B. Keith Peshak, Principle Engineer
Modern Avionic Development
Commercial single and multi-land, instrument airplane, BSEE & MSEE, faculty University of Florida and Texas & Embry Riddle Aeronautical University, 20 yrs avionics design. First GPS, First GPS-Loran, first GPS moving map, first GPS-ILS, first GNSS Cat. III c, first transponder proximity detector, first passive radar, first working and affordable collision avoidance system, first $10,000 ASR
_____________________________________________________________________________________________
This paper is "too hot" for the Air Traffic Control Association to publish:
Mr. Peshak -
After panel review and a review by our general counsel, we have decided
not to publish your paper entitled "What Is Wrong With ATC Transponders,
and How to Fix Them for Just About Free" in our 46th Annual Conference
Proceedings. We feel that your statements concerning FAA's technical
incompetence are not substantiated.
Your paper has technical merit, and we are willing to consider
publishing it in our "ATC Quarterly" or our "Journal of Air Traffic
Control." If you wish us to reconsider your paper, please edit
it to
alleviate our concerns and to make your case in a more academically
rigorous manner.
Please contact Ms. Matthews with further questions and to resubmit the
paper.
Sincerely,
Mike Perie
Our response:
I am sorry you feel that my statements concerning FAA's technical incompetence are not substantiated. I thought I had enclosed, via URL, sufficient proof. And, of course, the purpose of the paper is to itself prove beyond reasonable doubt that the TSO (technical standards order) for transponders is wrong. Simply put, if you build transponders the way that they are specified by the FAA, then they don't any of them work right. Even the FAA technical report admits to problems. If you read any transponder repair manual theory of operation section, you will see that it disagrees with the FAA TSO. Certainly one or the other is wrong, and all transponder manuals agree. And, of course, if you wrote that FAA official report, then you are in trouble. If you don't know that you need to violate the TSO, by doing the opposite of that specified, then you run the risk of your manufacture compliance putting you out of business by bankrupting you when the FAA orders you to make your product non-compliant to the TSO (remember Terra?).
I looked up the dictionary definition of incompetence: lacking the qualities necessary to effective independent action, inadequate to or unsuitable for a particular purpose expressed.
What is it, exactly, that you want me to change to appease whom you are worried about?
_____________________________________________________________________________________________
We now have the worst air traffic delays ever experienced, and the margin of safety is decreasing inversely with the lowered level of separation. Radar is operating too much in "coast mode" - the operators, and automated equipment, responsible for our only form of collision avoidance don't have a chance. The present system capacity is limited, and presently, it would seem, stretched to beyond that limit.
The ATCRBS transponder problems are entirely hardware specification error. The TSO, technical standards order, of the FAA, Federal Aviation Administration, is wrong. If you build aircraft transponders the way that they are specified by the FAA, then they will not work right. This can be fixed with a simple repair type modification to existing transponders, at the cost of a integrated circuit.
The magnitude and truth of the existence of this problem can be evaluated by studying the FAA official test report on ATCRBS transponder technology (http://www.gtwn.net/~keith.peshak/tn97_7.pdf) which many people refer to as "96% Of Transponders Don't Work Right". The efficacy of FAA upper level management in dealing with this problem is evident by reading the incongruous statement by the permanently employed top FAA official in charge (4th paragraph of http://www.gtwn.net/~keith.peshak/mcfaa.pdf).
Last year, we told of how we solved this problem, resulting in a single integrated circuit, which we sell at cost to transponder manufacturers. We also told of how we added a second, free, independent, fully autonomous, collision avoidance system, that we now know provides approximately 1,000,000 times the NAS capacity of ADS-B on VDL mode 4, the closest competitive technology. And we now know that a fully operable ATC ASR type radar can be produced from this technology for under $10,000 because we built and demonstrated that. This is mature, proven, tested, evaluated, ready to go capability. The only thing preventing us from sharing in the benefits of this technology is your FAA management (http://www.gtwn.net/~keith.peshak/taillight.htm).
