Deviation from magenta line in GPSS (GTN750, KAP 150 and GDC31 GPSS)

[M252]

Hi,

I have had a GTN750 and a GDC31 GPSS roll-steering converter in my M20K 252 for a few years. Most instrument approaches I fly are ILS approaches, however, when I do fly a GPS approach I see that the plane has an offset to the left of the magenta line (~ a couple hundred feet). This offset is also visible enroute where its doesn't matter. It does matter for approaches and I am wondering if anyone knows how this offset can be adjusted?

Thanks for any help along these lines!

[StevenL757]

Some questions...

- Do you observe the issue only on a GPS approach, or both ILS and GPS (sounds like the former, but looking for clarity) ?

- Are you flying the GPS approach manually or with autopilot?

- Have you tested other GPS approaches to see if the same behavior appears across some or all?  If so, how many approaches have you tested, and what were your observations?

- Did the issue start just recently?  If so, have you had any work done to the airplane that would correlate with the behavior you're observing, and if so, what type of work?

Steve

[jclemens]

More info please.  What other equipment is installed? Is this when coupled?  What autopilot? 

[StevenL757]

He has a KAP150...above.  :-)

[M252]

Hi,

- I see the offset in GPSS mode with the KAP150 getting heading information from the GDC31 GPSS converter and in APPR mode during the GPS approach

- I always see that deviation enroute in GPSS mode (always left of the magenta line)

- I have tried only a couple of GPS approaches

- ILS approaches looking good (on localizer)

- Nothing changed on the plane nor on avionics/servos/A/P etc.

[StevenL757]

At the risk of drawing out the thread and speculating, you may want to read this article by Don Kaye (participant/contributor on this forum).  Although I've never met Don, he is a very well-seasoned CFI and consultant to the Mooney community on a wide variety of topics...this being one of them.

http://www.donkaye.com/donkaye.com/GPS_Approaches_Using_GPSS.html

It talks about the GDC31 converter; albeit, with a KFC-200 autopilot in this scenario, and provides some guidance - possibly around the issue you're encountering.  Have a read and see if this addresses the concern.

[M252]

Thanks Steven, will do.

[Marauder]

1 hour ago, StevenL757 said:

At the risk of drawing out the thread and speculating, you may want to read this article by Don Kaye (participant/contributor on this forum).  Although I've never met Don, he is a very well-seasoned CFI and consultant to the Mooney community on a wide variety of topics...this being one of them.

http://www.donkaye.com/donkaye.com/GPS_Approaches_Using_GPSS.html

It talks about the GDC31 converter; albeit, with a KFC-200 autopilot in this scenario, and provides some guidance - possibly around the issue you're encountering.  Have a read and see if this addresses the concern.

I actually think that what is posted on Don's site was written up by John Collins who is a contributor on all matters avionics and one the advisors to the FAA on the ADS-B implementation. Don may have been the contributor the additional content.  It is amazing how something to simplify our flying makes it a bit more complicated.

 

[N201MKTurbo]

An autopilot sums the error signal from the radio (this includes the signal from your GPSS) with the error from heading source. If your heading bug is not at the top of the DG, it will introduce an offset. 

I need to double check, but I don't think the heading error is summed while in localizer mode.

So, try flying with the autopilot in localizer mode and also try removing the offset by adjusting the heading bug a bit left or right.

[M252]

So where can you adjust for that offset? In the GTN750, the GDC31 or the KAP150?

[N201MKTurbo]

The offset I'm talking about will happen with a perfectly adjusted autopilot. 

I was talking about changing the heading bug a little left or right. If the heading bug is not aligned with your heading it will add an offset.

[M252]

OK, thanks. I thought that the heading error of the heading bug does not matter in GPS mode... Is that correct? http://dacint.com/wp-content/uploads/2015/06/DAC-GCD31-flyer-09-12-13.pdf

[N201MKTurbo]

OK, that makes sense, it replaces the DG signal. 

Most autopilots have a heading offset adjustment. If you had that adjusted for the GPSS it would fly without an offset while coupled to the GPS but would have an offset when in heading mode. 

Maybe there is a hardware or software adjustment on the GCD31.

[N201MKTurbo]

Here is the installation manual:

http://dacint.com/wp-content/uploads/2015/08/GDC31-Installation-Manual-Rev-V.pdf

it doesn't say anything about offset adjustment.

[StevenL757]

7 hours ago, Marauder said:

I actually think that what is posted on Don's site was written up by John Collins who is a contributor on all matters avionics and one the advisors to the FAA on the ADS-B implementation. Don may have been the contributor the additional content.  It is amazing how something to simplify our flying makes it a bit more complicated.

 

Ah...gotcha.  Thanks for the correction.

