Oral-History:Robert Kyhl

About Robert L. Kyhl

Kyhl was a Physics graduate student at University of Chicago, working on microwaves and a microwave magnetron. Recruited by Rabi, he joined the MIT Rad Lab in the summer of 1941. He first worked in the magnetron group, where his tasks included liaison with Raytheon. Much magnetron work was not theoretical, but figuring out the mechanics of making cranky magnetrons behave well, and go to proper modes reliably. He then spent most of the rest of the war in the test equipment group, where they devised and/or improved the directional coupler, echo box, spectrum analyzers, and variable attenuator. Kyhl came up with a version of the magic-tee, a little after Tyrrell of Bell Labs, but Bob Dicke a little later independently came up with the complete version. After the war he continued as a “microwave plumber” in one form or another. He received his PhD from MIT, and ultimately ended up as a professor of electrical engineering there.

About the Interview

ROBERT L. KYHL: An Interview Conducted by William Aspray, IEEE History Center, 13 June 1991

Interview # 095 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:

Robert L. Kyhl, an oral history conducted in 1991 by William Aspray, IEEE History Center, Piscataway, NJ, USA.

Interview

Interview: Robert L. Kyhl

Interviewer: William Aspray

Date: 13 June 1991

Location: Boston, Massachusetts

Educational Background

Aspray:

This is an interview on the 13th of June 1991 with Mr. Kyhl. The interviewer is William Aspray. This is part of the MIT Radiation Laboratory Oral History Project. Let me begin by asking you if you could describe what your education was, and what your background was before you came to the Radiation Laboratory.

Kyhl:

I was a graduate student at the University of Chicago in physics and, strangely enough, working on microwaves and even on a microwave magnetron. [Chuckling]

Aspray:

That's really quite unusual, isn't it?

Kyhl:

I was one of the few people to come to Radiation Lab who had worked in microwaves before.

Aspray:

Was there a faculty member at the university interested in the subject?

Kyhl:

Yes. I'm trying to think of his name now. I've gotten to the point where name recall is difficult. I was not getting very far with it, actually. I was pretty much thrashing around, and worrying about being called up for the draft.

Recruitment to Radiation Laboratory

Aspray:

How did you come to the Radiation Laboratory? What was the means?

Kyhl:

Prof. I.I. Rabi came through Chicago on a recruiting trip, among other things, looked into the room where I was wielding a blowtorch and doing some glass-blowing and said, "What are you doing?" I said, "I'm making a magnetron." [Laughter] This was a smooth anode magnetron, a very low-power device. And Rabi said, "What!!!" [Laughter] Then he said, "I think you'd better come with us." And I said, "What!!!"

Aspray:

I see.

Kyhl:

Very soon thereafter I found myself in Cambridge, and not too many months after, what was left of the Chicago physics department was taken over by the Manhattan Project.

Collaboration with Raytheon

Aspray:

Right. What date did you arrive at the Lab, approximately?

Kyhl:

That would have been in the summer of '41. So it was still a fairly small operation. I worked first in the room in Building 4, where the operation was started, where they have now the plaque marking the foundation.

Aspray:

Yes. What duties were you assigned?

Kyhl:

I started out with the magnetron group, naturally enough. We were, of course, trying to develop on the magnetrons that the British had brought over. We didn't have a tube shop at MIT at that time. So arrangements were made (I don't know how because I was an underling and not involved) with the Raytheon Company. They had a small plant out on Chapel Street in Newton where they made tubes for radio receivers, an old New England mill building. They also made some transmitting tubes, mostly for sale to radio amateurs.

Aspray:

I see.

Kyhl:

So they had facilities. They knew nothing about microwaves. I was sent out as the liaison person from the Radiation Lab to Raytheon and worked with them on the construction of the magnetrons. From that start, the whole development of the Raytheon empire has grown.

Aspray:

Yes. Do you want to describe in a bit more detail what it was like to work with them?

Kyhl:

I worked with the legendary Percy Spencer at Raytheon, of whom you may have heard.

Aspray:

I've heard the name, yes.

Kyhl:

Percy Spencer came from Maine. He had the temper of a bull moose in mating season. [Laughter] I was a mousy sort of person, but we got along very well together somehow. He had his machinists and instrument-makers who manufactured magnetron parts. Percy Spencer assembled them with his bare hands and a blowtorch and evacuated and baked them out. When they were sealed off and were cool enough to handle, I would drive them down to MIT in my 60-horse Ford. We had the equipment for testing them, and doing experiments with these magnetrons.

Aspray:

Audio File

MP3 Audio
(095 - kyhl - clip 1.mp3)

What was the strength and the weakness of the Raytheon group?

