Oral-History:Lee Davenport
About Lee L. Davenport
Davenport got his BA in Physics from Union College in 1937, his MA in Physics from the University of Pittsburgh in 1941, and his PhD in Physics from the University of Pittsburgh after the war. He was recruited for the Rad Lab in 1941. He spent his entire time there working on anti-aircraft radar, eventually as a project manager for the SCR-584. In addition to the basic development of the SCR-584, he was also involved in getting equipment to the field quickly, training troops how to use it, and in adapting small numbers of SCR-584s for particular uses; e.g., waterproofing them for D-Day, adapting them to guide bombers for close support of infantry, using them against buzz bombs, and adapting them for mortar location.
Towards the end of the war he helped develop the first radar-guided missile, put a patent on it, and used it for his dissertation thesis; this wasthe basis for much post-war missile work. After war helped build cyclotron at Harvard with Curry Street and Ken Bainbridge; then went to work for industry as an executive at Perkin-Elmer, Sylvania-Corning Nuclear Corporation, and General Telephone and Electronics Laboratories; ending up as Chief Scientist for GTE. During war had run-in with Luis Alvarez; Alvarez blaming Davenport for not getting XT-1 to work for Alvarez’ blind-landing idea, when the equipment just was not suitable.
About the Interview
LEE L. DAVENPORT: An Interview Conducted by John Bryant, IEEE History Center, 12 June 1991
Interview # 089 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc.
Copyright Statement
This manuscript is being made available for research purposes only. All literary rights in the manuscript, including the right to publish, are reserved to the IEEE History Center. No part of the manuscript may be quoted for publication without the written permission of the Director of IEEE History Center.
Request for permission to quote for publication should be addressed to the IEEE History Center Oral History Program, IEEE History Center, 445 Hoes Lane, Piscataway, NJ 08854 USA or ieee-history@ieee.org. It should include identification of the specific passages to be quoted, anticipated use of the passages, and identification of the user.
It is recommended that this oral history be cited as follows:
Lee L. Davenport, an oral history conducted in 1991 by John Bryant, IEEE History Center, Piscataway, NJ, USA.
Interview
Interview: Lee L. Bryant
Interviewer: John Bryant
Date: 12 June 1991
Location: Boston, Massachusetts
Education and Early Career
Bryant:
I'm John H. Bryant. This is Wednesday morning, June the 12th, interviewing Dr. Lee L. Davenport, as part of an Oral History Project for IEEE Center for History and the History Committee. Dr. Davenport, could you please give us some background regarding your family and why you chose to become a physicist?
- Audio File
- MP3 Audio
(089 - davenport - clip 1.mp3)
Davenport:
This is a tall order, and I'm sure everyone of us who eventually ended up in this technology area comes from a different background. Mine started in the City of Schenectady, New York — home of General Electric Company. It was fortuitous that Union College was located there too. I was born and brought up there and went to the Schenectady schools. My father was a high school mathematics teacher, head of the math department and assistant principal of the Schenectady High School. In that environment, interest in science began for me at a very early age. I was an avid reader of Popular Mechanics and made little electric motors out of paper clips and copper wire and 1-1/2 volt batteries. I made a crystal radio set and I got involved with all sorts of technical things and enjoyed it very much. My school history was uneventful except for the fact that I did really concentrate on all the technical subjects that were available to be studied. In grade school and through high school I majored in the science and technical side: chemistry, physics, general science, and the mathematics courses. It was clear to me by the time I was ready to leave high school that I wanted to be a physicist; I was going to study physics in college.
Since my graduation from high school was in 1933 and the Depression was well upon us, funding for a college education was a very important factor. I applied and fortunately was accepted by three colleges, close to Schenectady, where I could go, hopefully, with minimal expense. Union, RPI, and Rochester all offered me scholarships, and I accepted the one at Union. Although it wasn't the best paying of the three, I was able to live at home. I enrolled for studies as a physics major. There were but two of us who majored in physics out of that class that was to graduate in 1937. During college I was lucky enough to get employment through the Federal Youth Administration Program — FYA, which paid 35 cents an hour for work that was done in colleges by deserving students. So during that time I was able to help the chairman of the department, Professor Peter Wold, do a revision of Kimball's College Physics. He was revising the textbook and he needed someone to make drawings and illustrations for the new volume that he was producing. So I was the principal illustrator for that volume at 35 cents an hour from the Federal Youth Administration. I also picked bugs off of plants and did things of that sort during the school year, under the same Federal program.
I was very lucky at the end of my freshman year to get a summer job at General Electric. Most of us needed summer jobs. I started off the summer by cutting asparagus in a truck garden — a very unsatisfactory job — for a budding physicist. But I got a call from GE, and I went to work in their Payroll Department that summer. Thirteen dollars and 50 cents a week, as I recall it, as a temporary employee of General Electric. I worked on balancing the pension accounts for employees in the pension plan. Do you want a surprising little story in here? Nothing to do with physics, but I'll tell it to you just for fun. I was interviewed and hired on the basis that I had completed my freshman year at Union and that I was adept at mathematics. Of course I said, yes. I'd had calculus and trigonometry and descriptive geometry and so on, and I'm pretty good at mathematics. So they sent me down to the Payroll Department. I got there, and the paymaster said, "Now with your mathematical background, it's my understanding that you could probably help us out in this area where we have a special need in balancing this pension plan. Over there you will see a big long table with eight young ladies sitting there, punching Comptometers.
Using those calculating machines, they're calculating interest at 2-1/2 percent semiannually on the sums of money for each of the 27,000 employees here in the Schenectady Works." He said, "Now there's a second table over there, so we have 16 girls calculating the interest and entering the figures on cards." He said, "It has occurred to us that calculating interest at 2-1/2 percent is the equivalent to dividing by 4. And it's certainly something that you as a mathematician should be able to do in your head. We'd like it if you'd use that method and see if you could check the work that these girls are doing." So I started with my first stack of cards, and I looked at the first number, say $184.19 — and divided it mentally by 4. It took a little time, but at the end of the first day, I could do cards, dividing by 4 by inspection in a matter of maybe 10 seconds per card. By the end of the first week, I was down to about 3 seconds per card. It wasn't long before I could go faster than all 16 girls in checking their numbers. By the end of the summer, when we were ready to balance this whole affair, I could not look at a telephone number, a license plate or anything else without dividing it by 4. For many years after, any large number that I looked at, would automatically come up divided by 4. The end of that story is that I was rehired for the next three years and given the assignment to take charge of the 16 temporary girls that they would hire, balance the pension plan and the Mutual Benefit Association plan. I did both of those and got all the cards entered for all the employees before I went back to work at Union in the fall. The girls worked nicely for me, even though we were all temporary employees.
At the end of that last year, I was interviewed by General Electric for a permanent job. They knew I was ready to graduate from Union, and I was called into the offices of Mr. M.M. Boring. Mr. Boring was the top personnel officer of the General Electric Company at its headquarters office in Schenectady. Mr. Boring said, "You have developed a good record working with us over the years, and we would like to suggest that you consider coming into our engineering test program. We pay $35 a week for any young engineers." "But," I said, "I'm not an engineer. I'm really a physicist, Mr. Boring." He said, "Well, we've also considered the fact that perhaps with your background you might like to go onto graduate school." I said, "I intend to. It's my plan." I told him that I had been accepted and I had saved enough money from General Electric to go on and get a master's degree in physics at Princeton. He said, "I don't think we would recommend that. There really isn't much application for people with physics degrees at General Electric. We hire lots of electrical engineers, and we would recommend that you go to Iowa State or Ohio State — either of those two — and get a master's degree in electrical engineering."
I said, "But I'm really not interested in electrical engineering." This seemed to roll off Mr. Boring's back. It didn't occur to him that there could be anyone who would want to continue in physics when, according to him, the only role for physicists was to teach other physicists, and that they wouldn't have a place in General Electric for someone with my kind of training if I insisted on going to Princeton. While preparing to go to Princeton for my master's degree in early September around Labor Day, I got a call from the head of the Union Physics Department, Professor Wold, who said that he had received word that the University of Pittsburgh had a last minute opening for a graduate assistant in physics. He'd looked up the department; he knew Dr. Worthing, who was in that department, and Professor Hutchinson who was then the editor of the Journal of Applied Physics, and was Chairman of the department. He thought that I should seriously consider accepting that assistantship. I did as he suggested. I called Princeton, talked to them about it, asked if I were accepted at Pitt whether I could cancel out of Princeton. The answer was, yes. They were very gracious about it.
As a result of all of that, I trekked off to Pittsburgh for my graduate assistantship and hoped that I could go on for a Ph.D. at Pitt. Fifty dollars a month was the fee for a graduate student assistantship at that time. My job was to live on that in Pittsburgh, which was not terribly difficult. I began work there, developing a strong interest in ultraviolet spectroscopy to start with. For extra money I also worked for Mellon Institute doing spectroscopic studies for some of the projects there. I got a master's degree and was ready for a Ph.D., when I decided to concentrate on x-ray diffraction studies instead of the UV I had worked on earlier. As I was coming to the end of that doctorate study, expecting to get my Ph.D. in June 1941 the conference at MIT on Applied Nuclear Research under the auspices of the American Institute of Physics caught my attention.
