Experiencing LPV approach failures

[John Collins]

My guess is that there is interference from your transmissions on the tower frequency.

KTTN tower 120.7
KABE tower 120.5
KLNS tower 120.9

These frequencies are close to the interference test frequencies of 121.15 thru 121.25. If you have a noisy transmitter, particularly a KX155 or KX196/197, they can leak energy thru the face of the unit and kill the GPS signal. You could also have a bad antenna. Some older ELT systems can also re-radiate from excitation around these frequencies, even though they are not receiving power.

Try keying the mike on 121.15 thru 121.25 for 30 seconds at a time while watching the satellite page. It is OK for the satellite bars to slowly decrease or even momentarily collapse, as long as they come back while you are still keying the mike and you don't get an integrity warning.

Do this with both Com 1 and Com 2 as either can be the source of the issue. If either kills the signal and you see an integrity error, you can try a notch filter designed for this purpose. I have bunches of them that I would sell cheap. They go in-line between the Com unit and the cable that goes to the Com antenna. Also, there should be a minimum of 36 inches between your GPS antenna and your Com antenna.

I have personally seen this on those tower frequencies. In my case it was an especially noisy KX197 leaking energy thru the display. You can verify if this is your case by placing an antistatic bag that avionics are shipped in, over the front face of the unit. If it works when the bag covers the front, but not when it is removed, you will have isolated where this interference is coming from.

 

[Henning]

It has occurred at 3 different Class D airports thus far. One in central New Jersey (KTTN), one in eastern Pensylvania (KABE), and one is south-central PA (KLNS). KTTN and KABE are about 30 nm apart and KLNS is about 75 nm from the other 2. So if its an external interference it has to be pretty widespread.

Until the report of NO GPS POSITION on the ground and then on climb-out at Lancaster (KLNS), it had only happened on LPV approaches at each of those 3 airports (and on every LPV approach in the last 2 weeks after having not flown any LPV approaches in many many months).

BTW, I did a short 1.3 hr breakfast flight today. No approaches. No problems.

That really reduces the chances of external issues. My course of action would be to visually inspect the antenna and full length of the cable run for any signs of chaffing, corrosion, or other damage. If everything there looked good, I would be talking to Garmin or your avionics shop about refreshing all the firmware and software. After that it's either a bad antenna, or a failure in the box hardware.

Real quick thought, have you tried pulling the box and reseating it a couple of times, maybe use some cleaner spray on the plugs? That would have been my first move...

 

[gasfiltered]

Did anyone check the NOTAMS? There were some WAAS outages for the area recently.

 

[bkspero]

My guess is that there is interference from your transmissions on the tower frequency.

KTTN tower 120.7
KABE tower 120.5
KLNS tower 120.9

These frequencies are close to the interference test frequencies of 121.15 thru 121.25. If you have a noisy transmitter, particularly a KX155 or KX196/197, they can leak energy thru the face of the unit and kill the GPS signal. You could also have a bad antenna. Some older ELT systems can also re-radiate from excitation around these frequencies, even though they are not receiving power.

Try keying the mike on 121.15 thru 121.25 for 30 seconds at a time while watching the satellite page. It is OK for the satellite bars to slowly decrease or even momentarily collapse, as long as they come back while you are still keying the mike and you don't get an integrity warning.

Do this with both Com 1 and Com 2 as either can be the source of the issue. If either kills the signal and you see an integrity error, you can try a notch filter designed for this purpose. I have bunches of them that I would sell cheap. They go in-line between the Com unit and the cable that goes to the Com antenna. Also, there should be a minimum of 36 inches between your GPS antenna and your Com antenna.

I have personally seen this on those tower frequencies. In my case it was an especially noisy KX197 leaking energy thru the display. You can verify if this is your case by placing an antistatic bag that avionics are shipped in, over the front face of the unit. If it works when the bag covers the front, but not when it is removed, you will have isolated where this interference is coming from.

I do have a KX155 as Nav/Com2, but on a couple of the approaches I turned it off and still experienced the problem. Ruled it out on the basis of that. Should I still do the keyed-mic test on 121.2 for it?

One of my 2 com antennas may be slightly under 36 inches from the GPS antenna (based on the AA5 maintenance manual it should be com2, but I don't know that for certain). It is 31 inches from the rear tip of the GPS antenna to the front edge of the tapered base of the antenna. It's another 1.5 inches to the base of the com antenna shaft. And another ca. 2 inches to the center of the GPS antenna. So it depends on how one measured. FWIW, I thought the install manual specified >=30 inches separation. The other com antenna is on the bottom of the airplane.

