Scanned from a reprint of: IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS VOL. AES-16, NO.4 JULY 1980
The major radio aids to air navigation are listed. Underlined are those whose signal format is standardized by the International Civil Aviation Organization (ICAO) and they have now all been the subject of the IEEE Aerospace and Electronic Systems Society Pioneer Award, as follows: airborne direction finder/nondirectional beacon (ADF/NDB), Busignies and Moseley, 1959; VHF omnidirectional range (VOR), Stuart, 1962; instrument landing system (ILS), Kramar 1964, Alford 1965; air traffic control radar beacon system (ATCRBS), Williams, Bowden, and Harris, 1973; distance measuring equipment (DME), Dodington, 1980. A brief history of the development of the distance measuring equipment, which also formed the basis for an IEEE National Aerospace Electronics Conference luncheon address, is given.
Each year the Pioneer Awardee "earns" his award by giving a short talk at the NAECON luncheon, to give some of the background and personal anecdotes that help demonstrate that real live people are involved in development work.
Invited paper, presented prior to the conferring of the Pioneer Award to Mr. S. H. Dodington, 21 May 1980, at the National Aerospace Electronics Conference, Dayton, Ohio. Manuscript received April 19, 1980. Authorís address: International Telephone and Telegraph, 320 Park Avenue, New York, NY 10022.
Towards the end of World War II it was evident that a new series of radio navigation aids based on radar beacons, or secondary radar, was about to emerge. Such beacons had existed before the war, primarily for identification friend or foe (IFF) purposes, and had seen much use in the wartime British Rebecca-Eureka system . It must have occurred to many people that the addition of a relatively simple automatic tracking circuit would enable an interrogating aircraft to constantly read its distance with respect to a transporting beacon. However, it was apparently not until 1944 that such a device was actually built (by the Canadian Research Council), operating in the Rebecca-Eureka band of 200 mHz; and in November 1945 I witnessed a demonstration of an airborne distance measuring equipment (DME), built by the Combined Research Group at the U.S. Naval Research Laboratory, this time at 1000 mHz and, largely based on work that Hazeltine had been doing on the Mark V IFF, at 950-1150 mHz.
I had spent the war designing airborne transponders to receive enemy radar pulses and retransmit them at higher power with different pulse shapes for deception purposes. Hence, it was not too surprising that I was put in charge of the development of DME when International Telephone and Telegraph (ITT) made the corporate decision to enter that business as soon as the war was over. At ITT, I reported to Paul Adams who, in turn, worked for Henri Busignies (Pioneer Award Winner 1959), who reported to Maurice Deloraine. To these men DME was only the forerunner of a much grander scheme to provide a comprehensive navigation and traffic-control system, known as NAVAR, an air traffic control system proposed by ITT in 1945 in the 960-1215 mHz band. Similar schemes were being proposed by other companies, including Hazeltine, RCA, and Sperry. These schemes were all presented to the public at a U.S. Air Force symposium held at the Pentagon in February 1946. Nearly all companies received some contractual support as one result of this symposium. For ITT it was a U.S. Air Force contract in April 1946 for 25 DME airborne sets (AN/APN-34) and 5 DME ground beacons (AN/GPN-4). The cognizant Wright Field engineers were Nat Braverman and Carl Trout. NAVAR was to be added later. The principles of DME are shown in Fig. 1. An airborne pulse transmitter "interrogates" a ground receiver which, after a short delay (typically 50us) activates "replies" from a ground pulse transmitter. These replies are compared in the aircraft with the time of the corresponding interrogation, and the elapsed round-trip time, minus the fixed ground delay is converted to a reading of distance. The circuit making this comparison is called a ranging circuit. Large numbers of aircraft can simultaneously use the same ground equipment, since each aircraft ranging circuit only recognizes the replies to its own interrogations. With 100 aircraft per ground beacon, duty cycle per channel is only about 2 percent, and, from the very beginning of DME, there have been proposals to provide other services on the same channels. NAVAR was one such proposal, adding bearing and secondary radar response.
Crystal Control at 1000 mHz
Flight tests in August 1945 established to my satisfaction that a few kilowatts on the ground, with some antenna gain, and one kilowatt in the airplane, with a quarter-wave stub, would probably do the job. From the very beginning, I insisted that, since pulsecoding was later going to be used to differentiate the various NAVAR functions on a given channel, each channel be defined as a separate pair of radio frequencies, each frequency to be stabilized by quartz crystal control, to a small fraction of 1 mHz. This was violently opposed by almost everybody else in the industry. During 1946, Radio Technical Commission for Aeronautics (RTCA) SC-21 came up with a DME channeling plan which ignored crystal control. Its 13 interrogation channels and 13 reply channels were spaced 9.5-mHz apart and depended on free-running oscillators whose stability was to be 12 mHz. Col. Sam Mundell of Wright Field told me to ignore this and to just get on with the APN-34 which called for 51 two-way crystal-controlled channels spaced 2.4-mHz apart with each channel having a nominal stability of 200 kHz. Hazeltine, however, proceeded to build sets in accord with SC-21, receiving a contract in 1947 to build 9 airborne and 2 ground sets for the Civil Aeronautics Administration (CAA).
