was testing.

BTW, thanks for your previous post (which I didn't respond to as I was waiting for Steve to clarify) about CRS for a leg being set to the ILS course (143) and not the bearing to the intermediate waypoint.

So do you think I (we?) scared him away, particularly with my (clearly overboard) previous long post about the Pilot Guide corrections?

I won't be able to get back in the plane to do more exploration until we return from Machu Picchu/Galapagos as we leave shortly.

Maynard

Joe,

Yes, good points.

Does your testing/experience indicate that:

1. The autoslew happens at ILS approach ACTIVATION time in the 430?
2. If "slewed before its time" and CRS is incorrect (for whatever reason), does setting PFD CRS directly result in the correct behavior when NAV is pressed and the localizer comes alive (i.e. CRS is not just for reference but actually gets used)?

Maynard

Yes, sometimes writing does not serve the topic best. Being in the same room sharing a white board would be more effective and efficient. However, part of the problem is we're using undefined terms where I suspect we may mean different things.

Autoslew is one such term. Garmin doesn't use it and Avidyne only uses it on 1 page each in of their EXP5000 and DFC90 Pilot Guides.

From the DFC90:

The CDI course is automatically set to the inbound localizer
course resulting in a hands-free transition if the navigator has
been set to autoslew (automatic ILS CDI output switch) and
VLOC was not manually selected by the pilot on the PFD Pri Nav
line select key.

From the EXP5000:

Avidyne recommends that you set the inbound course using the
EXP5000 Course Set knob to serve as reference during the localizer
intercept and tracking. This is automatic if the Navigator system has
been set to Autoslew. The CDI deflection will be driven by the
localizer signal itself, regardless of the course setting.

Lastly, the Garmin manual merely allows the operator to set either "automatic or manual ILS switching" without further explanation.

So far, no clarity there.

My impression (and experience, insofar as my testing has been mostly ILS, with some GPSS) was that "autoslew" was a singular event.

What is not clear is several things:

1. What activates it? Presumably, this is the 430 doing something based on some event that results in data being sent to Avidyne which clearly affects both the MFD and the PFD.
2. In the case of an ILS approach being selected, the supposition is that this occurs only once as "autoslew" does multiple things:
1. It sends data
2. It switches the 430 CDI from GPS to VLOC
3. From what you are saying Joe, once VLOC is switched (either auto or manual), the slews stop
4. If GPS is still selected on the 430 CDI, it continues to send data periodically, based on some event occuring (passing a waypoint perhaps). This may be true, but it is not clear to me why the behavior of "slew" should be different. I get that the AP must/should follow the localizer in the case of the ILS and the GPS navigator in the case of the RNAV/GPS approaches, but we're also talking about the PFD/MFD data.
5. When activating a GPS approach, I still see data changing, maps changing, the HSI changing on the PFD/MFD etc as we proceed through the approach. Presumably this data is coming from the 430. But the same stuff happens in the ILS approach - is this not also coming (slewed) from the 430?

Regarding the CDI/CRS behavior, I'll have to try that test out - it sounds odd and inconsistent with their manuals (as I understand them), but that's the ultimate proof of the pudding!

I'm hoping we do get a response from Steve to explain all this as we can test and wonder for quite some time.

Having thought about it for a while now, it is not clear to me while the behavior of switching to VLOC before it's time results in selecting the bearing to the IAF was considered by Avidyne as useful. But perhaps that happens as if the localizer is not detected yet, they may not receive the actual ILS course from the 430 so they have to do something else that's not completely useless. Still, the MFD DB has this info, so I'm wondering why that doesn't get used ...??

Maynard

Lost in the Mud

Steve-

mvitalis wrote: Autoslew is one such term. Garmin doesn't use it and Avidyne only uses it on 1 page each in of their EXP5000 and DFC90 Pilot Guides.

Joe,

Thanks for your detailed reply to my never-ending questions.

It does occur to me that there is sufficient complexity here that begs a manual/class that actually explains all this in sufficient detail such that the PIC has some hope of understanding what's going on and why.

I do know that (some) Cirrus Certified Instructors do not, so I'm wondering about what training they get (by virtue of the responses I received from them at the COPA forums).

You have clearly learned more than the average bear about how all this works; does this come from years of experience in the cockpit of GA/767s as an ATP or are there places/books one may have access to where this stuff can be learned, or what?

Frankly, I find unacceptable just having pro-forma instructions from the Pilot Guides and instructors over such a critical area. When things go bump in the night during an approach to minimums in difficult weather, then your life could be at stake. This information needs to be made more available, at least for those of us who want to understand.

Maynard (one annoyed pilot!)

My schedule seems to have served me well with respect to this thread.  I just read Joe's 29 Apr, 11:27am post and he seems to have addressed most of Maynard's questions with a fantastic, and very accurate, series of explanations.

I'm having some trouble bookkeeping any open questions/issues that still remain.   Is it accurate to say they are:

1.  Some Pilot Guide enhancements can be made to both the PFD and DFC90 Pilot Guides.  Specific examples include, cleaning up the use of CRS/PFD Course/430 Course; cleaning up the explanation of "line select key", clarifying what happens if the pilot manually presses the 430 CDI at various stages of an approach, and maybe a few others.

2. Clarifying when NAV will or will not capture a localizer signal and the impact on the dialed in PFD Course on localizer tracking.

3.  Clarifying the triggers for Glide Slope (GS) mode to arm in the DFC90.

Do you agree that those are the only open ones or did I miss some others?

Steve

AviJake wrote: 3.  Clarifying the triggers for Glide Slope (GS) mode to arm in the DFC90.

Maynard,

Let's start with a specific topic in this response.  Specifically, let me cover the arming and capture of VHF-based GS in this response.