Kyhl:

It was a very small operation. Percy Spencer was an extremely ingenious and extremely inventive person with strong opinions and strong drives, who ran a tight ship in his little place there — a small room with cubbyholes and things — at Raytheon. We worked together on how to assemble these things, and how to design them. At the time, we were playing strapping games on the magnetrons. You connect jumper wires between the various anodes around the eight- or ten- or whatever-anode magnetron to try to improve the performance. The problem, of course, is that if you have an eight-cavity magnetron, then you have eight resonances. The question is, which resonance will the magnetron pick? It usually will pick the wrong one. So you then try by playing these strapping games to make the magnetron behave.

We were playing numbers games. At first we didn't try the obvious thing, which was simply to connect all odd anode segments with one ring, and all even anode segments with another ring — the so-called complete double-ringed-strapped magnetron. Why we didn't try that straight away, I don't know. Percy Spencer was one of the people who said, "You people at MIT don't know what you're doing. Why don't you do this." He was one of the people who suggested that; I can't say he was the originator of that because there may have been others at the same time who did. Incidentally, I sometimes tell an anecdote about the early Radiation Lab work with magnetrons, and I'm not sure it's true at all. So take this with a grain of salt.

Aspray:

Sure.

Kyhl:

There was a period during the submarine battle in the North Atlantic, when if one of these magnetrons tested at high power when I brought it down from Raytheon, the British attaché would have a Lancaster Bomber warmed up at Logan. They would fly the magnetron to Britain that night. That meant they could put one more radar search plane in the North Atlantic. Now it sounds dramatic, and it may be totally made up. I don't know if there's any truth to it. Even if it's not true, however it's illustrative of the critical demand and the kind of atmosphere under which we worked.

Aspray:

Was there a kind of requirement for quality or performance of these tubes that was not customary at Raytheon? Or was this well within the design constraints that they would normally work with?

Kyhl:

Raytheon knew nothing about microwaves. They were making low-frequency tubes, and they had a tube assembly line. The magnetron, of course, was made of OFHC copper, which ordinary tubes aren't. Ordinary tubes are made of glass and tungsten and tantalum and this sort of thing. So working with the copper, and sealing glass to copper, these were the techniques that they had to work out. Later on, of course, when they expanded the facilities at Raytheon, Percy Spencer made an extremely important contribution to high-volume production of magnetrons. The original way of doing it was to take a block of copper and machine and bore out the cavity holes, and cut through the gaps with milling.

It had to be done with high precision, and it was an extremely laborious job requiring the highest skill from the machinist. Percy Spencer and his group had such a skill, and they proved they could do it. But it was a hard way to run the game. One of Percy's important contributions was to say, "The obvious way to do this is to make the anode block by stampings." You'd stamp a whole lot of lamina out of the copper. Having stamped them out, you'd then stack them up and silver-braise them together to make an anode block. If the stamping wasn't quite accurate so that some of the holes weren't quite the same as others, you just rotate all of the lamina around on the mandrel and then braise it together. It averages out, and you get a precision job out of it. It was the kind of inventiveness that Percy Spencer had. There are lots of anecdotes about the life of Percy Spencer, but I don't think that belongs necessarily in this history. [Chuckling] That's part of the Raytheon annals.

Magnetron Group

Aspray:

Right. Come back and tell me about what you were doing at this same time in some more detail.

Kyhl:

Other than working back and forth between MIT and Raytheon, I was working at MIT on measuring up the magnetrons, playing with the magnetrons. We did such gross things in those days as manually bend the cathode lead stems to realign the cathode in the center of the magnetron. [Laughter] We measured the power from the magnetrons by the heating of water — average power. We measured the water flow with a beaker and the clock on the wall. One day we got a magnetron efficiency of 120 percent, and we were scratching our heads over this. I.I. Rabi came out of his office, which was in the corner of the lab we worked in, in Building 4, and we told him we were disturbed, there was something wrong. [Chuckling] And Rabi looked up on the wall, and said, "Isn't that clock going awfully slow?" [Chuckling] The sweep second on the electric clock was running at half speed.

Aspray:

I see. [Chuckling]

Kyhl:

Which is possible, but unusual. We silly people hadn't noticed that. Rabi noticed it instantly.

Aspray:

I see. How large a group were you in the magnetron group at the time.

Kyhl:

George Collins headed it up. There was Louis Moose, and there was Al Clogston. There was Larry Walker. There was Foster Rieke, for whom the Rieke diagram is named. There were many people I don't recall. We all worked on measuring properties of the magnetron, and the Rieke diagram was developed at that time. Allaire, who's now with Raytheon, was also there. It was a fair-sized group.

Aspray:

What was the style of work? Was each person assigned a special task? Or were you working as small informal groups? Did Collins direct things closely, or was it fairly loose?