[jlunseth]

I don't have your exact combination, I have a 430AW+ KFC 200 + Icarus SAM, but they work pretty much the same.  It doesn't make any difference where your HDG bug is set if the GPSS is in GPSS mode rather than HDG mode.  In GPSS mode, the GPSS is sending the signal from the 430 (or in your case the 750) through the HDG bug circuit to the AP and the bug itself is disabled.   

As for your offset, I don't have a good answer for you, although I don't think it makes any difference.  You need to switch to APPR mode before the FAF, and once coupled, the magenta line and the GPSS have nothing to do with the approach.  The AP is using the RWY waypoint to set a course just as it would use the localizer to do so in an ILS.  You need to set the course needle to the inbound approach course so that it can do that.  The AP is then "pointing" to the end of the runway and the AP itself is keeping you on the final approach course without any help from the GPSS.  The only way I could see that an offset might affect it, is this:  the GPSS course you fly before the FAF is the equivalent of the HDG course you fly to intercept the localizer.  So if you were doing an ILS, you would be given a course to fly to intercept the localizer, and eventually a clearance to intercept the localizer.  You set the HDG bug to the intercept course, and then when cleared to intercept, you would press the APPR button and the annunciator would tell you it is ARM, in other words, it is letting the HDG bug fly the plane until intercept occurs, and then you are going to see CPLD and the HDG bug will not longer be in charge, the course needle and APPR mode will be pointing to the localizer and will be in charge.  The GPSS course you fly at the beginning of an LPV serves the same function, the AP uses the signal from the GPS, through the GPSS, to intercept the final approach course, but at the final approach course the AP will couple to the approach and the GPS/GPSS signal will no longer be in charge.  Assuming you correctly dialed in the final approach course, the AP will be pointing at the RWY waypoint.  The only thing you might see is if the offset is too much, the aircraft may never get close enough to the final approach course to couple and will stay armed, which means the GPSS is guiding you on the final approach course which should not be the case, and it also means you will not get coupling to the glide slope.  So if that is what is happening, then I would take it in to your avionics shop to have it corrected.  In the meantime, you can overcome the problem by using the HDG bug to fly the intercept course without the GPSS.  Either way, if the AP is able to couple to the final approach course, then it is not worth worrying about, the only thing the AP is using at that point is the RWY waypoint.

Now, if the RWY waypoint itself is offset for some reason, that would be a problem.  I would fly it in VMC to see if you end up at the runway, or offset from the runway by some unacceptable amount.   

One issue that I experienced early on was switching to APPR too late.  If I make the switch after the FAF in my aircraft, sometimes the AP just won't couple even if it is on the correct final approach course.  I read (I think it was a John Collins article) that this happens because the ARM function calculates and uses a closing rate to the final approach course.  In other words, you are flying at some angle to the course, it calculates the rate at which you are closing to the course, and uses that to switch to CPL.  And if you are already exactly on the final approach course there is no closing rate, so the AP does not know what to do and never couples.  This is solved if you make sure to switch the APPR before the FAF.

[M252]

Thank you jlsunseth. This is a very useful explanation of what happens.

 

Yesterday I tried to replicate the behavior during two RNAV approaches and saw the following.

- I flew the standard approach with GPSS until about 3 NM before reaching the final approach fix (SB141 on the photo). As always the plane was slightly left of the magenta line. The heading bug was set to the final approach course. By the way the wind was cross by about 50° from the left.

- Then, I switched the GDC31 on the A/P SEL button from GPS to HDG and pressed APPR on the KAP150

- In both cases after pressing APPR the plane banked left by about 15-20 degrees leaving the approach course but eventually captured the "virtual localizer" and GS. This is clearly no problem in VMC but can be quite scary in IMC (I would usually disengage the A/P at that point and fly manually).

- The A/P kept the plane left of the approach course (as always)

Could this be due to the autopilot not being set up properly? Do others see this with the KAP150?

Edited May 13, 2016 by M252

[jclemens]

One: disengaging your gpss is incorrect. Leave it on, just hit the APP button when you are on the intercepting leg inbound. By selecting heading the autopilot loses its wind correction angle. Leaving it gpss will make the transition smoother. 

Two: your autopilot needs to be aligned. Take it somewhere that's qualified to do so, or call autopilots central. 

[jlunseth]

First, I don't know why you would switch from GPSS to HDG mode on the GDC, before the FAF, unless you were just running a test.  You should be able to leave the GDC in GPSS mode, press APPR, and the GDC will fly the plane to intercept with the virtual localizer.  At intercept, and when the AP couples, the course needle and the AP will then automatically take over.  Unless, as I said in my earlier post, the GPSS course is not getting you close enough to the final approach course for the AP to couple in which case the HDG bug trick will do it.  

It is not unusual for the plane to bank and adjust at intercept, when it changes over from the GPSS course to the APPR course.  In my aircraft, 15-20 degrees is a standard rate turn, so not a reason to take control back unless it persists (the bank goes higher than standard rate or the plane is truly flying off course).