We have shown to the Washington state and the Texas regional ACOs what the problems are with transponders not working right. We have shown it to the Atlantic City New Jersey technical center. We have shown it to the Washington DC FAA. We have sat down with the Administrator, and shown it to her. Since the FAA, at all technical levels, still cannot understand how TSO C74c (http://www.faa.gov/avr/air/AIR100/tsocur/Current.htm) is wrong, we need to publish at least two significant technical errors, in hopes that other engineers will come forward and explain this to the FAA management better than we in avionics, we in air traffic control, we pilots, are able to accomplish. This is the third proof of complete technical incompetence of FAA management and technical levels outside the ATC branch.
We here explain two major points of what we fixed in the erred TSO C74c. The FAA management has been unable to discover these two primary logical faults since 1973, and remains unable to understand them. The repair is to redesign the transponder by pulling out the digital logic in the transponder, and replacing with this IC in its place. You wouldn't believe how much better ASR radar sees the airplane, and how much further.
It is possible to independently research how FAA management damaged the formerly functioning transponder based ATC system capability. The transponder interrogation classically once consisted of three pulses, in time order, designated P1 then P2 then P3. If the transponder ever hears P2, the "side lobe suppression rule" was to thwart any reply from the transponder. This was done to prevent the "radar" from seeing the airplane, under that circumstance of that airplane positioned within the side lobe of the "radar" antenna (a wrong thing, if it were to reply from there). Since the P1-P2 spacing was always to be 2 microseconds, and since there would never be any other case of an interrogation containing two pulses spaced 2 microseconds, the rule oft implemented in transponder detection circuitry was "if you ever hear two pulses spaced 2 microseconds, thwart reply".
When C74b was replaced with C74c, two things happened. Only one thing should have happened, and it wasn't either one of the two that happened. One correct and critical thing was removed, two wrong things were done. We shall turn our attention to the following text:
"2.6 Decoding Performance c. Side-lobe Suppression. The transponder must be suppressed for a period of 35 ±10 microseconds following receipt of a pulse pair of proper spacing and suppression action must be capable of being reinitiated for the full duration within 2 microseconds after the end of any suppression period. The transponder must be suppressed with a 99 percent efficiency over a received signal amplitude range between 3 db above minimum triggering level and 50 db above that level and upon receipt of properly spaced interrogations when the received amplitude of P2 is equal to or in excess of the received amplitude of P1 and spaced 2.0 ±0.15 micro-second from P3."
The FAA decided that there should be a P4 following 2 microseconds after P3 for new radars and for airborne TCAS collision avoidance interrogator equipment. Remember, the transponder is supposed to wait 3 microseconds after P3 before taking action (starting the transmission of F1). There is no specification that prevents the transponder from listening, and processing, during that 3 microsecond dwell. Remember, the rule oft implemented in many transponder detection circuitry designs certified by the FAA was "if you ever hear two pulses spaced 2 microseconds, thwart reply". You see the problem (side lobe suppression, always, if you get the always transmitted P4 from new radars and TCAS received clearly). Not a problem for the older radars (didn't transmit a P4 in the interrogation).
The FAA modified C74b into C74c. Two modifications. The first is remove the requirement to suppress reply on P1 and a pulse following by 2 microseconds (P2 is the only pulse within 2 microseconds of P1, and is only there when you should suppress to a side lobe interrogation). The second was add the requirement to suppress reply on P3 and a pulse within 2 microseconds (P4 is the only pulse within 2 microseconds of P3, and is always there on the new radar and TCAS). And the FAA mislabeled P4, calling it P2 (P2 isn't ever anywhere near 2 microseconds from P3, P4 is the only pulse within 2 microseconds of P3 and is always there on new radars and TCAS).
You now should understand one critical logic error in C74c. Since your FAA management has received this explanation, at many levels, on many occasions, but cannot understand it, a more complete and unmistakable explanation is required for your FAA management. Let's parse the various legal interrogations to the transponder, including those containing a P0, and look at what the transponder should do, and contrast that with what the FAA TSO C74c says should happen.
This isn't the only problems with C74c, but they are all solved with the IC to repair ATCRBS transponders and add, free, AIS-P on mode S (if you just plug in a GPS to the transponder).