Kyhl:

There really wasn't much in the way of administrative organization. But Collins assigned particular parts of the job to people as their interests would show. It was basically a common enterprise, everybody getting together on it. If something interesting had to be discussed, anybody who was interested got together and discussed it. It was a free-wheeling sort of thing with everybody playing whatever part he could. When something obviously needed some special attention, one person would say either, "I'll take this over," or Collins would say, "Why don't you take this over." But that wouldn't mean he'd go off by himself. The interplay continued. That was true of the Radiation Lab in general, that there was an amazing amount of interplay, and an amazing esprit de corps like I have never seen in any other organization before or since. I'm sure you've heard this from other people.

Work Environment and Administration

Aspray:

It's a comment I have heard, yes. How would you compare it, say, with a university laboratory environment?

Kyhl:

Well, that, of course, has also evolved during the period that I have watched it. Originally in the university environment there tended to be a professor working in a certain area. He had his students who worked under him and, depending on the professor, either at his direction or floundering on their own and getting help from the professor when they could. It depended on the professor whether he supervised closely or almost with neglect. This little group was an entity, and if there was another group in the department, they were a separate entity. Nowadays, of course, there's much more group effort because of the monstrous projects that people are involved in. Today as a graduate student you have a particular professor whom you work under. But you may go out to Kitt Peak for observations. Or you may go to Brookhaven or CERN or wherever. So, many of the projects require enormous numbers of people. One curious result is the published articles with many authors. It reached absurd proportions with the articles in the Phys. Rev. Letters with 175 authors. [Chuckling]

Aspray:

Could you comment on how information was exchanged within 52? Were there formal seminars? You've talked about the informality of people coming together. But were there formal seminars or regular meetings in addition to that informal exchange?

Kyhl:

I don't remember whether there were formal meetings, but any piece of information that was developed was spread to the group either by Collins saying, "Everybody get together, we've got something to talk about." Or perhaps there were regular weekly meetings. But that I don't remember. But the information was exchanged.

Aspray:

Did you have to write progress reports?

Kyhl:

I don't remember.

Aspray:

If you needed equipment or supplies or more personnel, did you have to put in a proposal, written proposal?

Kyhl:

I don't think so. There was remarkably little formal paperwork. There were occasionally administrative battles. We, of course, had a machine shop, and eventually we got to making our own magnetrons and had our own tube lab. There were administrative problems like the one involving the consolidation of the many individual group machine shops. They were going to combine these together into a Rad Lab machine shop under common direction. Some of the machine shops rebelled against this. They didn't want to be amalgamated and kept their doors locked, and we had a password to get in. [Laughter]

Engineers' Relationship with Physicists

Aspray:

In Henry Guerlac's account of the Rad Lab, he wrote that the lab was a physicists' world, run for and completely as possible by physicists. Were there any significant problems in integrating people with different kinds of disciplinary backgrounds into the research program, whether they were electrical engineers or scientists from other areas? Did you see any of that kind of problem?

Kyhl:

No, that's part of the miracle of Rad Lab.

Aspray:

With the group working on the magnetron, were there people with other disciplinary backgrounds besides physics?

Kyhl:

They were mostly physicists. I'm not sure about Louis Moose whether he was a physicist or an engineer.

Aspray:

He was an engineer.

Kyhl:

I knew him well, so I had a feeling he probably wasn't a physicist originally. But that made no difference in the group.

Aspray:

Not even in the sorts of projects they were assigned to work on?

Kyhl:

No, no, no.

Contributions of Theoretical Physicists

Aspray:

How significant was the theoretical work of John Slater to the ultimate development of the magnetron designs and to a comprehensive theory of magnetrons?

Kyhl:

Eventually profound. At the time, the experimental work by and large was groping, and the theory had not yet provided that much information. Later, toward the end of the war and after the war, the theory then coalesced and gave us a firm idea of things. But in the beginning the theory couldn't keep up with the trial-and-error approach. But Slater's contribution was a large one. The biggest contribution, of course, was by the British physicist, Douglas Hartree, of the Hartree-Fock equation. Hartree did theory for the British on the magnetron. He essentially did the basic fundamental theory. He is the father of magnetron theory. But Slater made very profound contributions.

Aspray:

Let me ask the same question with regard to Foster Rieke. What was his overall importance to both the practical work at the time and the theoretical work in the long term.

Kyhl:

Foster Rieke was a very good man. It's a little hard for me to separate out and say this man is responsible for this. Foster Rieke is credited with the Rieke diagram, which bears his name. He certainly deserves a great deal of credit for that. But that also was a group effort.

Aspray:

Are there any individuals you'd want to credit? I mean, not just individuals, but we've mentioned two names. Are there other names that you'd want to see represented in an historical account of credit for the work of the magnetron group?

Kyhl:

Oh, I don't know that I'd want to show any partiality.

