It has been observed by the FAA (DOT/FAA/CT-97/7) that a number of ATCRBS transponders fail to work to at least some degree.  While developing our transponder core, we made a remarkable discovery.
 
 

The Problem.

In short, the P4 pulse, which was added to the system to allow the more advanced Mode S, or selective, interrogation, causes a number of older transponder designs to not respond to the interrogation.
 
 

Why this is a problem.

Air traffic control radar is based on a transponder response, rather than imaging off of the radar reflection from the airplane skin.  When a transponder fails to respond, the plane is effectively invisible.  Either one of two things can happen.  Either the plane never appears on the radar screen, or, when radar contact is lost, air traffic control radar extrapolates the current position from the last two known positions (coast mode).   In either case, you can have airplanes in the sky whose location is unknown to the air traffic controllers.  And that can lead to mid-air collisions.

The problem is aggravated by the presence of the TCAS system on commercial aircraft.  Basically, each airplane can send out it's own interrogation.   All of these interrogations and responses can interfere with each other.  The 1030 MHz radio band becomes full of commands for transponders to "not respond".   The busier the airport, the more likely that a plane will not be seen.
 
 

Background - What is a P4 pulse, and why does it matter?

Transponders respond when they detect a series of pulses.  The original system had three pulses, named P1, P2, and P3.  P1 is the reference pulse.   P2 is a sidelobe suppression pulse.  And the position of P3 in relation to the P1 pulse indicated the type of response desired.

The radar system is designed to send out most of it's energy in a single direction.  This main beam contains the P1 and P3 pulses.  However, any directional radar beam will also send out weaker signals in other directions.  These are called sidelobes.  Even though these sidelobe signals are weaker, some airplanes will detect them and respond to them.  Since the antenna is not actually pointing toward these planes, their response implies that they are physically in front of the radar, when in fact they are off to a side.

To solve this problem, a second pulse (P2) is transmitted from a second, omni-directional antenna, and at a reduced amplitude.  The P2 pulse amplitude is set to be at least the same strength as the sidelobe beam, whose interrogation is to be suppressed.  The transponder receiver then compares the amplitude of the first pulse (P1) to the amplitude of the second pulse (P2).  If P2 is not below P1 by a certain amount, then the transponder is assumed to be listening to a sidelobe, and should not respond.

At a later time (1984?), a new system, called Mode S, was added to the transponder system.  Mode S provided an enhanced transponder interrogation.  To suppress a mode S response to an ATCRBS valid interrogation, a fourth pulse, P4, was added to the radar interrogation.  The P4 pulse was placed 2 us after the P3 pulse.  Note that the P3 - P4 separation is the same as the P1 - P2 separation.  That is the source of the problem. This P4 pulse presence causes a stoppage of the response from a non-mode S transponder.  This was not supposed to happen, when the mode S interrogation P4 pulse was envisioned.
 
 

Background - Where did the receiver design come from?

We have observed that many transponders have identical circuit designs.   Many even use the same component numbers.  Obviously, they were designed from the same master schematic.  While we have been unable to locate that master schematic, we believe that it was supplied to manufacturers at the time the original system was set up.

In this original design, there was no provision for the suppression of the detection of a P4 pulse.  It had not yet even been invented.
 
 

Background - Why do some transponders fail?

In the original design, the circuitry responsible for the suppression of the response when P1 = P2 is still active when the P3 - P4 combination occurs.  This is interpreted as a detection of a sidelobe, and the transceiver is turned off.
 
 

Background - Why do some old transponders pass?

The circuit used to detect all of these pulses is a combination of one-shots.  These one-shots have their timing values set through a combination of capacitors and adjustable resistors.  These components are subject to aging effects, and will tend to become detuned as time passes.  If the circuits detune enough, they will no longer work correctly.  As a result, they will misinterpret the P3-P4 timing, and, by failing, actually work correctly.

Since these circuits are adjustable, it is possible to set them to correctly respond when a P4 pulse is present.  However, this may produce a circuit that even though it responds correctly to a specific P4 timing, it might not respond to a second transmitter with slightly different timing, and might miss the P2 suppression.
 
 

The Solution

Instead of trying to modify an obsoleted design, we have designed a state machine that performs correctly in the presence of the P4 pulse, while maintaining the correct windows for the detection of the P2 and P3 pulses.  This state machine can be adapted for use in almost any brand of transponder, fixing the problem on new designs, and providing a path for the modification of old transponders.  Because transponder designs differ, our design is parameterized, so that it can be adjusted for each transponder transmitter design ringing duration, and each transponder receiver design phase delay.
 
 

FAA Test Results on Tansponders

Here is a sumary of the data collected by the FAA in a random sampling of general aviation aircraft trasnsponders.

A Field Study of Transponder Performance in General Aviation Aircraft

DOT/FAA/CT-97/&

Federal Aviation Administration

William J. Hughes Technical Center

Atlantic City International Airport

N.J. 08405

Page 12

Comments:

1 The problem is based in the receiver or transmitter or external equipment altitude encoder.

2 Our design controls this problem completely - without adjustments to fiddle with.

3 Minimum pulse width discrimination provided in receiver, Maximum commonly not, we could add that.

There is an interesting article, written by Vernon Barr in the August 1994 issue of US Aviator, which details the problems that the Terra transponder models 250 and 250D had with this problem.

There is another interesting paper published at the international meeting of the Institute of Navigation, Albuquerque, June 2002.  A complete peer reviewed paper covering the issue of FAA TSO Technical Standards Order C74c is wrong.