September 11, 1999
Thomas G. Lusch
2185 Olde Sawmill Bl
Dublin, OH 43016-8221
Tel 614-889-9809 (Voice & Fax)
Email:tomlusch@columbus.rr.com
Nicholas J. Talotta
Technical Program Manager DOT/FAA
FAA Technical Center
Atlantic City Airport, NJ 08405
RE: Transponders becoming invisible
Dear Mr. Talotta,
I had a chance to speak with you briefly on the flight line at this year's EAA Airventure convention (I was the one proudly wearing my pink Oshkosh Tower ATC controller's shirt). I had a great time being part of FAA's ATC Team that helps to make AirVenture a safe operation.
Because of your interest in transponder operations, I promised that I would forward to you a report I wrote several years ago concerning aircraft not appearing on ARTCC controller scopes. As I was preparing to send that to you, I came across the report you wrote entitled "A Field Study of Transponder Performance in General Aviation Aircraft," dated Dec 5, 1997. I found your report extremely interesting, especially your key finding (page vii) that nearly 1 in 5 transponders "...would create functionally significant problems when interacting with ground Secondary Surveillance Radar (SSR) processors, TCAS, or both." You elucidate that behavior when you describe the "Terra Problem" wherein "...a design characteristic of certain Terra transponders prevent them from responding consistently to...interrogations." During your study you were surprised to learn that some Narco AT150 transponders exhibited similar problems. You stated "Prior to this study, no reports have been published to suggest that the 'Terra like characteristic' had been observed in other transponder models."
THE PURPOSE OF THIS LETTER IS TO REPORT TO YOU THAT YET ANOTHER MAKE OF TRANSPONDER EXHIBITED THAT "TERRA LIKE CHARACTERISTIC." I documented this anomaly when I posted the following message on CompuServe's AViation Special Interest Group (AVSIG) forum...
Message: #852440, S/3 Air Traffic Control
Date: Sat, Jul 15, 1995 1:11:18
Subject: Xpdr problem @ 1 tracon
From: Thomas Lusch/CMH 76545,2004
To: ALL
Here's a baffling radar problem...
Around two or so weeks ago one of our club aircraft began to get bad transponder reports from ATC. Columbus Ohio (CMH) approach control would assign us a squawk, but they wouldn't "see" our transponder reply. Members began "squawking" <g> about the problem.
On Wednesday I took the "bad" transponder (a King KT-76) to Capital Aircraft Electronics at Port Columbus Airport. They let it run for several hours, running it hot and running it cold, and it "bench checked" with flying colors. They stated that the problem might be in the aircraft system. On Thursday I flew the aircraft over. Their test equipment showed it working just fine in the aircraft. They released it to me saying it was performing as per specifications.
Shortly thereafter I departed CMH and flew west for 25 or so miles. The transponder's reply light indicated it was replying to interrogations. However, not even once did any of my transponder replies show on CMH's radar! I was a primary target the _entire_ way.
>>>> Here's where it really gets interesting.... <<<<
Directly after terminating radar, I called Dayton (DAY) approach, asking for a transponder check. They assigned a code and my transponder replies showed up on their radar just fine. I immediately switched back to CMH to see if they were seeing my DAY assigned code, but I still only showed as a primary target!
After a fuel stop, I departed and flew midway between CMH & DAY. DAY saw my transponder replies just fine. CMH didn't. I then checked with Indianapolis Center, and they saw my CMH assigned squawk just fine! CMH didn't see my transponder replies at all that Thursday.
This is particularly hard for me to understand, as I happen to be a controller at CMH. Nearly three weeks ago our new ASR-9 was commissioned. From my vantage point behind the scope I am very impressed with how much _better_ the ASR-9 is compared to the ASR-8 that we had been running. Our new radar has a much cleaner presentation. We see primary targets much more reliably, and the same goes for transponder targets. We can see aircraft much more reliably, at a greater distance, and at lower altitudes. (And the precip display is a quantum leap beyond the old ASR-8!)
As you may have ascertained, our club aircraft's transponder problem appears to be coincident with about the time CMH's ASR-9 was commissioned.