Aspray:

That's fair enough. I don't want to push on that question.

Kyhl:

In addition there is a Professor Smullin [who] shows up here because when it got to the point of producing in quantity and worrying about lifetime and things like that, he designed and produced the magnetron test stand for doing life tests on magnetrons, and set up a whole room full of these things. So he had a role in this also, although most of Smullin's work was in other groups. I may have neglected some people entirely.

Aspray:

I know it can be invidious to start just naming names, because you're doing it on the spur of the moment, and it's hard to assign credit for group effort.

Kyhl:

What I should have done before I came here was go over the roster on the pictures of the magnetron group in the book and refresh myself on people's names, but I haven't had a chance to do that.

Practice and Theory

Aspray:

Fair enough. I think this will be taken in the right spirit. Maybe you can make a few comments, though, about what your trial-and-error method was like. Give me an example of what would happen on the day-to-day basis in the magnetron group. What kinds of problems you were addressing, and how you would go about them.

Kyhl:

That's a long time ago to remember that kind of detail except for those few incidents that stick in your memory. Certainly a fair amount of the work was just the instrumentation problem of connecting up the magnetrons, the pulsers, getting the magnets to run, providing the cooling water for everything. Tunable loads to provide loads for the magnetron. The circuitry for observing the pulse shapes, current shapes. All that sort of general electronics which is the everyday life of any experimenter, really. A lot of time goes into that. We were trying different kinds of designs. We asked, "What happens if we make the anode longer? What happens if we make it shorter? What happens if the cathode size is changed? How critical is it to have the cathode centered?" It seemed for a while that it ran better if the cathode was off center. You try things. You'd ask visitors who were working on magnetrons what they were doing and what they were finding out. Notably British visitors and Bell Labs, because our main competitor, or colleague, in the magnetron development was Bell Labs. They were making magnetrons in considerable quantities and doing research on them.

Aspray:

To what degree did theory inform your trial? For example, let me take an extreme: the story that's known about Robert Boyle in the 17th century was that in trying to understand vacuums, he'd look around his laboratory and say, "Oh, I haven't tried that. So let's see what happens if I put that into the vacuum." There was ostensible no theory informing his judgment in those sets of experiments. Clearly there was more going on at the Rad Lab, but can you talk about how theory informed your trials?

Kyhl:

Oh, that's pushing it far back. We did have some information, for example, the Hull theory of the cut-off magnetron and the Hartree line, which gave us some clues as to what voltages and dimensions we should be using. So we had a rough guide of this sort. But beyond that, an awful lot of the struggle had to do with the questions of moding and starting. The magnetron, as I said, has as many modes as there are cavities, and you want to have a situation where it starts in a mode you like. It has a tendency to start in a mode that doesn't couple to the output. [Chuckling] Although there is network theory of these problems, it was not that well developed. So an awful lot of it therefore was trying to make the magnetron behave itself, rather than going by the fundamental theory. It wasn't until a number of years had gone by that we discovered, for example, that the S-band (or the X-band) magnetrons we were building (I don't remember which.) were actually operating not in the fundamental wave mode of the electromagnetic field in the structure, but in a higher space harmonic. [Chuckling]

Aspray:

I see. This next question requires me to read you about a paragraph, so bear with me. "The Rad Lab volume, Microwave Magnetrons, edited by George Collins, devotes approximately ten chapters to theory and analysis, four chapters to design, and three chapters to 'practice.'" Some scholars assert that several bodies of knowledge may be distinguished, including basic science, engineering science and design. Do you agree or disagree that knowledge about magnetrons as developed at Rad Lab can approximately be placed in these three sorts of categories? If so, should the kind of practical knowledge or know-how included in the last three chapters of the book be distinguished as a fourth body of knowledge, one that might include details on fabrication methods such as hobbing and brazing?

Kyhl:

Oh, that's a tall order to discuss this late in the game. Certainly one cannot neglect the fabrication problem. In many ways, that's the thing in which the most time and effort is spent. What to make the cathode out of. How to keep the tubes from gassing up. These are extremely important things. The original development of the magnetron in Britain shouldn't have happened according to people who were playing with magnetrons in the early days, before the "British magnetron." They observed that you had strong back bombardment of electrons onto the cathode in the early magnetrons. So the cathode would melt. Then they made the cathodes out of tungsten wire. Even so, they would burn up. The people at Birmingham, whether from intelligence or ignorance, put a big oxide cathode in the magnetron. We all would have said, "That won't last. It'll burn up immediately." But it worked fine. It put out so many electrons that it provided a protective cloud against the back bombardment. There was some back bombardment. We had a magnetron that ran after half of the cathode was melted away, exposing the heater wires inside, but it still worked. A lot of things went on having to do with things really divorced from fundamental theory or basic design, but just from the practical aspects, which are very hard to predict. The magnetron is a strange beast. Its starting properties, for example: It builds up from nothing to peak power in an amazingly short number of cycles. I'm not sure that's even fully understood to this day, although I haven't been following the magnetron developments lately.