Remember, the remaining two rules for the civil aviation transponder are, "if you ever hear two pulses spaced 8 microseconds, reply squawk code" and "if you ever hear two pulses spaced 21 microseconds, reply altitude code" (once only if no P2, referring to P1-P3). Then there was another rule about "if you ever hear P0, a noise pulse, ignore it". This was done to prevent the loss of a transponder reply from interference caused by there being too many "radar" interrogators in the area, and to cover a combined 3A or 3C with a mode 2 interrogation.
There are, therefore, sixteen possible cases:
Old Original Radar:
1. P1
then
P2 @ 2us then
P3 @ 8us
This is a primary mode 3A interrogation with the P2 SLS pulse.
Suppress and do not reply
C74c does not require suppression, correct
operation in the NAS does.
2. P1
then
P3 @ 8us
This is a primary mode 3A interrogation with only the P1 and P3 pulses,
without the P2 SLS pulse.
Reply squawk
3. P1
then
P2 @ 2us then
P3 @ 21us
This is a primary mode 3C interrogation with the P2 SLS pulse.
Suppress and do not reply
C74c does not require suppression, correct
operation in the NAS does.
4. P1
then
P3 @ 21us
This is a primary mode 3C interrogation with only the P1 and P3 pulses,
without the P2 SLS pulse.
Reply altitude
New Improved Radar:
5. P1
then
P2 @ 2us then
P3 @ 8us then
P4 @ 10us
This is a primary mode 3A interrogation with the P2 SLS pulse.
Ignore the P4 pulse.
Suppress and do not reply
6. P1
then
P3 @ 8us then
P4 @ 10us
This is a primary mode 3A interrogation with only the P1 and P3 pulses,
without the P2 SLS pulse.
Ignore the P4 pulse.
Reply squawk
C74c requires suppression, correct operation
in the NAS requires reply
Many transponders suppress on any two
pulses 2 microseconds apart.
7. P1
then
P2 @ 2us then
P3 @ 21us then
P4 @ 23us
This is a primary mode 3C interrogation with the P2 SLS pulse.
Ignore the P4 pulse.
Suppress and do not reply
New Improved Radar and TCAS:
8. P1
then
P3 @ 21us then
P4 @ 23us
This is a primary mode 3C interrogation with only the P1 and P3 pulses,
without the P2 SLS pulse.
Ignore the P4 pulse.
Reply altitude
C74c requires suppression, correct operation
in the NAS requires reply.
Many transponders suppress on any two
pulses 2 microseconds apart.
Old Original Radar, with noise:
9. P0 @ -5us then
P1 @ 0us then
P2 @ 2us then
P3 @ 8us
This is a secondary mode 3A interrogation with the P2 SLS pulse.
The primary has no valid P3, so is not an interrogation.
Suppress and do not reply
C74c does not require suppression, correct
operation in the NAS does.
10. P0 @-5us then
P1 @ 0us then
P3 @ 8us
This is a secondary mode 3A interrogation with only the P1 and P3 pulses,
without the P2 SLS pulse.
The primary has no valid P3, so is not an interrogation.
Reply squawk
11. P0 @ -5us then
P1 @ 0us then
P2 @ 2us then
P3 @ 21us
This is a secondary mode 3C interrogation with the P2 SLS pulse.
The primary has no valid P3, so is not an interrogation.
Suppress and do not reply
C74c does not require suppression, correct
operation in the NAS does.
12. P0 @ -5us then
P1 @ 0us then
P3 @ 21us
This is a secondary mode 3C interrogation with only the P1 and P3 pulses,
without the P2 SLS pulse.
The primary has no valid P3, so is not an interrogation.
Reply altitude
New Improved Radar, with noise:
13. P0 @ -5us then
P1 @ 0us then
P2 @ 2us then
P3 @ 8us then
P4 @ 10us
This is a secondary mode 3A interrogation with the P2 SLS pulse.
Ignore the P4 pulse.
The primary has no valid P3, so is not an interrogation.
Suppress and do not reply
14. P0 @ -5us then
P1 @ 0us then
P3 @ 8us then
P4 @ 10us
This is a secondary mode 3A interrogation with only the P1 and P3 pulses,
without the P2 SLS pulse.