That leaves com1, the 530w. It's the one I was using to speak with ATC for all of the approaches. And its possible that its antenna is the one on the top. So I guess that if it is transmitting on one of the 120.X frequencies it could be interfering with the GPS antenna. The problem with that is that none of the signal loss events coincided with a transmission (or a reception, for that matter). I'm guessing, but my estimate is that there was at least 30 seconds and usually more like a minute + between the last transmission and loss of satellite signals. Knowing that, is com1 still a plausible culprit?

The ELT antenna is new (installed about a month ago). And the ELT is a 121.5 Ameri-King 450, which I have read are particularly prone to this problem. It's antenna is located 6 ft away from the GPS antenna, though. Could it still re-radiate enough from that far away?

Finally, should i be worried about radiation from the face of the 530w? Are they known to have this issue?

FWIW, today I did a test using my yoke mounted tablet. I set it to display satellite bars using its internal GPS. I flew 2 RNAV-16 approaches into Trenton. One actually went to completion. The 530w satellite strength bars dropped twice to about 25% of the normal instensity for an instant and then recovered each time (about half way from the FAF and a crossing fix, and shortly after that). The second time the 530w satellite bars dropped to zero inside of the FAF and the unit called to ABORT the approach.

The satellite bars on my tablet's display did not change at all during this period. In light of that, do you still think that cabin radiation is a significant part of the issue?

 

[bkspero]

That really reduces the chances of external issues. My course of action would be to visually inspect the antenna and full length of the cable run for any signs of chaffing, corrosion, or other damage. If everything there looked good, I would be talking to Garmin or your avionics shop about refreshing all the firmware and software. After that it's either a bad antenna, or a failure in the box hardware.

Real quick thought, have you tried pulling the box and reseating it a couple of times, maybe use some cleaner spray on the plugs? That would have been my first move...

I meant to pull the box and work the connections in and out a few times today but forgot my long 3/32 allen wrench at home. Will have to do that tomorrow. I have talked to the shop and they will start on it this week. I think the first thing they are going to try is to swap in another 530w and see how it behaves.

 

[bkspero]

Did anyone check the NOTAMS? There were some WAAS outages for the area recently.

Nothing that I saw on my online briefings (1800wxbrief.com) and for good measure today I called FSS for a phone briefing. Other than satellite 4 being out of service, everything else was ok (as far as the briefer could find). I did not check RAIM today, but did early on both online before flying and in the air using the 530w. Both said RAIM was good.

My photos several posts above are too blurry to see this detail, but just about every bar had an inscribed letter D indicating differential lock (which I understand is WAAS lock). That only went away when the bars disappeared.

Do you still think that a WAAS outage is a reasonable possibility in light of this information? Is there a different spot other than a AFSS briefing where I should be looking for satellite NOTAM>

 

[Henning]

I meant to pull the box and work the connections in and out a few times today but forgot my long 3/32 allen wrench at home. Will have to do that tomorrow. I have talked to the shop and they will start on it this week. I think the first thing they are going to try is to swap in another 530w and see how it behaves.

That is a wrench that should have a copy with a home in the plane. It's amazing how many issues this solves in avionics.

 

[gismo]

I do have a KX155 as Nav/Com2, but on a couple of the approaches I turned it off and still experienced the problem. Ruled it out on the basis of that. Should I still do the keyed-mic test on 121.2 for it?

Yes. Something else (or even a powered off KX155) could be rectifying transmissions from a strong ground station and making harmonics that interfere with your GPS. In fact I'd try transmitting on the frequencies John recommended on both of your comm radios.

The ELT antenna is new (installed about a month ago). And the ELT is a 121.5 Ameri-King 450, which I have read are particularly prone to this problem. It's antenna is located 6 ft away from the GPS antenna, though. Could it still re-radiate enough from that far away?

Yes. The separation requirements are intended to prevent overload of the GPS antenna preamp from out of band signals. A harmonic that's in band can interfere from a much greater distance.

And ELTs are a common source of the re-radiated harmonic problem. You can eliminate that possibility by disconnecting the ELT from it's antenna.

 

[John Collins]

I would not conclude anything until you run the com test on both units. This will let you know if the interference is coming from your own aircraft. Otherwise, it could be coming from the ground or a bad antenna. It is most likely coming from your own aircraft.

 

[Henning]

I would not conclude anything until you run the com test on both units. This will let you know if the interference is coming from your own aircraft. Otherwise, it could be coming from the ground or a bad antenna. It is most likely coming from your own aircraft.

Why do you suspect comms interference as highest?

 

[bkspero]

I would not conclude anything until you run the com test on both units. This will let you know if the interference is coming from your own aircraft. Otherwise, it could be coming from the ground or a bad antenna. It is most likely coming from your own aircraft.