Collaboration with Bell Labs

Aspray:

You've mentioned Bell Labs. Could you comment on the significance of the work that was done there and at some of the other places?

Kyhl:

Later on, Columbia University developed the very high-frequency magnetrons. They were given that specific assignment and did a splendid job on that. I remember particularly Si Sonkin at Columbia, because later I worked with him at Stanford on linear accelerators. They did splendid work at Columbia.

Aspray:

Was there close interaction with the Rad Lab group?

Kyhl:

Yes, very close. That came along rather later, though, because we weren't prepared to do much on K-band [Chuckling] in the early days. Our interaction with Bell Labs was close.

Aspray:

Was close?

Kyhl:

We took magnetrons down to Bell Labs. We brought magnetrons back from Bell Labs to our lab. People went back and forth. I think of Jim Fisk in particular in that group at Bell Labs, who later became, I believe, their research director.

Aspray:

How would you characterize the difference of the research programs of the two institutions?

Kyhl:

At the technical level I don't know that there was that much difference. There was a difference in administration. Bell Labs tended to play things very close to their chests. They were secretive. Although if you could get together with the technical people, the interchange was fine. There was even a rule at Bell Labs that a technical letter could not be sent from an engineer at Bell Labs to an engineer at Rad Lab, but had to be approved by the executive office of Bell Labs, and then transmitted through DuBridge's office and then passed down. So the engineers would send us a preprint. On a few occasions they would say, "This letter was not approved [Chuckling] by channels. Please disregard what we said to you."

Aspray:

What about the objectives of the two organizations. Did they have a different agenda than you had?

Kyhl:

Not really. We were all trying to find out how to make good magnetrons with power and long life and few moding problems.

Aspray:

What about in terms of their method? Were they basing their approach on slightly different design principles than you were? Or was it simply parallel work that was going on?

Kyhl:

My recollection is that it was pretty parallel.

Aspray:

What about some of these other organizations, such as General Electric and Westinghouse? Did you have contact with them?

Kyhl:

I don't remember many contacts with General Electric and Westinghouse. I'm not saying there weren't any, but I don't recollect any at that time. We had a man named O'Kress with our group who was, I believe, from Westinghouse.

Test Equipment Group

Aspray:

Okay. Let's move on with your own career. What happened? Did you stay with the magnetron group throughout this period?

Kyhl:

I moved to the test equipment group, Group 55.

Aspray:

What was the objective of that group?

Kyhl:

The objective of the group was to provide equipment for measuring the performance of radars or other equipments that the Laboratory was producing or that other organizations were producing. This included equipment measuring transmitter power, receiver sensitivity, antenna pattern. Whatever you needed to measure, in the field. After the radars had been sent out into the services, you needed to have tests to make sure that they were operating properly. There's a very dangerous point here because a radar which is down many, many tens of d.b.'s in performance, can still see the ships in the harbor. It's because of the fourth power of distance-dependence that you can't tell just by looking at the scope and seeing targets that the performance is all right on the distant targets. You have to have more sophisticated test equipment, and that was the basic assignment of the test equipment group.

Aspray:

To use modern computer terminology, was this mainly a hardware problem, or was it a software problem also?

Kyhl:

Well, that was hardware. There was very little in the way of software development at that time.

Aspray:

I mean, it wasn't that you used existing test equipment, but that you used it in unusual ways, which I would regard as software. Was it that you had to design new equipment, or to design new ways of using it? I guess that's my question.

Kyhl:

All right. Some of it was fundamental design, new equipment. One obvious example is that if you want to measure the power going out of your antenna, you need to know something about the net power going to the antenna through the transmission line. For this purpose the directional coupler was developed. The test equipment group was active in directional coupler development. Another example was the development of the echo box. I don't know who actually came up with the ideas for these devices and whether it was at Rad Lab or at some other organization, but we worked a lot on the echo box. The echo box is a very high Q resonator, and you connect it to a pick-up antenna. It provides an overall performance test for your radar. You mount it on a ship's mast, for example, and if the ship's radar beam sweeps past the mast, it excites the echo box high Q cavity, which then rings and retransmits the signals back again to the radar, and you can look on the scope and see for how long a period on the trace you can detect the signal coming back from the echo box. This gives you a combination of the transmitter power and the receiver sensitivity. Because you must tune the echo box, it tells you the frequency you're on. It's a nice passive device for overall performance check.

Aspray:

I see. In terms of the commercial equipment that was available were they well-suited to the tasks at hand?