Ignore the P4 pulse.
The primary has no valid P3, so is not an interrogation.
Reply squawk
C74c requires suppression, correct operation
in the NAS requires reply.
Many transponders suppress on any two
pulses 2 microseconds apart.
15. P0 @ -5us then
P1 @ 0us then
P2 @ 2us then
P3 @ 21us then
P4 @ 23us
This is a secondary mode 3C interrogation with the P2 SLS pulse.
Ignore the P4 pulse.
The primary has no valid P3, so is not an interrogation.
Suppress and do not reply
New Improved Radar and TCAS, with noise:
16. P0 @ -5us then
P1 @ 0us then
P3 @ 21us then
P4 @ 23us
This is a secondary mode 3C interrogation with only the P1 and P3 pulses,
without the P2 SLS pulse.
Ignore the P4 pulse.
The primary has no valid P3, so is not an interrogation.
Reply altitude
C74c requires suppression, correct operation
in the NAS requires reply.
Many transponders suppress on any two
pulses 2 microseconds apart.
Do you see which cases of interrogation the new radars and TCAS fail
to work with if the transponder is TSO C74c compliant as implemented out
there?
1, 3, 6, 8, 9, 11, 14, 16 don't work right
Some transponders also do not have alternate detectors:
1, 3, 6, 8, 9, 10, 11, 12, 14, 16 don't work right
As long as both 2 and 4 work (even on the new ASR radars but only if they have the "Terra patch" properly working), at least you will have, although only very occasionally, P4 "radar contact". Don't abrupt these guys, including turn, or change altitudes, or pass anything near them, if you can at all help it. If you do, get used to they did, you just didn't see it (happy coast moding, don't look for the "C" on the data block). And remember (even on the old reliable radars) to watch for phantoms (unreal target echos at same range but different radials from the radar). Your ATC computer is now highly challenged to remove those from the planform display, but, when it fails, you need to make sure to service the real transponder reply, and ignore the P2 spooks.
You get to guess which transponders operate how, because you have no way of knowing. One thing you can count on is non uniformity of operation between manufacturers and marks of a manufacturer. Some are intelligent enough to ignore provisions of C74c, all swear they comply fully, not all get all violations correctly implemented. Putting a "TSO" on the data block immediately for each transponder that fails any patterns of radar returns would help you know. But, then, there would be so many of them; so your FAA management thinks that you would become upset if you were allowed to know.
We stabilize the present ATCRBS system technology, and we eliminate, at the same time, the need for biennials to "adjust" the transponder, to keep it working within the ATC system. Fear not, this wasn't hard - all it does, for these examples, is to return the SLS function to be keyed off of P2 within 2 microseconds of P1, while ignoring the other P2 within 2 microseconds of P3. It, simply, does the opposite of the specifications contained in TSO C74c. You wouldn't believe how much better ASR radar sees this transponder, and how much further, and shows all the abrupts that the airplane does.
The story doesn't end here, because there are additional faults in the ATC system hardware architecture. We remind you that we solved all of those last year, in this same IC that was used to correct for the FAA generated P2, P4 problem set.
We solve this problem set, completely, with a retrofit interrogation detection system on a chip (custom VLSI IC SoC), to replace the large number of small scale integration chips and analog parts in the existing aircraft transponder (http://www.gtwn.net/~keith.peshak/PartsIzLabor.jpg) with a fixed tuned, always compliant VLSI circuit "fixkit" (http://www.gtwn.net/~keith.peshak/aispkit.jpg). Like any avionics shop has always been authorized to do in the past, without any STC, remove that and replace with this. For free, you also put out the information to allow anybody else to get a picture in the cockpit of you (http://www.gtwn.net/~keith.peshak/T3.gif) or from the ground of you (http://www.gtwn.net/~keith.peshak/Range.gif), which solves the operators of the system have noticed deficiencies in system performance, to the extent that aircraft seen by radar are not showing on the radar planform display (http://home.columbus.rr.com/lusch/rtudslide00.html).
This may be impossible for your FAA management to understand, but that doesn't mean it is complicated, nor that it doesn't work.