Well, I attended that conference. My attention was drawn to it, partly because of my interest in x-ray diffraction studies. I also knew Alexander Allen very well, a professor at the University of Pittsburgh, and Alex Allen was working on nuclear studies and had encouraged me to take an interest working with him on some hydrogen lamp sources and other nuclear things. So in attending that meeting, I was quite surprised to find out a number of the principals disappeared from the meeting sessions, of which Alexander Allen was one. At the conclusion of the meeting, he dropped by to inform me that he was going to be staying on for a few extra days in Cambridge and probably was going to move up for a short period to do studies up there at MIT. That surprised me a bit. He said he thought there might be a place for me up with him at MIT, but I felt reluctant because I was just about ready for my comprehensive examinations for the doctorate that I was hoping to get in June. I hadn't finished the thesis, but I was well along.
Recruitment to Rad Lab
Davenport:
To make a long story a little shorter, I returned to Pitt and in a matter of a week or two got a phone call from Alex Allen that said, "Yes, it's an important project. If you can come up here, we'd appreciate it." I said, "I prefer not to come right away. I'd like to get those exams out of the way." He said, "That might not be convenient for us, but we'll consider it, and you'll hear more." Well, when I heard more it came from Wheeler Loomis who called and said he thought I should show up rather promptly, that my draft board would look favorably upon it if I came to MIT. I finally was able to work out an arrangement under which I could finish those exams in January and agreed to show up early in February of 1941.
"Roof System"
When I came to the Lab, the first exposure I had other than signing up the necessary forms and getting identification pictures and things of that sort, was a meeting taking place under Louis Ridenour's direction. The subject was Project 2. Project 2 was just starting, and Project 1 had already been put in motion. It appeared to me this project was the most complex system that one could ever dream of for knocking an airplane out of the sky. It required a combination of radios and computers and guns and all sorts of paraphernalia. As far as I was concerned, you ought to be able to aim a gun at an airplane without a big computer and shoot it out of the sky. Little did I know. It wasn't long before I was heavily involved with this program, working for Louis Ridenour. After getting some understanding of what it was all about in the first few days my principal task was to demonstrate that a scanning system of the type we were planning to use would actually work. The system was to rotate the radar beam around an axis through the target in such a fashion that there would be equal signals from left to right and from up to down, a conical scan.
So I was expected to get a demonstration system together that would determine this. I acquired a 36-inch parabola with a dipole antenna and ran patterns. Then adding a motor drive, made the first scanner by physically tilting the parabola axis from the spin axis around the dipole as a center. Dynamic balancing was a problem to be solved since the whole parabola rotated and therefore wobbled. In other words the first concept to generate a cancel beam scan was to tilt the parabola axis away from the spin axis so both axes would cross one another at the focal point. This approach created a very wobbly rotating parabola since its center was off the axis of rotation. Obviously it would need dynamic balancing. To complicate matters further it would be necessary to change the angle of tilt in order to optimize the error signal and not sacrifice too much range sensitivity. It might even have to be changed while the parabola was spinning. Now that was a real challenge. It was typical in those start-up days that we had some clever ideas and some weird ones too. Here's one of the latter. We shared this variable dynamic balancing problem with one of the MIT professors of Electrical Engineering. He said "You know there's a washing machine somewhere that is dynamically balanced when it spins. I'll look up how it works and, if necessary, we'll buy a machine and take the mechanism out of its." Well fortunately we didn't have to go that far. A better and simpler concept saved the day.
We made all our first tests up on the roof to see whether we could obtain adequate correction signals. In the meantime, Ivan Getting was working on obtaining a servo controlled gun-mount from General Electric on which we planned to mount the antenna. I rather soon found that I was the principal gatherer of pieces, assembler and operator of what became the Roof System, the concepts for which had been pretty much put in place before I came to the Laboratory. The conical scan idea and many other concepts were already there. My contributions were far more in proving it out. Getting the pieces put together, getting the system on the roof operational. I developed into a systems manager which probably fit my skills a lot better than trying to understand the theoretical aspects of microwave transmission lines. That led me into the eventual assignment of being the project manager for a series of versions that finally became SCR-584. So I was with the same project for the entire Laboratory period. Does that give you enough introductory background?
Bryant:
That's excellent. Was that first called the XT-1?
Davenport:
Well, the first version was called the "roof system." The roof system was put together on the roof of Building 6, using the conical scan concept and using the General Electric designed aircraft turret for machine guns which had been acquired. XT-1 was the second version.
Bryant:
Was the roof system servo controlled?
Davenport:
It was Servo controlled using 400 cycle aircraft Amplidynes made by General Electric.
The machine gun turret was being designed and built for use in a bomber where it could be aimed by remote control. The servo system operated from DC and 400 cycle aircraft power and signals for the Amplidyne drive were derived from 400 cycle synchros. We employed the synchro driven handsight to aim our parabola at a target. When we were ready to aim at a real target it was clear that we needed something up in the air away from ground clutter. Preferably a stationary target we could use for adjusting things. The few aircraft then flying were never around long enough for us to lock on. Ivan Getting came up with the idea that we might use a tethered meteorological balloon made reflective by painting it with Aquadag and flown from the roof.
Bryant:
Was that colloidal carbon?
Davenport:
Colloidal carbon, right. The carbon, we thought would give us enough reflection. So we went ahead and we got the meteorological balloons, we got the Aquadag, and I got dirty. It wasn't too easy to paint a balloon with Aquadag, and the reflectivity of carbon was not really the greatest.
Bryant:
It probably absorbed more than it reflected:
Davenport:
That's right. It certainly absorbed more than we expected. So I recall we went on from there by having a silk cover with foil sewed for the next balloon. Dr. Getting's wife sewed it on her sewing machine. We went so far as to buy a fishing pole and reel for deep-sea fishing, and heavy fish line, to launch and retrieve the balloon. I thought to myself at the time: some purchasing agent is going to wince when he sees an order for a fish pole and fishing line come across his desk. What will the auditors think of our sport fishing equipment that we're using on this secret project? But things like that went on all the time. I can't recall whether that idea helped much since the rage was so short and receiver recovery was not very fast either. It really didn't matter because we quickly concluded we had to have an airplane under our control from the ground.
Range Tracker Testing
Davenport:
You probably have heard the story of how we got our first airplane target. Ivan, who by then had taken a more important role in Project 2 had been at Harvard. He knew a geologist who was a Fellow at Harvard, whose name was David Griggs. Griggs had his own airplane. Dave's Luscomb, an all-metal two place light airplane, might be available if we could recruit him into doing this job. Dave agreed that we could use him as a target without knowing what we were doing. He was eventually cleared to a point where he was able to provide extensive services to this secret project and later a staff member. It was important, however, to have some communications with the plane. The only sort of portable communications one could purchase was a battery operated transceiver systems. A transceiver was about as big as a typewriter — and not a little portable typewriter, either. This transceiver system had to be put on the airplane. Well, Dave couldn't fly his airplane and run the transceiver all at the same time. It therefore turned out that somebody had to ride the airplane with David, and I held the transceiver in my lap. We strapped the dipole antenna onto the landing struts of the airplane. So our first tests on a real airplane came when we flew that plane over MIT at an altitude of about a thousand feet. That was illegal at the time, too low an altitude over the city. But we did it, and found that the roof system could lock on and track.
Once we'd gotten to that point, the Signal Corps — which had expressed an interest in this whole project were invited to come up and take a look at the system. It was Ivan Getting's responsibility, as the principal liaison for the project with the military, to keep them informed. Their arrival at the Laboratory, probably May '41, was a cause for a very considerable amount of preparation, as you can imagine. That first demonstration for the military was a Laboratory event. The target plane flew over, the tracking was pretty good — a little jittery, a little shaky — but there wasn't any question that the roof system was doing its job automatically. Many elements of this first system had been built by other groups in the Laboratory. For range measurement and gating the circular sweep J scopes worked beautifully. We didn't have any plan position indicator (ppi) for searching purposes. We got on the first targets by using the handsight telescope on a tripod. So, although the system was put onto the airplane by an optical tracker once it got there and it got a signal, you could lock it on.
Bryant:
Did you have range tracking at the same time?
Davenport:
We had range aided tracking at the same time, not automatic.
Bryant:
Was that pioneering work in range tracking?
Davenport:
To the best of my knowledge, it was the first high accuracy range tracker.
Bryant:
So having range and angle tracking by the middle of '41 was quite an achievement.
Davenport:
Yes, particularly the automatic angle tracking made possible by narrow beam microwaves. Remember, range tracking was not automatic in the same sense that angle tracking was automatic. Angle tracking was a complete close-loop servo system. In range tracking, we employed a physical pointer over the range signal as displayed on a greatly magnified last sweep. As the signal moved with the changing range the motor-driven pointer also moved. That's fine so long as both rates are the same and constant. But of course range rates didn't stay constant. They increased or decreased. So the system that we used, which is referred to as aided-tracking involved a single control wheel which when turned physically moved the pointer ahead if it was falling behind and increased the rate of speed of the motor too. So it allowed an operator with a single control to do much more accurate tracking. That aided tracking system for range worked out well. The reaction of the Signal Corps, as I recall it, to that first meeting was: "This is great! This is a good idea! You fellows seemed to have made it work; but it's no good to us sitting here on a roof in Cambridge. We've got to have a chance to test the radar down at our own laboratory."