Kyhl:

People worked on improving scopes for us. We always needed faster scopes. The spectrum analyzer was being developed in those days, and it was extremely important for work on magnetrons. We, in fact, developed our own spectrum analyzers, and then other commercial people took them over. Those were important developments. They were engineering improvements on previous ideas, but they were very important for the development.

Aspray:

Why were you moved over to the test group?

Kyhl:

I'm not sure. I don't remember just what caused the move. I guess I was something of a microwave plumbing expert at that point, and perhaps test equipment needed some people in that area. But I don't really remember why I left the magnetron group.

Aspray:

Would that have been at your choice, or somebody else's? Or don't you remember that?

Kyhl:

I don't remember. I'm sure I would have had some input in the matter. I don't remember whether there was any conflict, or whether it seemed like a good idea or not.

Aspray:

I was asking about your particular case to get information about a more general phenomenon of how people were moved around within the Laboratory; if you had any general remarks to make about that, we'd welcome having them on tape.

Kyhl:

I don't really have many recollections on that. I'm sure there was some of the usual phenomenon, that if you developed something and it went out into the field, you may well follow along because you're the man who's best able to make it work in the field. I'm sure there was some of that going on. My own work was always in components at the bench. I never got out into systems development, and never out into the field. Many people stayed with one group. Of course as the needs changed, people shifted. If there were more people than were needed in one place, they'd go to somewhere else.

Aspray:

What did it take to be a good productive member of the test group?

Kyhl:

Microwave waveguide abilities on one side. Obviously circuit abilities on the other because much of the work was circuit design. Mostly experimental skill. You had to make equipment that would work, would work easily and reliably. Then all sorts of strange things. For example, the test equipment group went heavily into attenuator design because one of the ways of measuring performance is to have a variable attenuator between the transmitter and the detector, and measure power that way. We got into such strange things as, "How do you make attenuation in a waveguide?" We would be evaporating platinum films on a glass substrate, and things like that. So there were a lot of oddball techniques that got into this type of work. But basically it was microwave plumbing, waveguides, coaxial lines and circuitry. It was the circuitry of that time. There were no integrated circuits, and no printed circuits.

Aspray:

How large was the group, approximately?

Kyhl:

That was also a fairly large group. I can't give you a number. Frank Gaffney was the head of it, but under it there were a number of different subgroups. One of the outside people who helped us at that time was Prof. Ernst Weber of the Polytechnic Institute of Brooklyn (later an IEEE president, I believe).

Aspray:

What did you find your greatest challenge to be in that group?

Kyhl:

Well, I was developing microwave plumbing components. I did some work on high-precision impedance measurements, where the ordinary slotted lines weren't good enough. We soon developed other types of impedance measurements. This gets into the story of the magic-tee.

Magic-Tee

Aspray:

Okay. Why don't you tell me that.

Kyhl:

This story gets to be a little confusing. Bob Dicke is credited with the invention of the magic-tee, and that's correct. But there's some general confusion as to what you mean by a magic-tee. Many people, if you show them a waveguide structure of a straight waveguide and an E-plane and an H-plane branch coming off, they will call that a magic-tee. On the other hand, some people will say that it's not a magic-tee unless you have matching irises in the waveguides. I'm not sure what the official IEEE standards notation is on that. [Chuckling] But the structure, consisting of a waveguide with an E-plane and an H-plane side guide coming off of it, was invented independently three times, at least. The first time it was invented by Tyrrell at Bell Labs.

It was then invented by me a year or so later. Invention is perhaps the wrong word. Somebody brought into my office a strange device from Sperry Gyroscope on Long Island, which consisted of a fairly large ring of waveguides with some side guides coming off, some E-plane and some H-plane. Somebody said, "What's this thing? What's it good for?" I looked at it and realized that it had some symmetry properties and bridge-balance properties, but that was the wrong way to do it. I said: "Oh, but that isn't the thing! What you want to do is this. You want to take a waveguide that has these two things coming off the side. Then you have a nice object, and you can make an impedance bridge out of it." Then we started making impedance bridges in the test equipment group, using this new structure.

After we'd done this for a little while, Bob Dicke, whom I had never met, walked into my office and said, "I hear that you've invented the magic-tee." I said, "What?!!" He pointed to the thing on my desk. I said, "Oh, that." He had invented the magic-tee a short while after I had invented it, and then somebody had said, "Oh, but they're making these things over in the test equipment group." But neither Tyrrell nor I had put in the matching irises. Bob Dicke had realized that with the matching irises, you came up with the full beautiful symmetry of the scattering matrix. So that's the strict terminology for the magic-tee. The ones we had been using in impedance bridges, however, didn't have the matching irises. They were just a balanced bridge arrangement. Only after we were building impedance bridges did we find that in the Rad Lab document office was a report from Tyrrell describing a bridge circuit for microwaves. The document room had filed it under circuits instead of under microwaves. So none of us had ever seen it.