I took the aircraft over to CMH again on Friday. I spoke to the FAA radar techs about our problem, and they find it very strange. They hadn't received any other such problem reports. Capital Electronics ran their tests again, and everything appeared just fine. With a spare transponder in hand, myself and a Capital Electronics technician decided to fly it, and see if the problem was still there. We were prepared to swap out the transponder while in flight. As you might suspect, our "bad" transponder performed flawlessly.
We're all really baffled by this anomaly. If anybody out there might be able provide some guidance with this problem, we're all ears...
Tom
The only replies to the question I posed back in 1995 dealt with similar anecdotal problems...but no explanations. The radar technicians never did provide a plausible theory to this mystery. The FAA certified avionics shop was baffled too, especially since the transponder passed their repeated tests with flying colors. Then, some time later, the problem mysteriously went away. All indications pointed toward the newly running ASR-9 radar...but then something changed. It was sort of like the transponder magically got better. It didn't make sense. I now understand, having read your report, that modifications have been made to the Mode S ground radar interrogations (as described on page 18) to fix the "Terra problem." That is the most plausible explanation as to why these anomalies occurred.
I feel it is important to bring this to your (and the industry's) attention because it attests to the possibility of a more widespread problem in the fleet of GA transponders than may be realized or reported. This concerns me a great deal, as proper operation of transponders is the backbone of the ATC radar beacon & TCAS systems. It is disconcerting to realize that possibly 1 in 5 transponders, those that exhibit this "Terra like characteristic," are likely to be invisible to TCAS equipped aircraft!!!
The pieces of this puzzle all point to Mode S interrogations.
On an unrelated, yet extremely similar topic, I have enclosed a copy of my paper, "Real Targets - Unreal Displays..," as reprinted from the March '92 issue of the Journal of Air Traffic Control. That paper has nothing to do with transponders, or radar, or Mode S. It is purely a radar data processing problem...a result of ARTCC software not being able to handle the voluminous amount of excellent radar data available. However, it results in a similar condition...aircraft becoming invisible to the controller. Having visited Cleveland Center just last June, I observed that the problem still exists.
Sincerely,
Thomas G. Lusch
cc: FAA Administrator Jane Garvey, AOPA, ATCA, EAA, NASA-ASRS, NATCA, and others...
Here's a copy of what I wrote on a NASA-ASRS report which I submitted just one month ago. This report dealt with a similar "invisible transponder" anomaly, wherein the Mode A (transponder code) and Mode C (altitude) data was not received by one facility (CMH's ASR-9 system), yet simultaneously the transponder worked just fine on a totally separate beacon radar (London, Ohio "QWO", Air Route Surveillance Radar). In ASRS fashion, I have omitted distinguishing information regarding the pilot's identity and the avionics shop identity, as I have not asked for their permission to identify them.
In this instance, I was working "local" (tower frequency) at CMH, and the traffic was light when the aircraft returned, so I had an opportunity to investigate if this aircraft's transponder was showing up on Indianapolis Center displays. This aircraft was at pattern altitude and on about a 3-mile left base for Runway 10R at CMH when I verified that it was "seen" okay by Indy Center (but not us)....
> 4/10/00 (evening): Nxxxxx (PA28) departed for XXX airport. I asked pilot to > recycle transponder as no beacon showing. Pilot said he was showing a reply > light. Entire way to XXX (xx nm) _never_ saw beacon reply. Pilot said he just > got transponder worked on @ Xxxxxxxxxx. > > About half hour later same airplane inbound fm XXX. Pilot says he is taking > aircraft back due to transponder problem. I decide to check other facility > (Indy Center) since pilot says he has reply light. Indy Center sees the > aircraft squawk code & Mode C altitude with no problem. This is from > antenna site ~ 30 miles away (aircraft is 3 miles from CMH). > > Supervisor was notified of this evening's problem.
Such a perplexing problem lead me to get in touch with the pilot and share my similar experience. Here is a portion of this pilot's reply shortly afterwards...
> Tom...I spoke with Xxxxxxx this afternoon, and they performed > extensive system checks (with everything installed using a test antenna as > well as bench tests -- AOK all the way)...I did send him the > info on the "invisible transponder" syndrome -- but only after they had > completed their retests and deemed the transponder fit...
Interestingly enough, the transponder in this aircraft was a KT76, exactly like the one I had problems with as described in my letter of 9/11/99 to the FAA Tech Center (see above).