Bryant:
Was this Fort Monmouth?
Davenport:
This was Fort Monmouth. They wanted us to move it there. "Can you pick this system up and move it to Fort Monmouth? And we can give it a test down there." "Well, how would you test it?" "One of the ways we'd like to test it is we'd like to drive a searchlight with it. Get a plane up there at night, lock this thing on the airplane, and let the plane fly around. We'll see whether we can follow it with a searchlight. If we can, then that's the first step in shooting down an airplane. Getting a light on it. Then we can also track it with our standard systems optically and compare results. We agreed, generally, to do that. At that point in time those agreements were reached with the Signal Corps above my level. I was not involved in that discussion. I do know the relationships were excellent. My contacts with all Signal Corps people were most friendly. I was a little surprised in one sense. Even in those early years, we had been exposed to what we refer to today as the "NIH problem," the "not invented here" problem. The Signal Corps had its own separate development programs. They had the SCR-268 already running. They were talking about the SCR-541, which to the best of my knowledge, even at that time had been put on the docket as a requirement. But they did quickly accept our project as a concept that was worth a further test.
XT-1
As a result of that, it became clear we had to have a mobile system which we called XT-1. (for experimental truck number one.) So Ivan Getting and I started out to find a truck on which we could mount a unit. Can you imagine a couple of 20-, 30-year-old, MIT types, coming across the river to the White Truck Corporation's showrooms on Commonwealth Avenue in Boston with their yardsticks, their measuring equipment, and their calipers to determine the thickness and rigidity of the frame of a White truck as compared to the frame rigidity of a Mack truck? We were like a couple of kids walking in here, since they usually sold their trucks to construction people, not MIT types. We decided on a White truck, and then we had to find a custom body-builder. We designed a rigid moving van body, and we got Lacey Bodies up in Medford, Massachusetts, to build it. We designed the elevator system to raise the parabola through the roof. Thus began the project of putting a system into a truck to drive to Fort Monmouth for its test. I look back on the procedures that were followed in putting that together as being somewhat hair-raising.
Davenport:
That gives me a chance to give you a little feel for the climate of the Laboratory at that time. There was a spirit within the organization that I look back on as being unique for its day — and unique for today as well. First, I think everyone was under the impression that we were going to get into the war, and that this fact was damned important. Next, what we were trying to do was equally important. The Battle of Britain and other events pointed to the importance of our role as a laboratory in microwave radar developments. We considered our mission to be an all-consuming mission and not to be deferred or deflected for any purpose; and early results were critical. We had no compunction whatsoever about bending the rules a little bit if necessary to get something done. We would buy a washing machine if it would bring a quicker result. Second, we were all under the impression that this was not a really long-term project. We were in a rush. Overtime was expected.
We had to get this work done but I for one was expecting to go back and finish my Ph.D. in six months — certainly by a year we could finish. So we had a feeling — partly generated because of the secrecy that we were expected to hurry and we were not to be put aside for anything. That permeated the building of XT-1. The first experimental truck was a project that we moved heaven and earth to get done quickly. We had no reluctance whatsoever about going to an outfit that could spin a 6-foot aluminum parabola and telling them this took priority over anything else on which they were working and that it was being done under the auspices of MIT, who had an important contract with the federal government. We couldn't tell them anything about what it was for except that it was important. And people listened. They listened to us, even though we didn't appear to have the trappings of authority. As scientists we were personally involved in selecting the things we needed and the contractors who were going to do the work rather than having a purchasing agent do it. So our involvement was total, as it were, from design to finished model. We worked as a team without any problems at all. Everyone was cooperative. The sense of urgency was communicated throughout the whole Laboratory, and things moved fast. No question about it, red tape was minimal. If we needed added help, the most serious problem was getting people who could be cleared and moved in quickly. Money seemed to be available from the level at which I operated. I didn't have administrative or financial problems if any existed for the laboratory management. We were encouraged to move right ahead. So putting all of this together was a real challenge, a lot of fun, and called for imaginative ideas.
Imaginative ideas? I'll give you one. We had to design an elevator to raise the parabola from its stowed position inside the moving van body and raise it to the roof level. That accomplished, we next had to do some tracking in the Boston area to check performance. We couldn't allow the public to see what was happening up there on the top of this truck if we tested it at Logan airport where the planes were. So we built a collapsible baby buggy cover, that would conceal the whole antenna mount. It had wooden bows and a light canvas cover and we could track through it — a collapsible radome. Looking back at it now, it didn't seem to be startling at all that we would approach it that way. It was quick, it was easy, it was something we could do readily. It may not have met military specs, but it would get the job done. The tests that were run with XT-1, worked out very well in Boston. I'd say we made one unnecessary error. We assumed correctly, that a mobile system would have to supply its own electrical power since it would need to be operated in the field. Our solution was to attempt to employ the truck motor to run a large generator rather than to buy a gasoline powered motor generator set that would require a separate trailer or another truck. We designed and installed an under floor mounted generator, driven from the truck drive shaft through a large clutch and multiple vee belts. It was a disaster. But except for that, almost everything could be made to work. Now XT-1 was much more than a breadboard by the time it was ready to go to Ft. Monmouth.
Bryant:
Would you say you duplicated what was on the roof?
Davenport:
Essentially, we duplicated what was on the roof rather than physically remove the roof system. Since the roof system was still needed for testing improvements XT-1 was an effort to field a system that could be moved for military testing.
Bryant:
Did the Laboratory have shops at that point that could do quite a lot of it?
Davenport:
Yes, they had some shops, but certainly limited. We used outside machine shops for some pieces of this system. However, the electronic chassis in the radar part was built by the Lab component groups. I recall the servo electronics were made by Sid Godet who was a GE employee assigned to us. By the way, we painted the truck red and gray, MIT colors. It looked like a moving van with no identification on it.
When the time came to move the equipment to Fort Monmouth we were furnished with an armed military guard to ride in the truck. A Laboratory beach wagon completed the convoy with the balance of our team aboard. Now it's a long run from Boston to the Jersey shore and we had to make good time in order to do it in a day. Old Route 1 went through every town, but there was a time saving route for a good part of the way. It was the newly opened Merritt Parkway, a toll road, but it was open only to passenger cars. Now this was in November of 1941 before war had been declared and our truck certainly didn't look like a military vehicle. Nevertheless, at the first toll station, we talked our way onto the Parkway with the help of the military armed guard and liberal use of the words "secret" and "emergency." We may even have mentioned "camouflage."
Our first appearance at Fort Monmouth resulted in general failure. We set up the equipment and we hooked up to drive the searchlight via special 60 cycle synchros which we had added to the turret to transmit the azimuth and altitude data. There was no ranging necessary for the searchlight test, of course. We aligned the searchlight pointing axis with the radar axis. We scanned around; the searchlight followed. After dark a target airplane flew over so we could lock on. After re-aligning the search we started to track. The searchlight soon fell behind the airplane. So we quickly readjusted, and, although XT-1 was tracking, the searchlight fell behind again. The searchlight would not stay on the airplane. This left us very disturbed because at that moment, we could not determine what could have gone wrong. It didn't take us very long to figure it out though. Searchlight pointing was to be controlled by high and low speed 60 cycle synchros which we had installed on our gun turret antenna mount. We had the wrong high-to-low gear ratios on these drive units. It is my recollection that this happened because we had incorrectly assumed that the ratios for searchlight drives would be the same as the ratios for driving the pointers on the Army anti-aircraft gun directors. Our eventual goal, of course, was to feed data using these same synchros to the M-7 gun director which in turn drove the guns, so it was a natural mistake.
We brought the whole system back up to MIT and installed new gears on it. We changed the parabola to a larger size at the same time. Then we went back for a second show at Fort Monmouth, at which time the performance was completely satisfactory. It may be interesting to note that on December 7th, a Sunday off for us at Fort Monmouth, during these tests, we all decided to take the company station wagon and drive up to New York City. It was not too far away. So we went to the movies at Radio City Music Hall, which, of course, was the movie house in New York at the time. I obviously don't remember the picture that we saw, but when we came out, we went down to the lower level of the Music Hall on which the restrooms and a big lounge were located. There on the lower level were some clattering teletype machines with news reports. There seemed to be quite a crowd gathered around the machines and we curiously asked why. We quickly learned that there were reports coming across the wire that the Japanese had bombed Pearl Harbor. That meant to us we might be needed back at Fort Monmouth in a hurry. If the Japanese had bombed Pearl Harbor, everyone felt submarine attacks against a place like Fort Monmouth might occur at any time, and we were the proprietors of an important piece of equipment that might be helpful in such a situation. We hot-footed it back to Fort Monmouth in our station wagon, breaking a speed limit or two as we went.