Making 1000-mHz airborne crystal control work was, however, much easier said than done. Aside from business trips and Christmas day, I spent some part of every day, seven days a week, at the ITT Nutley laboratory in 1946 and 1947. Making crystal controlled receivers and the fixed-frequency ground transmitters was relatively easy, but the one-kilowatt multichannel airborne transmitter was something else again. By mid-1946 the following system had jelled:
(1) Reception would be in the upper half of the 960-1215 mHz band, with a tunable preselector, and a crystal-controlled local oscillator.
(2) The local oscillator would be furnished by a broad-band multiplier working from a turret containing 51 crystals in the 42- to 47-mHz region. The IF would be 64 mHz.
(3) Transmission would be in the lower half of the band using a pulsed 2C39 triode oscillator, with motor-driven automatic frequency control (AFC) referenced to the crystal-multiplier chain. Since transmission occurred on approximately the image frequency of the receiver, the same IF served both for reception and for AFC. Transmission and reception were 126-mHz apart. Pulses were rectangular, of 1.5us duration, and adjacent-channel selectivity would have been intolerable but for the timely appearance of Hal Ferris of Trans-Canada Airlines. I had dinner with him during the Pentagon meeting in February 1946 during which he described a marvelous scheme whereby three diodes at the IF output were arranged so that almost any kind of spectrum could be rejected by the adjacent channel. I had it changed to work with two diodes, named it the Ferris Discriminator, and it has been part of the DME scene ever since.
For three weeks in October 1946 the CAA played host to the Provisional International Civil Aviation Organization (PICAO) at Indianapolis. Just about every navaids contractor was invited to participate. ITT was no exception; we were there with our DC-3 flying laboratory, airborne and ground DME, and some elements of NAVAR, the latter including an airborne azimuth indicator and a ground secondary radar display (this used S-band interrogation and L-band reply from the airborne DME transmitter). The radio interference between the various exhibitors was horrendous and of 41 demonstration flights in which I participated, only 25 were satisfactory.
One of these was planned for the French delegation on a Sunday, to reduce the chances of interference. We had scheduled the flight to be preceded by a champagne lunch at Mrs. Gammonís family restaurant on the north side of the airfield. However, we had forgotten that Indiana law precluded public drinking on Sundays, so we had to take the champagne upstairs to Mrs. Gammonís bedrooms, where a couple of cases were consumed. During the subsequent flight, most of the guests fell asleep, but nevertheless congratulated me afterwards on an excellent
The equipment at Indianapolis was far from deliverable. For one thing it only worked on a couple of channels. Not until April 1947 was a contractual delivery made to Wright Field of a preliminary set which only had to work on 6 adjacent channels, but that set was in almost constant transit back and forth between Dayton and Nutley for repairs and "fixes" during the summer of 1947.
A major result of the Indianapolis demonstrations, both by ourselves and by Hazeltine, was that ICAO decided that the future short-range navigation system would be rho-theta (as opposed to GEE, which was the British VHF hyperbolic navigation system, GEE standing grid, and that the DME part would be at 1000 mHz, not 200 mHz. (Australia decided to implement 200 mHz and still uses it.)
In October 1947 a series of demonstration flights was given in the Air Transport Associationís DC-3 named "Beta," piloted by Charles McAtee. We operated from Grumman, LaGuardia, and Washington National Airports, with ground beacons at Nutley, NJ and Allentown, PA. By the end of the year 5 airborne sets had been delivered to the U.S. Air Force. These each occupied a full air transport rack (ATR), a standard airline black box size, 19" by 10" by 7", weighed 55 Lb, and used 41 tubes, mostly of the so-called "ARINC-reliable" plug-in type (ARINC being Aeronautical Radio Inc.). These sets only worked on 6 adjacent channels. Construction took the form of plug-in subunits and the original specification had called for operation from 28 Vdc, necessitating a horrendous series/parallel arrangement of 6-V tube heaters. Fortunately, this was changed to 400 Hz ac before delivery.