Aspray:

I see. [Chuckling]

Kyhl:

That was the way I first met Bob Dicke. Then we struck up a warm friendship. Later on when he was recruiting a crew to go to Florida for his radiometer experiments, he invited me to join that group. But that was toward the end of the war.

Aspray:

Is that the next stage in your career at the Laboratory?

Kyhl:

That was just a brief period at the end. My stay was divided basically between the magnetron group and the test equipment group. There were a couple of other minor things. At the end I got involved a little bit in lighthouse tube oscillators because they had a problem there that they needed help on. But I spent most of my time in those two groups. Then at the end joined Dicke's group for the trip to Florida.

W.C. Brown

Aspray:

Let me give you an opportunity to talk about some of the things we haven't already mentioned.

Kyhl:

Yes. I don't know if these things are interesting or not. I should mention W.C. Brown who joined Percy Spencer when they expanded their magnetron facilities at Raytheon. He played an important part in developing magnetrons out there. I might also mention that Percy Spencer claimed the magnetron patent for some work he had done long before the war, in which he had tried to improve the performance of a thyratron, which is a gas-filled switch tube, by applying a magnetic field to it. I don't know how his experiments came out, but he filed a patent, and he claimed that that patent covered structures which contained a cathode, an anode, and an applied magnetic field. [Laughter] Nothing ever came of that, but it's an amusing bit. : Another anecdote is on the Bethe-hole directional coupler. This was invented by Hans Bethe. One of the standard designs for directional couplers is the so-called quarter-wave coupler. Two probes probe the main transmission line. They're separated by a quarter wave. You recombine them in the second line, and you get phase interference so that you only get a signal in one direction. That tells you the strength of the wave moving in that direction, which is what the directional coupler is designed to do.

They had built these with two waveguides side by side with holes in the narrow side of the waveguide, and they had worked this way. Then they'd tried putting holes in the broad side of the waveguide and putting them together. They tried that, and they didn't get any coupling at all. [Chuckling] They were scratching their heads. We all went down to our regular seminar on microwave circuitry in waveguides, and Hans Bethe gave us a talk on the couplings through a single hole, in which, he pointed out, that under proper dimensions the coupling was basically in the backward direction. So, if you had two holes, you got no coupling in the forward direction because of the Bethe principle, and you got no coupling in the backward direction because of the quarter wave. [Chuckling] So you got no coupling at all. They heard what Bethe had to say at the seminar, and they ran upstairs and covered up one of the holes [Laughter] and had a Bethe-hole coupler. There was one example of the theoretician beating the experimentalists to the draw.

I'd like to comment on a totally different subject: the pulsers for driving the oscillators, the design of pulsers, and in particular the universal use of the line pulser. When this problem was first presented the question was how to make a high-powered device that produces a short square voltage pulse with power behind it. Ernie Guillemin who essentially founded the science or art of sophisticated network theory teaching at MIT, [didn't found] network theory, but he founded the teaching of sophisticated electrical principles to the engineering students. He was a network theorist.

He took the job and did it the way a network theorist would do it. That is to say, he "Fourier-analyzed" the square pulse. He found where the poles were, he built a resonant circuit for each pole and connected them together into what was then called a Guillemin line. You could make pulses this way, but this was the hard way to play the game. Each circuit was tuned to a different frequency, and the couplings spoiled everything. It was a very critical thing, a very hard game to play. Instead, if you simply built an artificial transmission line, charged it up, and had the switch at one end, you'd have the wave traveling down the line and back, and you'd get your pulse. It's a beautiful idea, but Ernie Guillemin missed it. That was a real invention. I forget who was credited with that invention. It may be Mort Kanner from our pulser group; I'm not sure. It was called modulator group rather than pulser group, to throw spies off. [Chuckling]

W.W. Hansen

Kyhl:

I might mention the role of W.W. Hansen at Radiation Laboratory. Hansen sort of created the role of microwave electronics for us by a series of lectures that he gave at MIT. He also gave, I think, the same lectures at Sperry Gyroscope on Long Island because he had contacts with Sperry for contracting or consulting. So he gave us these talks, and we all attended, and this is where we learned our sophisticated microwaves. He didn't give out notes, but somebody sat in the front row and took it all down, and published in mimeograph what were called Hansen's Notes. This became an early bible. I'm not sure who did that. It may have been Dave Saxon. [pause] Hansen did theory the way he had played football at Stanford as a student. He had been a lineman, and he did theory by lowering his head and plowing through. [Chuckling] And he could plow through and come out the other end. This was entirely different from Slater's style, but he was a very great man. Later on, I went to Stanford and worked for Hansen. : An amusing point [concerning] the klystron, which was the low-power tube used for local oscillators throughout the war, invented by the Varians and Hansen (the Varians had had the patent for it). The motivation for inventing the klystron was to provide high microwave power because Hansen wanted to build a linear accelerator. The first ones that came out were low power and were very useful. But the original incentive was to power an accelerator.