Bryant:
Who was with you on that trip?
Davenport:
I think George Harris and Ivan Getting were there, possibly Leo Sullivan on that occasion. I don't remember anyone else.
Bryant:
But you had other project people?
Davenport:
Yes. We had a group of project people. When we checked in at Fort Monmouth that evening, they wouldn't let us on the base. We were civilians. Only military personnel were admitted to the base. It took us a little while the following day to be able to get the issue straightened out, to be permitted to get back on the base again, and get our passes. But we did so, and from that point on, the success of the XT-1 had been demonstrated. There was clearly an interest from the Signal Corps to move forward.
The next major phase, of course, was to arrange a test where it could really get coupled to some guns and see if it would perform in actual use with a gun director. That required a trip to Old Point Comfort, Virginia, to Fortress Monroe, the headquarters of the Coast Artillery Command. They were responsible for anti-aircraft fire control tests at that time. When we got there we set up for test operations and anti-aircraft fire. It's my recollection that at that time we had an M-7 director, a three dimensional cam driven mechanical computer — and did not yet have the M-9 electronic director available for coupling to the guns. The tests there were dramatic, no question about it. The ability to track planes was phenomenally good. The automatic tracking jitter presented almost no problem for the pointer matching operations on the M-7. Even the ability to stand up under fire of nearby 90 mm. antiaircraft guns — though we had never planned it for anything like that — turned out to be excellent. Targets for our relatively low level firing tests were cloth banners with copper wire woven into the fabric. These were towed well behind the aircraft, nevertheless great care had to be exercised during firing tests. Drone aircraft targets for our tests came much later.
We learned two very surprising things: One was that we could see the echo from the shells as they left the gun and watch as they climbed toward the target. The most startling thing was the slow upward progress of a shell on our scopes. It took what seemed to be an infinitely long time and gave plenty of opportunity for the airplane to move out of the way. There was much surprise that we could see a shell at all. But of course the rear diameter of these 90-millimeter shells wasn't so different from the 10 centimeter S-band that we were aiming at the back end of the shell on its way up to the airplane. The other thing that surprised us a great deal was the presence of spurious low altitude signals over water which we later tracked down to being seagulls. So the sensitivity of the unit was really quite remarkable with our home made crystal detectors.
Before continuing I'd like to point out that seeing the 90 mm. projectiles in flight toward the target turned out to have much more significance than we realized at the time. Somewhat later the shells were purposely tracked when we became suspicious that the fuze timing seemed to be consistently in error by an amount that changed with range. I recall that XT-1 was used to measure gun muzzle velocities and to recheck the master ballistic tables that had been used to set fuze timing and gun aiming. Those tables turned out to be in error for the 90 mm. gun. They were recomputed and all gun directors readjusted by the time the SCR-584 went into service.
I still look back on my earliest days at the Radiation Lab when each of us made his own crystals. We'd sit in the crystal laboratory and adjust crystal cat whiskers contact on the silicon. Then we'd bounce them a few times, and if they held up okay, then we'd bring them upstairs and try them on the test system and promptly burn some of them out. But it was not long before the reliability of the equipment had reached sufficiently high levels so that even our brass board, or breadboard XT-1, could perform remarkably well.
Bryant:
I'm curious about two things. You were measuring range. Was the proximity shell available at that time?
Davenport:
No. The proximity shells were not available nor was there any hint as to when they might be.
Bryant:
That was later.
Davenport:
Much later.
Bryant:
Could the computers make use of the measured range?
Davenport:
Yes. Range translates into time-of-flight of the projectile to cover that distance. And that interval varies with elevation of course. Longer time-of-flight means the computer must aim the gun farther ahead of the target. Incidentally, the gun directors were more often referred to as predictors rather than computers at that time.
Another point needs clarification here. The M-4 and M-7 gun directors to which we initially fed data were mechanical units that had existed at the start of the war. They were normally fed by optically tracking the target with slant range information coming from an optical range finder. The Signal Corps hoped to introduce night firing capability with the forthcoming SCR-584 while still keeping the ll optical inputs in reserve. Also planned was a shift to a high precision electronic gun director, the M-9. It was not known when the M-9 would be available so 584 units had to be capable of working with both types of directors. The M-9 was fed data from three large potentiometers mounted in the 584 so no pointer-matching human link would be needed. To complete the system, the output from the M-9 went directly to servo driven guns and fuze cutters, in the final form, only the operators in the 584 and the gun crew to load shells would be required. To the best of my knowledge by early 1944 all of these final elements were in use. The way the system was initially tested, the M-7 gun director was a 1-cubic-yard box rotatable on a pedestal.
Bryant:
Big analog computer?
Davenport:
Yes. It was a mechanical analog computer with gears, differentials and three dimensional cams inside and direction dials outside. There were two telescopes on the outside that operators sat and looked through, one azimuth and one elevation. They physically moved computer box to aim it at the airplane. We used our synchros to drive the concentric indicator dials also on the side of this computer where, by matching pointers, operators caused the computer to track the unseen target. The synchro indicators were driven from our SCR-584.
Bryant:
So this was manual tracking?
Davenport:
This was a manual-tracking link to put data into the computer. We provided automatic tracking to the indicators, and operators did the pointer matching and drove the computer physically back and forth. The range data went through to a fuse cutter, which is located at the back of the gun. The projectile was inserted into the fuse cutter, pointer matched against the timing indicator there. The fuze "cut" means twisted to the correct angle that set the timer in the fuze. The shell then was loaded into the gun, and fired as quickly as possible. As long as the plane stayed on a predictable course, you had a chance. But proximity fuses did not come along and were not used, in fact, until the buzz bomb attacks on London. That was the very first use, to the best of my knowledge, of proximity fuses in the war effort.
Bryant:
Nineteen-forty-four.
Davenport:
June of 1944 marked the first buzz bomb attacks. I was there at the time that happened.
Bryant:
Were you in Britain?
Davenport:
I was assisting the 584s that did that job.
Bryant:
I'm curious about one other thing in your development. When did you switch from rotating the parabolas to rotating the feed?
Davenport:
Very early on. We quickly determined the wobbly parabola was not necessary. So even the first roof system, that tracked did not use a wobbly parabola. But with the very first one that I put together, which was only about a 3-foot aluminum parabola, we didn't know enough then about our beam pattern to try antenna nutation rather than the parabola rotation.
Bryant:
You had the physics all correct?
Davenport:
Yes, I think so.
Learning to be Engineers
Bryant:
But your engineering wasn't developed?
Davenport:
Engineering began to creep in fast. This is an important point that your touching on. We started this program as physicists, looking at things with much more theoretical attention. We quickly found out we had to be practical engineers to make anything work. Our examples of failures included such things as how to drive a generator off a truck engine. It turned out, of course, we never really made use of our generator. They had trailer mounted motor-generators down at Fort Monmouth that would work.
Bryant:
Would it be going too far to say that physicists quickly became very good development engineers?
Davenport:
No, it took a little time. I'd like to digress here. A substantial number of us who were experimental rather than theoretical physicists found it possible to put on a different hat. It did take some time however. Initially of course we were delighted when we could make our first roof system work at all, even with clip leads and laboratory clamps. At the XT-1 stage when we were preparing to leave the Laboratory far behind with our equipment, our construction ideas began to change. However at this point we Laboratory types still always operated the system and maintained it, and often modified things in the field. The real education began for my team when we took equipment to Fort Monmouth, New Jersey, to Fortress Monroe, Virginia, to Patuxent, Rhode Island, to Dahlgren Virginia and to the Naval Airfield at Elizabeth City North Carolina for various tests. Each step increased our awareness that reliability and repairability were virtues that could not be overlooked. Optimum performance which we could easily achieve in Boston could not be taken for granted. Military operators who tried to help us out could not be expected to match our expertise or our tolerance for inconveniences. It was much later that we really learned about manufacturability, spare parts, instruction manuals, and all those things that an engineer expects to handle. However we did learn very early that nothing ever works in the field as well as it does in the laboratory. Nothing! You must over-design for performance by significant amounts to be able to compensate for field conditions.. But fortunately we were over-designing in our mind. We believed this radar should shoot planes down at 50,000 feet if they'd ever get planes up that high. Over-design was a part of our tradition to push performance to the limit. That gave us safety factors in performance so that when equipment got out in the field, it would work quite well even in severe conditions.
Bryant:
By over-design, are you stressing only performance?
Davenport:
Yes. Performance, and not necessarily over-design in terms of reliability because we didn't know how to improve that. None of us had ever seen a big gun fire alongside 300 vacuum tubes. In fact, none of us had ever seen 300 vacuum tubes in a single system before we put this one together. At the start there were serious questions in our minds as to whether anyone could keep complex electronics working at all in the field. Remember we were using the identical electronic parts that were then available for commercial radio sets. They were mass produced and the reliability and life expectancy of these 1940 components was not very high. None had been designed for extreme temperatures, humidity, dust, shock or vibration. Beyond that we were using many parts in circuit applications that often subjected them to signals not found in analog radio sets. So this is a little of the history of the 584's early days.