I don't pay much attention to the magenta line during the final part of the approach, and here's the reason.  Let's take one of my favorite RNAV practice approaches as an example, which is the RNAV 31 at KGYL.  The final approach course on the chart is 308, or at least it used to be, they just changed the plate a little.  But my 430 would consistently tell me to turn to a different number (there is an annunciator strip at the bottom of the 430).  That course would usually be 306, but sometimes I would see something like 304.  The 430, in other words, has computed a different final approach course than what is on the plate.  No one has been able to explaint to me why that is the case, and the 308 course is the one that the FAA has used in clearing the approach path, so I always dial in what is on the chart even though it is slightly different than what the GPS says.  The magenta line is driven by the GPS, so the magenta line will be a little different from my final approach course, I just know that and have never looked at the line to be honest.  I am focused on the HSI.

There is something baffling going on here though.  I am wondering why, if your HSI is showing the course to be to the right, the plane would bank left to couple.  It seems as though the AP thinks the course is left of what it is supposed to be.  So I am wondering if the problem is in your AP, and not in the GPSS or GPS.  

This is the point where I would take the thing in to my avionics shop and have them straighten it out.

[jlunseth]

11 minutes ago, jclemens said:

One: disengaging your gpss is incorrect. Leave it on, just hit the APP button when you are on the intercepting leg inbound. By selecting heading the autopilot loses its wind correction angle. Leaving it gpss will make the transition smoother. 

Two: your autopilot needs to be aligned. Take it somewhere that's qualified to do so, or call autopilots central. 

What he said, but he was alot more succinct.

And I just checked the new plate for the RNAV 31 at GYL and I see the final approach course is now 304, so my GPS would be telling me to fly 302 or something slightly less.  You get the point.  The GPS final course and the magenta line are not what is on the chart.

Edited May 13, 2016 by jlunseth

[PTK]

I don't think it's the autopilot at all.

I think your symptoms are due to the GDC31 gain being set too low. The GDC31 computes the bank angle it needs according to the GPS data, and uses the reference voltage from the autopilot with the gain (scale facor I think they call it) to closely match it to the HSI heading signal. IOW you may have too large of a mismatch between HSI and ap signals due to gain set incorrectly. The result is the ap cannot accurately reproduce the intercept angle to capture what the GPS is telling it to do. Your KAP 150 (I have the KFC150) will bank 30 degrees max. With the gain too low it never banks enough to intercept. 

Your installer may have not been familiar with the GDC31. I think you need to take it to someone who knows how to configure it properly. It's a straight forward installation. Especially so with the King HSI and ap. 

[M252]

Thanks everyone for the help. I will take it to a proper avionics shop to get either the GDC set-up and/or the A/P fixed.

When I said 15-20° left bank I actually meant the course difference it would bank to (i.e. from 135 to 115 or even less in this case, which seemed a lot to capture the virtual localizer since I was already left of the approach course line).

4 hours ago, M252 said:

GDC31 on the A/P SEL button from GPS to HDG and pressed APPR on the KAP150

If I do not disengage the A/P SEL button from GPS to HDG how will the A/P get a HDG signal from the HSI? It seemed to me that I would have to do that to fly the RNAV approach. I thought this is what Jim Collins says: http://www.donkaye.com/donkaye.com/GPS_Approaches_Using_GPSS.html

[jlunseth]

No, you don't need to do that to fly the RNAV approach.  The AP is already getting a HDG signal because that is how the GPSS signal is fed to the AP, it is fed through the HDG circuit.  Command is then switched to APPR mode and the course needle when the AP couples to the approach.  There is sometimes some banking or adjustment at that point because the AP goes from one nav source (the GPS/GPSS through the HDG bug) to another (the course needle on the HSI, pointing to the RWY waypoint).

This still confuses me though: "When I said 15-20° left bank I actually meant the course difference it would bank to (i.e. from 135 to 115 or even less in this case, which seemed a lot to capture the virtual localizer since I was already left of the approach course line)."  The picture you posted shows the aircraft already left of course, whether you look at the magenta line or at the course needle on the HSI.  According to the HSI you would need to steer right to get on course.  If it were just the magenta line saying the course is to the right, well fine, the line is painted in the wrong place.  But what I am reading and seeing on your HSI, is that the HSI itself says the course is to the right, but then the aircraft steers left to intercept?  That confuses me.

Edited May 13, 2016 by jlunseth

[M252]

Thanks jlunseth.

- Next time I will leave the GDC31 in GPS mode. One very helpful thing I learned today :-) thanks again.

- I am also confused why the AP banks to the wrong side. I will ask an AP expert next week.  I have observed this several times... I will post what I find...

[jlunseth]

I looked at your picture again and it looks normal.  You must have taken it after the AP was coupled to the approach.  The HSI is showing the course to the right of the aircraft, the magenta line is showing the same thing, and both the AI and the turn coordinator are showing the aircraft banked slightly right, as if it is trying to get back on course to the right.