Ok, will pull and reseat 530w a few times and do com tests on 121.2 for both com's. If the approaches work, then it's probably the connections. If the signal fail in lockstep with keying the mic, then it's my radio and I will disconnect the ELT antenna and retest to see if it is involved. If it fails in lockstep with radio keying, then it's probably a radio issue.

If it fails, but not when keying the mic, then it's probably not the radio.

 

[bkspero]

Ok, will pull and reseat 530w a few times and do com tests on 121.2 for both com's. If the approaches work, then it's probably the connections. If the signal fail in lockstep with keying the mic, then it's my radio and I will disconnect the ELT antenna and retest to see if it is involved. If it fails in lockstep with radio keying, then it's probably a radio issue.

If it fails, but not when keying the mic, then it's probably not the radio.

Did as described above and John gets the prize. With ELT connected, keying mic on 121.2 for either Com1 or Com2 causes all satellite signal bars to drop to zero even on the ground in ENR mode. And they stayed at zero until I stopped transmitting.

Disconnecting the ELT and repeating the test on the ground and there was no effect from Com1 (the 530w, which I think has its antenna on the underside of the fuselage), but Com2 still caused signal loss (its antenna is about 34 inches from the GPS antenna on top of the fuselage). In the air with the 530w tuned to Trenton (KTTN, 120.7) I was able to fly an LPV approach to the edge of the Delta (about half way between HORDE, the FAF, and CULEV without any decrease in signal intensity (I kept Com2 tuned to the ATIS, 126.775). After breaking off the approach I retested both coms on 121.2 and again only Com2 caused signal loss. Signals had recovered by the time I reached the pattern at Princeton and I tested Com2 again and this time it did not cause signal loss(???). On the ground just before shutdown I retested both coms again and neither Com1 nor Com 2 caused signal loss (???).

I reconnected the ELT, restarted the engine and retested both on 121.2. Both coms caused signal loss.

So, it is clear that the ELT is a big part of the problem, and correcting it would eliminate transmission issues with Com1. And it makes some sense as to why the issue appeared recently, because, as I wrote in an earlier post, a new ELT antenna was installed a month ago because the old one (which was the wrong one for my ELT) broke. Maybe the old one was less susceptible to resonating with the GPS antenna.

The story with Com2 (the KX155) is less clear. I can't explain why sometimes it causes satellite signal loss and sometimes not. Anyone have any suggestions?

John, you wrote earlier about installing notch filters on the com antenna(s). Exactly what type and on which radio would you recommend (the shop tells me that the 530w already has one built in, but it clearly isn't enough to deal with the signal when the ELT is connected). Will a filter on the ELT antenna help with it, as well. You can respond here, or send me a PM, or an email (I will send you my email address by PM).

 

[John Collins]

Why do you suspect comms interference as highest?

Multiple locations, all with tower frequencies near the frequencies which cause the issue. In my experience, a good percent of aircraft fail the Com test and there are multiple failure modes, including antenna location, ELT, noisy Com unit, etc. Even if the test was performed at installation, systems degrade over time. In my case, the issue was coming thru the display and although was not antenna related, the only way of fixing it was to replace the Com unit with a newer one of the same model.

When these systems fail due to interference from the Com system, it takes some amount of time before they recover and that is dependent on the mode, so to recover from enroute or TERM mode is much quicker than to restore LPV.

 

[bkspero]

An update. Installing the GPS notch filters from John on the KX155 (Com2) antenna coax and ELT antenna coax substantially reduced interference when broadcasting on 121.2 (or 120.7) on Com2 to the point that 3D differential lock is not lost. Before the filters were installed, broadcasting on those frequencies on Com2 would instantly kill the GPS signals whether the ELT was connected or not.

Broadcasting on those frequencies on Com1 (the 530w) still killed the GPS signals when the ELT was connected (even with the filter installed). Covering either the KX155 or 530W face with aluminum foil did not change behaviors. It did seem to take longer to lose satellite lock due to Com1 interference (compared to pre-filter behavior). I was able to fly an RNAV approach to KABE even broadcasting on Com1. It did downgrade to LNAV, though, after calling the tower inbound on the approach.

Disconnecting the ELT and removing it from the plane eliminated the issues with Com1 (as it did before the filters).

My shop tells me that the 530W is supposed to have a GPS notch filter built into its output, so our next step is to determine if mine has failed (and as a result is exciting the ELT) or if it is behaving as a normal 530W and even the filtered ELT is still too sensitive to external excitation.

Before we go to an ELT swap we are going to swap in another 530w to see if it also loses satellite signal when broadcasting on 121.2 and the ELT connected. If it does, then my next step would be a new ELT. If swapping 530w's eliminates the interference even with the ELT connected, then its time to correct an antenna filtration issue on my 530W.