Once the contracts were signed for purchasing SCR-584 systems — then came the job of taking this same breadboard system and arranging to get it promptly produced by General Electric and Westinghouse as the two prime contractors. Chrysler Corporation was the contractor for the parabola, pedestal, and the precision gearing that went in it. Freuhauf Trailer made the semi-trailer unit in which 584 was to be placed, and Palmer-Bee Corporation was the manufacturer of the mechanical elements of the range units. Palmer-Bee did the range aided tracking assemblies, which were anti-backlash gear-driven systems with synchros and potentiometer to transmit range data for pointer matching in case that had to be used. This liaison with General Electric was a whole new ball game for most of us.
Relations with Private Industry
Bryant:
Liaison with industry?
Davenport:
Yes. I'd never had that experience. I'd once worked for General Electric on college summer jobs in the payroll department of the Schenectady Works. But when I showed up at the door of the Schenectady Works with a bundle of secret drawings, so to speak, and began to sit down with engineering people and say here is what this is all about, and you're going to have to build it this way, it was a whole new role for me. Suddenly a group of college professor types, full of enthusiasm, appeared with responsibilities for directing two of the major electrical corporations in the world — General Electric and Westinghouse. We were not letting them re-engineer a product to a point where it would not pass the performance tests that we had demonstrated in our model. We had to be the principal judges of that. I must say that it was a surprising degree of responsibility that came rather suddenly to some of us. To me, certainly. But on the other side of the coin, the people that we dealt with in industry from K.T. Keller at Chrysler right on down treated these 24 to 30 year old kids coming in from MIT with respect. We were treated with deference because of the technology that we were bringing.
The cooperation that we received from those industrial people was miraculous in my mind, looking back at it now. If there was any NIH problem at all, it never seemed to surface in any of those relationships. Part of it, probably, was our own enthusiasm. We had a belief, an understanding and we were experts. Part of it probably was the label "SECRET" on what we were bringing to them and their ignorance of what had been accomplished. Some didn't know how the part they were making was going to fit in the overall system. So they were respectful, trusting, suggested improvements but did not insist. However, I'm sure that each one recognized that if the assembled system did not work, it would be our responsibility not theirs. We were smart enough, I think, by then to know the difference between what we did know and the things that we didn't know about engineering. Our field experience surely helped. So that period was especially interesting and required a lot of traveling. I was on the train all the time, back and forth to Detroit and Schenectady and Baltimore and various other locations where the subcontractors were working. At the same time that was going on, we were continuing the development work because we were constantly improving the system.
Testing the SCR-584
Davenport:
The first systems to come off the line — 584s — were tested at Camp Davis, North Carolina, were the Anti-Aircraft Command then had its headquarters.
Bryant:
Would that have been before 1942 was over?
Davenport:
Oh, no. This was '43ish.
Bryant:
Was the final assembly done at General Electric and Westinghouse?
Davenport:
Yes. By that time the first of the M-9 electronic gun directors designed by Bell Labs was ready. As I mentioned, these remarkable analog electronic computers required inputs from large precision potentiometers. All SCR-584s had been designed for and equipped with azimuth and elevation potentiometers in the antenna mount and a range potentiometer in the console mounted range unit to transmit data to the external M-9. The Signal Corps had planned and coordinated the combination of the two systems to drive the servo controlled 90 mm. and subsequent 120 mm. anti-aircraft guns.
For the Signal Corps and the Antiaircraft Command these tests were the culmination of a Multi-year effort to revolutionize antiaircraft fire control. Obviously they were very well attended by the working types even though we all expected a few bugs and a failure or two. The top brass were not to be united 'till we had removed all the gremlins and had prepared for a proper demonstration. When all the equipment alignment, parallax adjustment and other setup steps had been completed we called for the target towing airplane. On it's first pass, the 584 searched, found the target, locked on, we verified that we were tracking the target not the tow plane, the M-9 sent output to the guns and all four began to track. For all of us it was a rare example of everything going right for a change. In planning the second pass we decided to have the gun crews prepared to fire on command if all went smoothly. It did, and the four 90 mm. guns began normal rapid fire, aggregating about one round every three or four seconds as I recall. The fifth round destroyed the target and we called a halt. But by the time the fifth round hit, three or four more rounds were already in the air and they each exploded along the track where the banner would have been. All together it was phenomenal debut of a complex project. I teletyped the results back to the Laboratory with great satisfaction.
I have one follow on episode that fascinated me, that I will pass along to you. The liaison with the military at the highest levels was mostly through Bowles's in Washington. Stimson, who was then Secretary of War, whom I had a chance to meet only once or twice, leaned very heavily on Ed for his professional advice on matters relating to this high-technology work.
Bryant:
Is this Edward L. Bowles?
Davenport:
Yes.
Bryant:
A professor of electrical engineering at MIT.
Davenport:
Yes, that's right. He was based in Washington in the Office of the Secretary. Well, for reasons that had, I believe, something to do with 584, Secretary Stimson soon visited Camp Davis, North Carolina. It may have been stimulated by Bowles I suspect. But we of course had the 584 set up right there at the time since we were continuing these tests. It wasn't long after the day of our first successful test that the entire base was readied, everything polished, the generals lined up, and the Secretary of War arrived for an inspection. As he was escorted around and shown various things, he finally came to the firing range on the beach. At this time the military had no indoctrination course or trained personnel so we civilians were in charge of the test firing operation. Now I have always felt that our first-off-the-line -line system was intended to be the highlight of the entire tour. Naturally we had prepared carefully for a repeat of our successful test firing with gun crews at attention and everything ready to go. I stood by at the door of the 584 when Stimson appeared with his party.
He came over and I was introduced to the Secretary. He said "What do we have here?" I replied "This is the latest thing in radar equipment. The SCR-584 is a microwave radar antiaircraft system and we are prepared to fire at a towed target to give you a demonstration of its accuracy." He interrupted "Oh, I've seen radar before." Turning to his aide "What else do you have for us to look at?" as he turned and left. So much for the demonstration for the Secretary of War. The Secretary never saw the inside of a 584 even though we would have been delighted had that happened. But Ed Bowles had done his job, I guess, because the production contracts had been approved. Among the very first 584s were some that were sent over on lend-lease to the British. They were very anxious to test them. You know there were competing systems too. There was the Bell Laboratory's-Western Electric SCR-545, which was Bell's version of an automatic tracking system. I'm delighted to say that ours outperformed it. Only a few SCR-545s were made.
Bryant:
The Canadians must have had a competing system of some sort, too. It's only on the record. I don't know if anybody used it.
Davenport:
I don't recall any tests against any Canadian systems.
Rapid Response
Davenport:
Now let me digress for a moment. In what became Division 8 of the Laboratory in which I worked, everyone felt a special responsibility for making his project succeed. Each system project represented the cumulative efforts of hundreds of Lab people in the supporting groups. Looking back at it now I thing we systems people felt that the reputation of the whole Laboratory depended upon our performance in demonstrating our combined product to the Military Services, particularly in the XT-1 days — and it better work.
There was another factor that helped make our systems projects go smoothly. We operated as a team of equals. Especially when in the field, there was a feeling that regardless of rank you would help out in any way necessary. Project Managers soldered wires, connected cables and drove trucks in addition to their management roles. The name of the game was to move every obstacle out of the way as quickly as possible.
As time passed I personally felt that there were many more applications for such a versatile radar as the 584. I did not aspire to move on to a more advanced project once 584 had gone into production. This follow-through by the same team of people from research through engineering to initial field use, to modifications for new tasks, to more applications was a never ending progression. It characterized my Group 81 team as it did many other systems groups in the Laboratory. As a consequence of these factors, responsibility was clear and rapid feedback by increasingly experienced people produced an almost continues stream of successes.
Another observation at this point concerning rapid response — sometimes referred to as quick reaction — projects may be helpful. The SCR-584 was produced in large quantities. Some three thousand were made and they worked reliably. As a result it was available for adaptation to solve a number of problems that were never envisioned at the time of its design. Many of these applications required only a few units but were needed in a hurry. None would have justified the design of a whole new piece of equipment from the ground up even though a new optimized design would have been better. Each new task required some modification or addition to the basic 584 followed by field tests in the US, and then often by trail in combat. My Group 81 team, supported by Laboratory facilities, became very skilled at rapid response jobs. In fact we sometimes ignored field testing completely. One of us would carry equipment made at the Labs directly into combat areas, installing, debugging it and training military personnel right there. Without exception someone from the group would be on the scene following up on each phase of every new project to assure its success. We also learned early how important operator training really is.
SCR-584 in England
Davenport:
In 1943 when the first of the 584s were starting to come off the line, several were promptly shipped to England as deck cargo. I went over to receive the first one and demonstrate its operation to the British forces. I dashed off to England for the unveiling and learned about British red tape. Though the 584 was shipped to a British destination I could find nothing about its arrival from the British. Ship movements were very secret in those days. On that, the U.S. Embassy came to my rescue. I checked out the arrangements for its testing with the British. Upon my inquiries as to when it would arrive, I found out that it was expected to take about two weeks after it landed in Scotland to move it a few hundred miles south to the test area. Well, two weeks was certainly not our style.