Others may not think it significant if a transponder doesn't work correctly once in a while, but mind you, these are both cases wherein a transponder was seen by one beacon radar, yet NOT by the other (when it was assuredly in range of both). Equally significant is that directly before and directly after these experiences, an avionics shop had certified these transponders and found no problems with their operation.
During the summer of '95, when I was repeatedly experiencing this problem firsthand with my club aircraft, we observed a large enough number of other aircraft that were not showing up on our CMH beacon radar, yet were observed on other radars with no problem, that we documented some of these for our Airway Facility technicians. There was no single make/model of transponder that was having the problem. It appeared random. Then, sometime later, the problems disappeared. It wasn't until late in 1999, when I got a chance to read "A Field Study of Transponder Performance in General Aviation Aircraft," (click here to download a PDF copy of that study), that I learned more about what was apparently taking place. That is when I discovered that to address the "Terra problem," the Mode S interrogations had been modified. Here's a short quote from page vii (Key Findings)...
These transponders fail to reply consistently to interrogations used by the Mode Select (Mode S) radar system and by TCAS to acquire targets. This failure prevents the transponders from being detected by TCAS, and would be invisible to Mode S radar if the modification introduced to deal with the Terra transponders were removed from the Mode S processor. (emphasis added)
With the understanding that the Mode S interrogations changed at some point, that was the best (and only) explanation I have come up with as to why our KT76 transponder mysteriously began to work without problem (the shop never could "fix" it, as they never found a problem with it!) .
However, my experience from last month makes me wonder. If this is just the "P4 problem," why would this recent KT76 be invisible just to us, now that modifications have been made to our Mode S beacon interrogators? Is there something else at play here? Nevertheless, I can't help but ponder on the fact that if I have been able to see this specific anomaly occur firsthand, involving two separate KT76 transponders, how often may this actually occur without ATC (or the pilot) being aware of it? How many of those primary targets out there on my scope may actually have a "functioning" transponder???
This anomaly is especially relevant when it comes to aircraft squawking the VFR 1-2-0-0 code, and not in communication with ATC. The pilot would have every indication that their transponder is working correctly (i.e. recently certified, "REPLY" light flashing normally, other ATC facilities experiencing no problems, etc.), yet I as the controller may only have a primary (skin paint) target on my display. A primary target isn't nearly as conspicuous or helpful for a controller as a primary target that is enhanced by a transponder's beacon reply.
Due to the nature of this problem, it is difficult to diagnose, and very difficult to document. For instance, I attempted to document this most recent problem by requesting that radar data for a ten minute period from both CMH Tracon and Indianapolis Center be saved for analysis and comparison (i.e. it would be nice to have hard data to share with the Airway Facility technicians). However, I was met with resistance in my request. I was informed that such a request would have to be accomplished as a FOIA (Freedom of Information Act) request as described in the 7210.3R "Facility Operation and Administration" ORDER, Section 4-8, (see http://www.faa.gov/ATPubs/FAC/FAC0408.HTM) and that the cost would have to be borne by me (i.e. a portion of that cost is a charge for CPU use, which ranges from $99 to $275/hr as described in Para 4-8-4). Needless to say, I dropped the idea of obtaining this data.
The experiences above demonstrate yet another reason for the importance of redundant systems, and why it would be beneficial to be able to process all the radar data, and therein "fuse" the radar data from multiple radar sites (i.e. make use of the great redundancy in transponder replies that currently goes unused).
I have developed the following recommendations concerning anomalies like this...
Also, it would be well worth it for all of us to determine if the "P4 problem" that Keith Peshak has been harping on for the past couple years is a pervasive as he indicates. If so, the statement from "A Field Study of Transponder Performance in General Aviation Aircraft, wherein it says..."This failure prevents the transponders from being detected by TCAS..." is also of critical importance.
If you have had similar instances of unexplainable transponder problems, I'd be interested in hearing of your experiences.
Sincerely,
Tom Lusch
[Note: There is "transponder monitor" equipment that a pilot can purchase so as to monitor their own transponder output to verify if it is transmitting a signal, as well as monitor the data being sent. Information on this equipment can be found at http://www.airsport-corp.com/altitude.html. I have no commercial ties or interest in this company.]
© 2000-2002 by Lusch's Midair Collision Investigations. All rights reserved.