Make sense?

 

[chemgeek]

Definitely have your 530 checked out. I had a similar failure mode in my 430 where LPV approaches were downgraded to LNAV. At first it was intermittent then it became universal. Turns out I had a failure of some unspecified component in the vertical guidance circuit. In addition there was a hardware/motherboard update that Garmin performed while they had my unit. No issues since repair, and the loaner unit worked fine, too, in my tray.

 

[bkspero]

Thanks for the input, Chemgeek.

I have had another thought reading about notch filters. As I now understand them to work, they take signal in the notch frequency range (in this case about 1574 MHz) to ground on the coax. Enough so with the filters that I got from John provide about 50-60 dB of signal loss in that range while losing only about 0.1 dB of signal in the VHF range.

So it seems to me that if we added a 2nd notch filter to the ELT antenna coax (say at the antenna end with the first at the ELT end) we should see a total of 100-120 db of signal loss at 1575 MHz with only 0.2 dB at 121.5 MHz. That seems like a pretty good trade-off and worth testing. What do others think?

 

[John Collins]

Thanks for the input, Chemgeek.

I have had another thought reading about notch filters. As I now understand them to work, they take signal in the notch frequency range (in this case about 1574 MHz) to ground on the coax. Enough so with the filters that I got from John provide about 50-60 dB of signal loss in that range while losing only about 0.1 dB of signal in the VHF range.

So it seems to me that if we added a 2nd notch filter to the ELT antenna coax (say at the antenna end with the first at the ELT end) we should see a total of 100-120 db of signal loss at 1575 MHz with only 0.2 dB at 121.5 MHz. That seems like a pretty good trade-off and worth testing. What do others think?

I would try one on the GNS430W first, as it is the source of the remaining harmonic energy. Yes the GNS430W has a built in notch filter, but its broadcasts are still exciting the ELT. You seem to have made progress with the KX155 and ELT, so why not try the GNS430W?

 

[bkspero]

I would try one on the GNS430W first, as it is the source of the remaining harmonic energy. Yes the GNS430W has a built in notch filter, but its broadcasts are still exciting the ELT. You seem to have made progress with the KX155 and ELT, so why not try the GNS430W?

I considered that, John, and thought that if the issue was GPS range noise coming directly from the 530w it would interfere without the ELT installed. So the problem must be non-GPS range signal coming from the 530w that is exciting the ELT to broadcast noise on 1575.42 MHz.

A GPS notch filter on the 530w output wouldn't, I think, filter non-GPS range frequencies, so even if the 530w was broadcasting noise (outside of 1575.42 MHz) that was exciting the ELT, a 1575.42 notch filter on the 530w wouldn't suppress that excitation (since it only filters near 1575.42 MHz).

That's where I came up with the thought that maybe 55 dB of filtration isn't enough on the ELT if it is being strongly excited by VHF from the 530w. And that maybe another 55 dB would do the trick.

Either way I expect I will need another GPS notch filter. Before I pull the trigger, though, I posted my question about installing 2 notch filters in line on the ELT on a Ham Radio forum. If they say that there is no downside other than another 0.1 dB of insertion loss in my ELT VHF signal, I will be asking for another filter to try first in the ELT line, and if that fails, in the 530w output (recognizing my propensity to be wrong).

 

[fiveoboy01]

I'm glad I pay avionics guys to deal with this because I don't understand a word of it

 

[John Collins]

I considered that, John, and thought that if the issue was GPS range noise coming directly from the 530w it would interfere without the ELT installed. So the problem must be non-GPS range signal coming from the 530w that is exciting the ELT to broadcast noise on 1575.42 MHz.

A GPS notch filter on the 530w output wouldn't, I think, filter non-GPS range frequencies, so even if the 530w was broadcasting noise (outside of 1575.42 MHz) that was exciting the ELT, a 1575.42 notch filter on the 530w wouldn't suppress that excitation (since it only filters near 1575.42 MHz).

That's where I came up with the thought that maybe 55 dB of filtration isn't enough on the ELT if it is being strongly excited by VHF from the 530w. And that maybe another 55 dB would do the trick.

Either way I expect I will need another GPS notch filter. Before I pull the trigger, though, I posted my question about installing 2 notch filters in line on the ELT on a Ham Radio forum. If they say that there is no downside other than another 0.1 dB of insertion loss in my ELT VHF signal, I will be asking for another filter to try first in the ELT line, and if that fails, in the 530w output (recognizing my propensity to be wrong).

I would try it anyway. After all, your KX155 no longer introduces the issue and you installed a notch filter in its transmit output. It is easy to check out, just temporarily move the notch filter in the KX155 antenna line to the GNS430W. You can theorize or you can test. I would test.