So I went to Scotland and found the ship. The 584 was on deck. I found the Habormastor and explained the secrecy of the program and its importance to the war effort. Can we get this thing unloaded rather promptly? The harbormaster said, "Oh, certainly if you need that in a hurry it's deck cargo and we can get that off of there." I said, "Could I come back tomorrow and can you have it on the ground?" "Yes, we'll have it ready for you tomorrow." So then I found the nearest US base, located the transportation section and talked the colonel in charge into loaning me a driver and a tractor unit (it was a standard semi-trailer, so any standard Army tractor unit would pull it). With a driver and a truck, the following morning we got the 584, signed the papers, and drove off with it. I rode with the driver, and we had it at the test site the following morning ready to go. Now this was typical of the way in which we tried to operate. We attempted to eliminate red tape. We knew we had credentials if we really needed to use them but we tried to avoid going through the command structure to get orders cut that would cause these things to happen since that took extra time. As a civilian, I wasn't in uniform.
Bryant:
Were you informed as to what procedure you could use to do that?
Davenport:
Oh, I certainly knew how to get official orders on paper. But the short cuts you learned yourself by being convincing. I told people, "I'm a civilian on an important classified assignment. How do you handle things like this?" The cooperation that we got all the way through was remarkable. The 584 performed very well for the British — as one would expect. That began the flow of such equipment to the British.
Bryant:
Was this British Army?
First Combat Uses of 584
Davenport:
It was their equivalent of our Anti-Aircraft Command, but under the Army's auspices. The first use, as you know, was at the Anzio beachhead.
Bryant:
Do you mean the first use in combat?
Davenport:
Yes, in combat, it was at Anzio beachhead, the invasion in Italy in 1944. From that point on, my involvements were largely for operations which were rapid response emergencies. One emergency that I can remember was an important one. It was the arrival of the buzz bombs. D-Day was also considered an emergency. So since I was in England several months before D-Day I took on the responsibility for arranging and supervising the waterproofing of 584s so that they could be floated ashore. Now this was something we never dreamed would become necessary. That required a trip to Wales with a British group, and we jointly worked with a team that was preparing trucks and other vehicles constructing snorkels for the air intakes and exhausts.
Bryant:
Oh, they added snorkels so it could almost drive in the water.
Davenport:
The snorkel could suck in air without sucking in water, and they would run almost submerged. The same group of people worked on our 584s, and I briefly supervised that. We had ventilation ducts under the floors and all sorts of unusual things to take care of. That was pre-D-Day extra work. The reason I was in England involved a different emergency invasion project: could the 584 provide close support for our Ninth Tactical Air Force Group? They flew dive bombers. We could readily track their airplanes. They often couldn't find their ground targets working from maps. Our quick solution was to make a simple synchro driven plotter to show the aircraft position on a map inside of the 584 itself, and have a controller talk to the airplane and vector him to his target. We had excellent detail maps of all of Europe. Close support meant attacking targets directly ahead of our advancing ground troops. We quickly designed and built plotting boards over there in some quantities prior to D-Day so that they would be able to do that close support work. That was a significant new application.
Bryant:
Did you work through the British branch of the Radiation Lab?
Davenport:
Yes, BBRL at that time.
Bryant:
Did BBRL support all of the American operations regardless whether it was Air Force or Army application?
Davenport:
That's right, all of the Radiation Lab operations. And it used services and facilities of the Technical Research Establishment and U.S. military bases. American military bases worked on making up parts for me for the plotting boards. There were machine shops at one U.S. Air Force base that helped us with parts and assembly. Some things were made in this country at RRL and at some in our own facilities at Radiation Laboratory and shipped across. In general it was these sorts of rapid response adaptations of 584 that kept my team occupied.
Normandy
Davenport:
I went over to France shortly after D-Day to be certain that the first close support 584 was working properly in its role for dive bombing. I helped train the 555th Signal Air Warning Battalion at BBRL and in Normandy which had been assigned to operate the unit. Leo Sullivan, George Huff and I were involved with that exercise and we were near St. Lo at the 584 site when the breakthrough occurred. Prior to that we managed to get shot at a few times by German 88 mm. guns while searching for closer sites for the 584. We watched the St.-Lo breakthrough while we tracked the waves of Eighth Air Force bombers that were coming. They dropped bombs for several hours on the German lines to prepare the way for the attack. We watched and tracked bombers as they came closer and closer to the American lines. The dust that was blown up by the explosions from the first waves of planes was blowing toward our lines. The following waves of planes were dropping blind on that dust cloud. One of the most tragic experiences I have ever had was to watch that happen, knowing from our plots that the explosions were dangerously close to American lines and being helpless to reach those bombers. Of course we reported to Ninth Tactical what was happening but without radio communications links to Eighth Air Force no one could inform the bombers. That was a tragedy. American troops were killed, and we lost one of our generals in that episode.
London: Defending Against Buzz Bombs
Davenport:
After that I came back to England since the close support equipment was working very well. When I returned I discovered we were in the midst of the London buzz bomb attacks. They had started about the 6th of June D Day before I had gone to France. When we went to France, those of us who were civilians had to wear military officers uniforms. If we were captured, we would be shot as spies if we were not in uniform. So we were supplied with assimilated ranks and credentials as military officers, but without rank insignia on our shoulders. It confused a lot of troops. They didn't know whether to salute us or not. Only when we were in battle zones did we wear uniforms
Bryant:
Did you have to wear a steel helmet and all?
Davenport:
No I didn't wear a steel helmet. Although we got shot at, we didn't wear steel helmets. I came back and was preparing to return to the USA after the successful operation of close support equipment in France knowing that it was working well. But I found that we were in a real problem with the V-1 buzz bombs. While I was in France, this buzz bomb situation had grown serious. I learned it had already reached a point where the British had asked the United States forces to bring 584 gun batteries to the south coast of England in order to try to knock down the buzz bombs out over the water before they crossed the coast into England. Colonel Arthur Warner, who earlier had been one of my staff people at Rad Lab, was by then attached to the Anti-Aircraft Command reporting to Eisenhower in the European Theater of Operations. Art Warner, many years my senior, had worked in my Group for some time as a civilian Radiation Lab Staff Member before he was recommissioned with the rank of Lt. Colonel in the Anti Aircraft Artillery. His knowledge of the 584 and the associated gun batteries was extensive and he became a principal advisor on AA in the European Theater. Thus, when I got a message from Art asking me to stop by his office at ETOUSA Headquarters outside of London I assumed it was just for a friendly visit before my return to the USA the following day. My assumption was wrong.
Art began our conversation by intimating that the request for US Anti aircraft gun batteries had gone from Churchill to Roosevelt implying that it had very high priority and that our reputation was at stake. In order to respond quickly, AA Command had directed that 584s together with the new M-9 directors and the necessary 90 mm. gun batteries be diverted to the south coast. This was the preferred location for AA since firing could be over water and wounded bombs would drop harmlessly into the sea. Since our AA troops and equipment were already in France, they had requested that Washington HQ. dispatch other AA crews from other locations to England. They had arrived a few days earlier and Art said they seemed to be having trouble since the twenty or so batteries in place had managed to destroy almost no buzz bombs. He had requested immediate help from the Laboratory and was told that I was just returning from France.
Art had a car and military driver waiting to take us to the south coast immediately. We arrived as darkness was falling and I watched seven buzz bombs all within range of one four-gun battery fly over us and on to London. Only a few shots were fired. Several interviews with the two or three crews we visited that night turned up the problem quickly. The men previously had been manning SCR-268 equipment with mechanical gun directors in places like Iceland where they had never seen an enemy plane. Thus they didn't know how to align the radar with the M-9 or make the guns point in the correct direction. Under those circumstances they deserved a lot of credit. It was a wonder that they had been able to connect all the cables between the equipments and find the on-off switches. They were trying to fire at targets while still reading the several instruction manuals. [The three War Department Manuals that accompany the 584 alone total almost 1300 pages.] I could make the 584s perform beautifully in a few moments for them but that was of no real help.
We returned to London in the early hours of the morning. No easy trip with blackout headlights. On the way down had Art confided to me that these batteries had been equipped with proximity fused 90 mm. shells, still a very secret development. This was their first use in the European Theater and it was permitted only because unexploded rounds would fall into the sea. This statement increased my growing suspicion that Art Warner's recommendations lay behind this entire deployment of US anti aircraft batteries against the V-1 buzz bombs. Every step had been expertly planned to use our best equipment under optimum circumstances. Then Art said, "Obviously we've got to do something about this situation. What do you suggest?" I said, "I'll get Leo Sullivan who has just returned to BBRL and we'll go down and straighten it out. I'll delay my return trip and Leo and I will spend a week or so aligning equipment and training every one of those batteries. The others must all be having the same trouble. When we've finished, Leo will be on call at BBRL in case you need more help." I'm sure Art expected that answer even though our job officially related only to the 584s. I commandeered an ambulance in which we could sleep if necessary had the crosses painted over, put some test equipment aboard from BBRL and off we went to the south coast. Sullivan and I checked into a hotel on the beach. About five o'clock in the morning we heard the first buzz bomb approaching. We looked, and sure enough, coming over the water at an altitude of between 500 and 1,000 feet, was this damned little airplane going like crazy and making all sorts of racket.