The remaining 20 sets were to operate on the full 51 two-way channels and only weigh 35 pounds. We therefore reluctantly took the sideways step of abandoning plug-in subunits and building the set out of 3 large chassis, one each for rf, ranging circuit, and power supply. This set was described at the 1949 Institute of Radio Engineers (IRE) convention.
The principles of multichannel crystal-controlled DME were established by the end of 1947 and were reduced to practice during 1948. The CAA ordered 15 ground stations (DTA) and several airborne sets (DIA) similar to those built for the U.S. Air Force. The cognizant CAA engineers were Leo Wilbur and Sieg Poritzky.
We now come to a series of electro-political events that, taken together, were to delay the worldwide acceptance of DME by over 10 years. Many of them have been well described by Peter Sandretto (who joined FIT in 1946) and are far too complex to be repeated here. Instead, I will confine myself to the two main developments which, together, caused this delay: 1) The RTCA SC-40 (Gertrude Stein) channeling plan, 2) tactical air navigation (TACAN).
While ICAO had agreed that the future rho-theta system should be the VHF omnidirectional range (VOR) plus a 1000-mHz DME, the exact signal format of this DME had not been agreed upon. And while we at ITT felt we had the answer, there were many in the industry who felt that crystal control was impractical and that cheaper results could be had by the pulsecoding of unstabilized oscillators spaced 9.5-mHz apart. The leading exponent of this view was Hazeltine, backed by much Mark V IFF experience and some DME contractual work.
At the end of 1947 RTCA established SC-40; the bulk of 1948 was devoted to its deliberations, the protagonists being Knox Mcllwain, Bob Brunn, and Charlie Hirsch for Hazeltine, and Peter Sandretto, Paul Adams, and myself for ITT. Other members of the cast included Dick Borden of the CAA, and Bob Gilliland and Gene Jackson of Wright Field. The chairman was J. Wesley Leas of Air Transport Association (ATA). Knox Mcllwain took much ribbing over The fact that he wore spats and traveled to our meetings by train, rather than by airplane. Much of our argument against pulse coding was based on our wanting to reserve this for the separation of other (non-DME) functions on the same channels, as we had proposed for NAVAR and as had been recently endorsed by RTCA SC-31. (SC-31 had proposed that many existing and planned navigation and air traffic control aids be consolidated into one equipment operating in the 960-1215 mHz band. The report was widely acclaimed and was awarded the Collier Trophy for 1948.)
However, no one in SC-40 could define exactly what those other functions were, and we finally settled on a format which provided no other functions: a DME which was solely a DME, quoting Gertrude Steinís "a rose is a rose is a rose." It interrogated on 10 crystal-controlled channels at the low end of the band and replied on 10 crystal-controlled channels at the high end of the band, with channels spaced 2.5-mHz apart and with 10 pulse codes using multiples of 7us, starting with l4us. Thus there were 1000 combinations of frequency pairs and codes, of which it was proposed to pair 100 with the VOR/localizer. The CAA, moving with alacrity, adopted this plan as the official U.S. position, modified ITTís existing contract (DTA and DIA) to conform with it, got ICAO to accept it, and in March 1950 placed orders with Hazeltine for 450 ground stations. However, the enthusiasm for this system outside the CAA was almost nil, both within and without the U.S., and very little actual use was ever made of it. The main reason was that "everyone knew" that the U.S. military was developing an entirely different system in the same frequency band. When it was finally abandoned in 1956, there were only 340 airborne SC-40 type sets in existence.
During the summer of 1947, Sidney Pickles, who headed another laboratory at ITT (also reporting to Paul Adams), had been trying to interest John Loeb of the Navy in a new multilobe omnirange type of homing beacon for aircraft carriers, replacing the YE/YG. A contract was received in June 1948 and it was agreed that, while Pickles would build the antenna, I would supply the electronic hardware. While originally planned for the 1700-mHz band, it was hoped to add DME to it a year later (this was done) and it was therefore agreed by all concerned that it made sense to put it in the 960-1215 mHz band and use the same channeling scheme already developed for Wright Field. The system was later named TACAN, for Tactical Air Navigation. And so, by the end of 1948, the U.S. Air Force, the Navy, and the CAA had contracted for the ITT 51-channel crystal-controlled system.
As part of my DME work I had long been looking for a scheme for automatically maintaining beacon sensitivity at its optimum value. In October 1947 I came up with the concept of operating the beacon at constant duty cycle, raising its receiver "squitter" when there were no interrogations present. I also proposed that the resulting constant stream of output pulses could be amplitude modulated by a rotating directional antenna to provide bearing. When the TACAN contract was received, Peter Sandretto insisted that this principle be applied to it and it has been an integral feature of TACAN ever since.