Aspray:

I see.

Kyhl:

I've also heard a rumor, and I can't verify it, that the original motivation for the invention of the high-power magnetron in Birmingham, England, was to get power to run an atom-smasher. [Chuckling] So the original motivations on these things are sometimes very different from their practical uses.

Operettas

Kyhl:

I think that may cover most of what I wanted to say. I should mention — because I think it tends to be neglected — the two operettas which were produced at the Radiation Laboratory.

Aspray:

Oh?!

Kyhl:

Somehow, when you get a group of people working together well and efficiently (and of course we were very busy and working long hours because of the pressures and needs of the war crisis), people still find time to do things like putting on two original operettas with scenery which was manufactured in the Rad Lab shops and directed semi-professionally. I think it was put on at the Charles Playhouse in Boston to packed houses. The music was written by Art Roberts, who was a nuclear physicist, and the librettos by Kay Bolt, the wife of Richard Bolt, now of Bolt, Beranek and Newman. There's one little picture, I think, in the five-year book about that. It just shows the kinds of things that people can do when things are working effectively. [Chuckling]

Collaboration with Military

Aspray:

I have a couple of questions more for you. Can you comment at all on the interactions of Rad Lab with the military?

Kyhl:

Most of that interaction would have been at a higher level than I would have been involved in. I was not involved at decision-making levels or interacting contractually with other people. I wouldn't know anything about that. We had no problems with the interactions down at the technical level. We had the training program in the Harbor Building run by MIT, where the services had people to be trained as radar operators for the fleet.

Aspray:

But it seems to me there was work going on at one of the Army Signal Corps facilities. Was it Fort Monmouth? There was also work at the Naval Research Laboratory. Was there technical contact between your group and theirs?

Kyhl:

In magnetrons, I don't recall much. In the test equipment group, there was a lot of contact because many of the pieces of test equipment were designed for specific radars in the field. The services people were worried about problems like, "Will this still work with salt spray and vibration?" So in many cases they were bearing down on us to make sure that our products would meet those sorts of specs. The test equipment group was developing the mock-up, and somebody else would produce the production models. There were those kind of problems. But I don't recall any strains or anything like that. Everybody was working together.

Aspray:

I wasn't particularly looking for that. I was just trying to find out what the interchange was.

Kyhl:

There was fairly close interchange.

Postwar Career

Aspray:

How did your Rad Lab experience affect your own subsequent career?

Kyhl:

It changed me completely, of course. I was a real microwave plumber by the end of the war. I've remained essentially a microwave plumber, throughout the rest of my career. I did different things, but always there was a microwave component to it somehow. So I really grew up there.

Aspray:

When and how did you leave the Laboratory?

Kyhl:

I stayed there until the end, as many people did. Then, although I had been working at the University of Chicago on microwaves before the war, there was no point in going back there anymore because what had happened to microwaves had completely changed the situation. I stayed at MIT and picked up a Ph.D. there. Then I went out to Stanford to work for Bill Hansen on linear accelerators. Later I came back and did some work at GE Research Lab in Schenectady. I then came back to MIT and rejoined the Research Lab of Electronics, and ended up at MIT teaching in the electrical engineering department.

Aspray:

Okay. Are there any other points you want to add or points to emphasize?

Kyhl:

One anecdote. I don't know if it's worth telling. At the end of the war, Rad Lab realized that many of these people would want to be going back into physics again, since many of them were physicists. They thought that they should help get them back on line again. They got Wolfgang Pauli to come and give us a lecture in the main MIT lecture hall on modern quantum mechanics. A couple of weeks later, they got Julian Schwinger to come and tell us what Pauli had said. [Laughter] A few weeks after that, they had Ed Purcell come and tell us what Schwinger had said. [Laughter]

Aspray:

I see.

Kyhl:

The main thing I think that I would say is what I've already said, is that the spirit of cooperation, and the camaraderie, and the effectiveness of working together was amazing. If somebody was designing a system, and they needed a receiver for that system or they needed a transmitter, they'd go to the components groups, who would break their backs to get just what they needed and get it fast, and get it to them.

Aspray:

I see.

Kyhl:

It was an amazing experience. I'm sure you have heard some of the same things from other interviewees. I've also worked with other large groups since, such as the Stanford linear accelerator people, and there you also have this kind of cooperation. But I've never seen it to the degree that I saw it at the Rad Lab.

Aspray:

I see. Well, thank you very much.

Kyhl:

Oh, you're very welcome.