As it went right overhead, I heard the anti-aircraft guns go boom! and boom! and boom! way behind it. Nothing came anywhere near the thing. I said to Sully, "We sure have our work cut out for us." So we spent the next two weeks going from site to site. I think there were 20 or 25 584 batteries lined up all along the coast over a distance of fifty miles. We went from one to the next, taught the crews how to align and run the equipment, tuned everything up and left each one prepared to handle incoming V-1s. Nothing made us happier than going back to our beach hotel at the end of that period and watching the next buzz bombs come in. Not one of them managed to get near our hotel window before it was out of the sky. This was the first coupling of 584s, M-9 electronic directors, proximity fuses, and 90-millimeter guns. Looking back now it's easy to see that this combination was a brilliant concept when set up in this fashion. The V-1 was a very high speed weapon surpassing speds of 350 mph. at very low altitudes. SCR-584 ordinarily was not very effective at such low angels because of ground clutter. And no elements of the combination were designed for an application like this one. Buzz bombs were within firing range for only a short time and setting fuses by hand with pointer matched fuse cutters would have been much too slow. But the combination with proximity fuzes was deadly.
I recall one other important factor. I had asked the Lab back in Boston quickly to assemble and ship some modification kits for these 584s. The normal 584 located targets by scanning round and round with the automatic tracking off. For this application more than three quarters of that scan was wasted. Simple electronic Sector Scan Units that automatically oscillated the scan through a variable azimuth angle were promptly delivered and installed on all 584s in this service.
Bryant:
Was that August and September of '44?
Davenport:
That was in July of 1944. That was a very satisfying episode. I'm sure you've heard the numbers quoted often, but they ran essentially like this: Up to the time we got those 584s working, the Germans had been firing in the order of 100 to 110 buzz bombs a day, mostly in groups. Roughly 95 percent of them got through to London. There were only two defenses that made any difference at all. Short range anti-aircraft fire was useless because V-1 altitudes were so low and speeds so high, 350+ miles an hour. They could not be effectively followed by optical sights. Barrage balloons were up, but they were also useless. The third thing they had was aircraft. Now there were a few aircraft that would actually catch a buzz bomb, but very few. Remember, most of the fighter aircraft were built to go at high speed at 20,000 feet, not at 1,000 feet. But the British had a few water-injected engine Hurricanes that would catch a buzz bomb. The Americans had a few modified P-51s that could make that speed at low altitude which would catch a buzz bomb. But in no way could you keep them in the air enough. They had a total of 15 or 20 planes that would actually do it. It took rare courage for a pilot to chase a buzz bomb, catch up and shoot it down. It was a very small target. It required full speed to catch and a close approach to hit. And when you did hit it, the 1000 lb. charge inside usually exploded right in front of you. Then you had to fly through the debris for a good measure. After some practice a few pilots daringly dove on the bombs from above to get up speed, fired and then pulled up without hitting the ground. The Air Groups were delighted to turn over the defense to antiaircraft. At the end of the two weeks when I left to come back to the USA we were knocking something like 95 percent of them out of the sky before they got to the coastline. So that was one of the most successful and dramatic quick-reaction campaigns.
From that point on 584s continued to be used for all sort of different applications and those of us who were part of the original team stayed with that work. For example, I can remember going down to El Paso, Texas to Fort Bliss and testing one modified for mortar location, backplotting the flight of an enemy mortar shell to determine the point from which it had been fired. We fed those data by voice to nearby ground artillery so that they could fire at the mortar location. Mortars are quickly portable, and you've got to get a quick fix on the location. We developed a plotter that showed the trajectory of the incoming shell we were tracking.
Radar-Guided Missile
Davenport:
Among all the projects underway toward the end of the war one major development stands out. It was the first radar guided missile. That arose from an earlier project and we named our version Willy Orphan. An attempt had been made to knock out the submarine pens along the west coast of Europe that the Germans had built to protect their submarines. The Germans had heavily reinforced the roofs of the pens so that ordinary bombs were ineffective. Large doors covered the front to prevent damage from torpedo attack. It had occurred to someone that our retired war-weary B-17s might solve the problem. I believe it was Colonel Rand who proposed loading a B-17 with explosives putting a television camera in the nose and flying it through the pen doors. The plane took off with a pilot aboard who bailed out over England. It was flown by remote control form a mother ship, another B-17, that accompanied it across the channel. A television camera in the nose with display in the mother ship allowed the latter to guide the B-17 on its final approach at very low altitude into the pen doors.
I don't know whether any missions were successfully completed that way, but I do know that Joseph Kennedy was the pilot on a failed attempt and was killed when the B-17 exploded prematurely on takeoff. I recall that Colonel Rand was looking for a radar method of guiding the B-17 when the Laboratory became involved. Naturally it was our suggestion that we adapt the readily available 584 to do the job. At first it appeared not unlike a range extended version of our successful close support system. Before long however I was in charge of the design of a complete guided missile control system with a number of unique features. By incorporating a beacon in the aircraft we could extend the range reliably and also use it to decipher steering signals derived from pulse variations transmitted from the 584. Next it was critical that we develop a security technique that would prevent the enemy from taking over control or jamming. I recognized that the security problem became more difficult at long ranges as our signals weakened and the enemy's strengthened. I suggested and we tried a coding system involving multiple pulses precisely spaced which could be changed for each flight. The guidance information was transmitted by pulse position modulating of one or more of the coded pulses. Delay lines could be plugged into each beacon and into the 584 to set up any of thousands of combinations. A patent was filed on this system in my name and I understand it was extensively used for many years after the war.
The first laboratory model of this guidance system I took to Eglin Field in Florida for its tests. I recall that it was planned to fit a US made version of the V-1 buzz bomb being built by Ford Motor Co. It is my recollection that the testing was successfully completed using the beacon in an aircraft before the sudden end of the war in Europe. I cannot recall whether we completed the installation in a buzz bomb or whether we had reached a point where production was feasible.
Impact of Rad Lab on Postwar Career
Davenport:
When I was released from the Radiation Lab, I went back to the University of Pittsburgh to finish my Ph.D. degree. It was complete except for a thesis. Of course my original thesis topic was by then obsolete after five years. At Alexander Allen's urging I was asked, "Did you do anything at MIT that might be suitable for a thesis topic?" And I thought the most suitable and current one was the guided-missile control system. There was a patent filed on it, so I knew that it was protected and clear. So I asked whether there was a chance that the University would consider a classified MIT project. The University of Pittsburgh accepted the idea. This was a very generous concession to grant a Ph.D. on thesis work done at MIT. So my dissertation was written as a classified document. The review and oral examination was by MIT personnel who had secret clearance, and my doctorate was granted by the University of Pittsburgh on that basis. That's pretty unusual because theses are supposed to be published. This one wasn't declassified for 25 years. Even then it was considered sensitive I'm told because the same coding system was still in use.
Bryant:
What effects did your Radiation Laboratory experience have on your subsequent career?
Davenport:
It certainly had one big effect in getting my degree. Did it have others? You bet your life! For every one of us, the laboratory experience was surely a revelation. Some people must look back at the war period and say it was a difficult period. But for me it was just the opposite. It was a fascinating period. It was exciting. It opened opportunities for many of us to accept responsibilities that we never would have dreamed of, that we never would have been given, had there not been a war, and had we not been associated with MIT in this kind of a role. I never would have been a business executive in my succeeding career. I was going to be a college professor. It gave many of us confidence that we could do things that we had not necessarily been trained for. Many of us stepped into roles that were not normal or natural in our career development prior to that MIT period.
So I ended up with a wonderful job at the end of the war, having finished that Ph.D. It was at Harvard. I was asked by Curry Street and Ken Bainbridge to come to Harvard and build a new cyclotron for them. Consider the fact that I had been an x-ray spectroscopist, who learned a little about radar during the war, and was suddenly being asked to take on the building of a huge new frequency modulated cyclotron. I didn't know enough about a cyclotron to build anything of that sort. But we had confidence. There was nothing, after our Radiation Lab experience that we weren't willing, within reason, to tackle. So knowing Curry Street and Ken Bainbridge, I started off selecting a space for a building, had the building designed by the University architects, and hired the staff (many of them from the Radiation Laboratory). We settled on and designed a 92-inch machine. I managed to find sources to buy the steel, have it machined at Watertown Arsenal for the huge magnet. I dealt with General Electric for the coils and DC generator. We built a 92-inch cyclotron and a complete nuclear laboratory in the next four years. It was a tremendous challenge, and a tremendous opportunity. I wasn't skilled in that art, but the Radiation Laboratory experience was invaluable. I enjoyed the assignment at Harvard and did some teaching too.
Bryant:
Project engineering is a great discipline.