While the basic TACAN patent is in the names of Dodington, Pickles, and Stavis, and John Loeb received the Navyís highest civilian award, I have always felt that Sandretto provided the key catalytic action at the critical time.
Because of the groundwork already laid on DME, the development of TACAN proceeded quite rapidly. In June 1950 we gave the first flight demonstrations involving distance and bearing. We also solved the problem of how to reduce channel spacing to 1 mHz. On 23 June Bob Gilliland at Wright Field demonstrated to me a 4-stage pencil-triode amplifier which produced an output of 1 kW with an input of 1 W, and 4 days later I had three 2C39 triodes operating in cascade with a gain of 51 dB. I decided immediately to abandon AFC and, instead, use direct crystal control in our airborne transmitters. In October 1950, the U.S. Air Force, the British, and the Canadians joined the TACAN team, and the TACAN specifications, as we know them today, were essentially fixed, with 126 two-way channels, spaced 1 mHz apart. Gaussian pulses were chosen for the ground beacon to reduce adjacent-channel splatter (thus reducing dependence on the Ferris discriminator), and the pulse pairs were spaced 12us apart to miss the 3us of the IFF system and the 14us of the SC-40 system. However, at the same time, the number of bearing lobes was increased from 3 to 9, adding further to the development delay. It was not until September 1952 that a complete equipment, to the new specifications, was test flown. Production deliveries commenced at the end of 1953.
With the award to Hazeltine of the 450 SC-40 style beacons, ITT took little further part in the "Gertrude Stein" system. The NAVAR system, along with many of its contemporary rivals, slowly faded away (though its main feature, azimuth by means of a rotating narrow beam, was subsequently adopted by the U.S.S.R. who, some 25 years later, cited it as one of their reasons for endorsing a time-referenced scanning beam for the microwave landing system (MLS)). Peak production of airborne TACAN equipments at one time reached 400 a month in the ITT Clifton factory, with others being produced by General Dynamics and Hoffman, for a total of 20,000 ordered before 1958, together with 600 ground beacons. While the TACAN system remained classified until the summer of 1955, its magnitude was generally known and, understandably, neither the airlines nor general-aviation had any desire to invest in the SC-40 system, which obviously was in conflict.
In retrospect it seems strange that two entities of the same government, one military and one civil, could proceed with parallel, conflicting developments for years. But that is exactly what happened. It has been well described in an admirably dispassionate book by Stuart Rochester. During that period I was frequently asked whether some kind of scheme could not be devised whereby TACAN-equipped aircraft could receive the CAA DME stations (and vice versa) but the answer was always the same: the two systems were so inherently incompatible that very little hardware commonality was cost effective.
Finally, in March 1954, a solution began to take shape. Donald Quarles, Assistant Secretary of the Air Force, appointed a committee to resolve the conflict. Most importantly, the committee was to be assisted by three advisors: Hector Skifter, President of Airborne Instrument Labs, Jerome Wiesner, later president of M.I.T., Russell Newhouse, Bell Labs (Pioneer Award Winner 1967). They were men of proven integrity, who had not been involved in the development of either system.
Two more years of acrimonious debate followed, much of which could have been avoided, in my opinion, if the pro-TACAN people had not attacked the VOR, but instead concerned themselves solely with the real point at issue, namely, the existence of two incompatible DME systems in the same frequency band. The VOR versus TACAN azimuth controversy was particularly pointless. VOR was a land-based system, already well established, which provided omni range service at a lower airborne cost than anything else, particularly if the aircraft were also equipped with an instrument landing system (ILS). It was, however, not suitable for ships or portable tactical use, and the military were therefore justified in proceeding with TACAN azimuth, saving money by combining it with DME. During this two-year period, 1954-1956, many pseudo-issues were raised. One was an attack on the Ferris Discriminator, which was more needed in TACAN with its 1-mhz spacing, than in the SC-40 system with its 2.5-mHz spacing. Fortunately, I had anticipated this and had made plans for an air-to-ground link based on spectrum control and conventional airborne selectivity. This was demonstrated to the three advisors on March 31, 1956. While the anti-Ferris critics were concerned with its effect on TACAN azimuth, spectrum control became an integral part of the ground-to-air link not only for TACAN but also for DME, unduly raising its cost. Until this point it had been assumed that airborne DME channel selection would be by a separate control head. The airlines now objected that this would allow blunders in which VOR was received from one station and DME from another. In July 1956 I therefore proposed a "hard" pairing plan whereby 100 TACAN channels were paired one-for-one with 100 VOR/ILS channels, allowing a single channel selector for VOR/ILS/DME. This was accepted and has been standard ever since.