Davenport:
It is. At the end of that time in 1950, I had a decision to make: Whether to stay on at Harvard, accept a tenure appointment, and stay with physics research and teaching. Or do something else. I was asked to accept a job in industry, and I decided to consider it. I felt that I'd spent all of my life working in universities up to that time, and before I got to be too old, I thought I'd like to try a job in industry to see what challenges it offered. If it didn't work out, I could always go back to teaching somewhere, and I enjoyed teaching. So I accepted a job with the Perkin-Elmer Corporation and promptly became executive vice president. I stayed there for seven years, then went on to the Presidency of the Sylvania-Corning Nuclear Corporation. Through acquisitions I became the President of General Telephone and Electronics Laboratories for many years. Finally I became Chief Scientist of GTE. So yes, the Radiation Lab experience was a tremendously valuable thing for those of us who were given the opportunity.
Bryant:
You've mentioned two things: One, your own awareness and your own personal abilities.
Davenport:
It helped bring out the best in all of us.
Bryant:
As well as tools of knowing how to use groups of people and resources.
Davenport:
You find out quickly when you're thrown into something like the Laboratory that the individual science project that we carried out in trying to get a degree is very different from the team science work at a place like the Radiation Laboratory. It taught all of us how much more could be accomplished by organizing the subdivision of tasks, by the willingness to depend on the work of others, and by the ability to direct a team by setting objectives that are clear to to everybody. Big science projects became much more widely understood as a result of the Lab experience, and many more people were skilled. Now that isn't to say that no other cyclotrons had been built pre-war. Of course they had, and there were team efforts, too. But relatively few people had that experience. The Radiation Lab trained a lot of people on how to handle big science.
Project Management at Rad Lab
Bryant:
This brings up a point I'd like to go back to — 1941 — if we could. How formal and detailed was your scheduling for the projects?
Davenport:
Very undetailed by today's standards. There were no such things as program charts and plans.
Bryant:
What did you put on paper to schedule a project?
Davenport:
You raise a very interesting point. A tremendous amount of this work was done in one's head because the identification of bottlenecks presented no difficulty. We didn't have to put anything on paper to know where the bottlenecks were going to be. We worked on getting the total job done as quickly as possible. I don't recall at any time having the equivalent of a progress chart.
Bryant:
A Gant chart?
Davenport:
What I call a PERT chart. At no time that I can recall did we use anything like that. Of course we did have schedules, but the formality of the charting method did not figure in my project control.
Bryant:
Did your supervisor have a schedule as well?
Davenport:
No.
Bryant:
Was it simply handled verbally between you?
Davenport:
Yes. My first supervisor, of course, was Louis Ridenour. Then Ivan Getting inherited Louis's assignment and finally it became Division 8 instead of Project 2. I started with Project 2, and when Division 8 was created for systems work, I became Group 81. I was in charge of the system work under Division 8 until the number of identifiable systems began to grow. Group 81 had about twenty permanent members plus many others that were assigned to the project from other groups. Division 8 also had non-systems groups to assist us.
Bryant:
How did you control the drawings for parts and subassemblies?
Davenport:
We had our own drafting staff in Division 8 that handled drawing.
Bryant:
Were draftsmen assigned to you?
Davenport:
Yes, we had draftsmen assigned. And of course a great deal of the design and construction work was parceled out. Perhaps I can clarify how we coordinated the work in this way. In the XT-1 days I had specific people assigned to the project who worked in the several "labs." There was a receiver lab, a modulator lab, an indicator lab etc. that designed and produced parts for our project. The assigned staff members and their technicians worked full time on XT-1 until their component had been installed and tested, usually some months. Regular meetings were held in which all staff members working on the project met to report progress, share ideas and attack problems. Each lab produced schematic drawings. After the system was working those same staff members were on call whenever needed. There was minimal standardization and little concern over costs at that time since XT-1 was not engineered for production.
Now once the General Electric Company and Westinghouse were involved, that was a different era. All drawings then had to flow through a single source. I was in charge as the principal liaison with the manufacturers. I had personnel who were responsible for standards and for the continuing supply of updated changes and so on.
Bryant:
Did you keep some kind of message center to regulate the flow of drawings?
Davenport:
Yes, we did. Although that was under my general direction, but I do not recall what system was used for that. That was delegated. We had people in our Division who grew with our work. One was Jack Clemente who was a great jack of all trades. You could give Jack most any job to do. He'd handle purchasing and many other unassigned things. But the teams worked very well together.
Bryant:
What do you consider your most important contribution at Rad Lab?
Davenport:
I am personally most identified with SCR-584. There is no other contribution that I could single out other than the multiple successes of that particular project. The 584 and its several adaptations surely was my most important role.
Run-in with Luis Alvarez
Bryant:
Did you serve on the Steering Committee?
Davenport:
No. I did not serve on the Steering Committee of the Lab. The Steering Committee served as a policy making body and facilitator for the Laboratories working groups. Such a body was needed of course, but the majority of us who were responsible for getting things done were delighted to be freed from that kind of activity. The members of the Steering Committee were top notch scientists drawn into a rather non-scientific role. Each had been a successful competitive scientist and many were strong willed, which, as the Lab grew, lead to a certain amount of "political" struggle for power and influence. I'm sure that some members enjoyed both the prestige and the politics even though it may have detracted from their scientific contributions. Fortunately, very little of that atmosphere leaked beyond those of us who were Group Leaders. I was lucky that my boss, Ivan Getting, who was a member, grew very skilled in this committee job. My one experience with this phenomenon came very early in the Laboratory history when I became a fall guy.
Soon after the first XT-1 performed successfully at Fort Monroe for the Anti-aircraft Command, I was asked to take it and the team to Elizabeth City, North Carolina for some Navy tests on blind landing. I had been away from the Lab for some time, but off we went for these new tests. These were done under the Navy. It seems that this suggestion had come from Luis Alvarez who had been dreaming of ideas for the Navy, one of which was blind landing airplanes on a carrier at night. Luis had figured that since the XT-1 could effectively track an airplane, a controller inside the truck could direct the pilot onto the proper glide path and vector the plane right to the landing deck. With only a minor nod in Ivan Getting's direction, my then boss, Luis had arranged with the Navy for tests to be run at a naval Air station at Elizabeth City, North Carolina. It had a long flat runway with an approach over the water that might simulate a carrier landing. Our first test required bringing in the airplane on a low angle glide path. Our experience was all higher angle firing runs, not low over the horizon. The XT-1 rather quickly ran into problems. The problem was that it would not continuously track this airplane. It would lose the track periodically during the approach and would suddenly point into the ground.
Bryant:
Would it look at the image?
Davenport:
That's what we quickly figured out it had to be. It was an image problem. So I called Luis to say that I was sorry, but I didn't think the XT-1 or any automatic tracking system would be suitable for this application. If we were going to use XT-1 it was going to point to the strongest signal, and that the image, when it became stronger took over control. A manual override system didn't make any sense in my opinion. So what he really needed was a different kind of system from automatic tracking. Clearly it needed a scanner that would scan up and down so that the operator could see both the plane and the reflected image and could separate one from the other. A vertically-oscillating scan system with a proper indicator could be manually directed to follow only the aircraft signal ignore the image. I could not do that with XT-1.
Luis was pretty unhappy, when the XT-1 did not work out. When I returned to the Laboratory, I found that my name was not highly revered for having failed in an operation where I should have succeeded. Ivan called me into the office and said, "I don't know what you've done, but you certainly rubbed Luis Alvarez the wrong way. Luis is pretty influential and quite unhappy with your performance." And I said, "Well, I can tell you exactly what happened and what I reported to Luis." After explaining it, Ivan said, "I hadn't thought about that, but that's absolutely right. There's no way in which you could separate the two signals." I said, "I just told Luis the truth. It will not work. And he's got to recognize that this is not the right route to go for a blind-landing system." I ran into what I felt were political problems. I do not know whether this is the truth or not, but I later heard from several sources that Luis Alvarez had promised the Navy that he had an answer to their blind-landing problem, and his answer was XT-1. What I had done was to pull the rug out from under Luis Alvarez and possibly cause him to lose face with the Navy. Either he could admit to a failure based up on something he should have foreseen as a self professed expert, or have another reason why XT-1 didn't work. I was selected as the other reason. I could not make it do the job. But that was my only political upset in my entire history of relationships with Laboratory personnel from the top to the bottom. After that I don't think I ever got along with Luis Alvarez as well as with all the other people in the Lab. By the way, Luis recovered rather quickly by announcing that he had a far better concept that would take somewhat longer to develop — a vertical scanning system for blind landing.
Bryant:
You should have been given credit for getting the ground-controlled approach project on the right track.
Davenport:
Well, looking back at it now with the benefit of historical perspective, the vertically-oscillating antenna which I had suggested to Luis as a way around our problem turned out to be the way it was eventually made. But none of us knew then. Apparently I had embarrassed him without knowing it. I did not realize that he had made commitments to and promises to the Navy.
Bryant:
But he had not revealed that to you. You were not prompted to being cautious with whom you talked.
Davenport:
No, but very few ever know the full story. But these things happen. I'm sure there are many other things that would be interesting, and I hope I've covered some of the ones that will be helpful to you.
Bryant:
Well, I thank you very much.