On August 30, 1956, the Air Coordinating Committee handed down its decision: the SC-40 system would be scrapped and replaced by TACAN-compatible DME. Furthermore, enroute DME service would be provided by TACAN beacons co-located with VOR stations, this to be known as VORTAC. From the technical standpoint, an identical decision could have been made six years earlier, saving at least $10 million.
This decision has frequently been criticized as a political solution to a technical problem, with the taxpayer having to foot the bill for two separate azimuth systems. However, given the widely differing needs of the civil and the military users, I doubt that a much better solution could have been invented from scratch. It certainly has stood the test of time and popularity.
There remained the problem of ICAO. While few of its member countries had done anything about the SC-40 system, some of them now took the opportunity to drag out the old pro-hyperbolic arguments of 1946, but, this time, instead of GEE, it was DECCA, the British LF hyperbolic navigation system. More years of acrimonious debate ensued, but on February 27, 1959, ICAO adopted the TACAN-compatible DME. One of the issues raised during the latter part of the "Tacantroversy" (a word coined by Bill Carnes, then chairman of the Airlines Electronic Engineering Committee) was whether the 450 Hazeltine SC-40-type DME beacons could be converted to TACAN beacons. Contracts to attempt such conversions were awarded, but the general conclusion was that, while they could be made to work as TACAN-compatible DME beacons on a limited number of channels, they did not have enough average power to work as full TACAN beacons.
The CAA therefore awarded ITT contracts in 1957 and 1958 for several hundred TACAN beacons to be co-located with VORís, and these are still in service. While awaiting their delivery, the CAA (now succeeded by the FAA) also made do with military-type URN-3 and GRN-9 TACAN beacons. Today there are about 1000 DME beacons in the U.S. (mostly of the TACAN type) and an equal number in the rest of the world, many of them built by FACE-Standard, ITTís Italian associate. In 1979 the FAA awarded ITT Avionics Division a contract to update the electronics portion of all their VORTAC stations. A major feature is a 5-kW solid-state output stage. With the coming of the jet age, the airlines found DME to be a necessity, and in 1961 American Airlines was the first to have its entire jet fleet DME equipped, at that time with ITT equipment. By now the airline DME had been reduced to a half-ATR. Today, virtually every airliner in the world has DME. General aviation usage has been increasing, following initial development sponsored by the FAA, there being about 60,000 sets in the U.S. fleet of 250,000 aircraft. By 1970 moving parts were eliminated in airborne equipment, and by 1977 even the transmitter tube was replaced by solid-state electronics.
What of the future? The present ICAC "protection date" for VOR/DME is 1985. At a recent meeting of navigation users, it was strongly recommended that no existing system be scrapped until 15 years after its proposed replacement attains full operation, in order to provide an orderly transition. No such replacement system has yet appeared. I therefore believe DME will still be with us in the year 2000, not least because it will be part of the new MLS, which has barely started.
Furthermore, because of the relatively low dutycycle of the present ICAO DME, some of the added functions envisaged by SC-31 back in 1947 may yet come to fruition. Some examples that are already in the engineering development stage are Germanyís sector TACAN (SETAC) landing system and the U.S. Military Joint Tactical Information Distribution System (JTIDS). Increasing use is also being made of airborne area-navigation computers which, taking their inputs from the VOR and DME, allow an aircraft to fly a direct path from one point to another, whether or not this path is on a VOR radial. An interesting variation on this, taking advantage of todayís rapid channeling and range-search time, is an airborne DME which simultaneously tracks 4 or 5 ground beacons, providing a high degree of accuracy, coupled with redundancy.
A few words about the people involved in the development of DME. Some have already been mentioned. At ITT 130 employees attended a luncheon at Teterboro Airport on February 28, 1952, following a demonstration to employees who had worked on DME and TACAN. They cannot all be listed, but Etienne deFaymoreau certainly headed the list. In the U.S. government, besides John Loeb, Alexander Winick played a key role in the Navy, the Air Navigation Development Board, and the FAA. In the airline community, William Carnes, who headed the Airlines Electronic Engineering Committee for over 25 years, was outstanding. In general aviation, Gil Quinby, Vice-President of Narco, was a leading figure. At Wright Field it was Al Parker, and at the Naval Air Development Center (NADC), Johnsville, it was Oscar Shames: all dedicated and reasonable men who greatly aided the establishment of what has to be considered one of the most popular radio aids to navigation.
Scanned from a reprint of: IEEE TRANSACTIONS ON AEROSPACE AND ELECTRONIC SYSTEMS VOL. AES-16, NO.4 JULY 1980