Oral-History:Warren P. Mason
About Warren P. Mason
Warren Mason received his bachelor of science from the University of Kansas in 1921. Upon graduating, he took a position with the Western Electric Company. While employed by Western Electric (later Bell Telephone Laboratories), he completed work on both a masters and Ph.D. at Columbia University. His early work involved carrier research as well as quartz crystal research. Mason also worked with ferroelectric crystals, and after WWII, his work was transferred to Shockley's solid-state division, where he focused on piezoelectric crystals and dielectric properties. From 1948, until his retirement from Bell in 1965, he headed the Mechanics Research Department, where he was involved in investigating mechanical properties of a vast array of materials and structures as they applied to Bell System uses. Upon retiring from Bell, Mason held a visiting professor appointment at Columbia University, and was a Senior Research Associate at the Henry Crumb School of Mines. In addition, he served as a consultant to Bell Labs for two years in their development of monolithic crystal filters.
The interview begins with Mason's educational background and a discussion of his early work on carrier current transmission for Western Electric, including his work on the carrier system for coaxial cable. The interview continues with a lengthy and detailed discussion of his extensive work in crystal research, including his work with ferroelectric and ADP crystals. During WWII, Mason was involved with the development of delay lines for use in moving target indicators for the MIT radar systems. The interview covers his crystal work in Bell's solid-state division under Bill Shockley, including research on electrostrictive materials. The interview then shifts to Mason's work while head of Bell's Mechanics Research Department, a position he held from 1948 until his retirement in 1965. Mason discusses his use of torsional vibrating quartz crystals for the measurement of the shear properties of liquids, including polymers; his work with Minlen of Columbia on tangential force and shearing stress; his efforts in the field of wire spring relays; and his studies of fatigue in metals. This work carried over into other fields, such as bonding wires to silicon wafers and addressing the problem of decreased attenuation in underwater cables. The interview continues with comments on Mason's work on magnetostrictive phenomenon, sound attenuation in metals, and the stress-strain linearity of crystals. Mason's concluding remarks concern his publishing and patent achievements while at Bell, and his post-retirement work for Columbia University and the Henry Crumb School of Mining.
About the Interview
Warren P. Mason: An Interview Conducted by Frank A. Polkinghorn, IEEE History Center, March 3, 1973
Interview # 005 for the IEEE History Center, The Institute of Electrical and Electronics Engineers, Inc. and Rutgers, The State University of New Jersey
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, Rutgers - the State University, 39 Union Street, New Brunswick, NJ 08901-8538 USA. 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:
Warren P. Mason, an oral history conducted in 1973 by Frank A. Polkinghorn, IEEE History Center, Rutgers University, New Brunswick, NJ, USA.
Interview
Interview: Warren P. Mason Interviewer: Frank A. Polkinghorn Date: March 3, 1973
Western Electric
Polkinghorn:
This is an interview with Dr. Warren P. Mason who did physical research and development at the Western Electric Company and its offshoot, Bell Telephone Laboratories, from 1921 until his retirement in 1965. According to my information, you were born in Colorado and received a Bachelor of Science degree from the University of Kansas in 1921. Did you go directly to work in the Western Electric Company then?
Mason:
That's right. I came directly there. I was fortunate enough to be put in the research department in the division working on carrier current transmission. Previous to that, all long distance telephone had been carried out in voice frequencies. Radio communication at higher frequencies was being used. The object of the carrier work under the direction of J.W. Horton was to apply radio frequency communication to telephone long-distance communication. We were working on a three-channel system. Vacuum tubes were used to modulate the voice channel up to the higher frequencies. Coil and condenser wave filters, which had been invented by G. A. Campbell, were used to separate out the channels. I was in this department from 1921 to 1925.
Polkinghorn:
I presume you remained with this work after the Western Electric Company was changed over to Bell Telephone Laboratories in 1925.
Mason:
Yes, I did. I was changed over to the filter work with K. S. Johnson. He was carrying out work supplementary to that of Campbell and deriving equations for all types of filters and their design formulae.
Columbia University
Polkinghorn:
Did you continue with your education on a part-time basis?
Mason:
Yes, while with the carrier research group I worked in the physics department at Columbia University part-time. This work resulted in a master's degree in 1924 and a Ph.D. in 1928.
Polkinghorn:
What was the subject of your thesis?
Mason:
Since my work at Bell Laboratories had to do with filters, I chose an investigation of acoustic filters in the propagation of sound waves and tubes. This was accepted in 1927, and I won my degree in 1928. This was the first work, which applied distributive constant networks to filters.
Coaxial Cables and Crystal Research
Polkinghorn:
I believe you then did some work on the carrier system for coaxial cable?
Mason:
That's right. At that time a new mode of carrier transmission was devised, the coaxial cable — through the work of Herman Affle and Lloyd Espenschied. This had a much wider frequency range than the open wire transmission line previously used. The top frequency of the first ones were on the order of a megahertz. But with the coil and condenser filters then available, the Q- or quality factor of the coils was low enough to produce a high attenuation in the band, and a considerable loss in frequency space needed to separate adjacent channels for all the frequencies that kept very low ones. This used the full frequency range of the coaxial conductor. It was necessary to use a low frequency filter and modulate it up to the proper frequency by a series of modulation steps, or else use some elements with a higher cue to obtain a series of band pass filters with low-loss pass bands and small loss in frequency space. The Germans and the Japanese used the first alternative, but the Bell System used the second alternative — with quartz crystals replacing coils and condensors. The first work on quartz crystals was done by Professor W. A. Cady and Carl Van Dyke of Wesleyan University and Professor G. W. Pierce of Harvard, who showed that they had very high Q’s but only over a very narrow frequency range. The problem that I worked on was how to utilize a high Q and produce a bandwidth large enough for a telephone channel. I showed that if crystals were incorporated into lattice-type filters with coils on the end, then wide bands up to ten percent could be constructed with an attenuation due to the low cue in the coils, which was a constant independent of the frequency. It also showed that, by dividing the electrodes on the surface, the crystal could appear on both sides of the lattice structure, thus cutting the number of crystals in half. A two-section filter used in the coaxial conductor of microwave radio in undersea cable system requires four quartz crystals, three electrical coils, and several condensors to adjust the bandwidth and attenuation peak. As many as five hundred thousand filters a year and two million crystals have been used to satisfy these requirements. This type of filter has been used from its conception in 1935 to date. It is only now being replaced by another type of crystal filter, the monolithic filter, in which all the sections are formed on monocrystal plates by evaporating electrode layers of controlled thickness onto a single crystal plate. On the account of a lower cost and a decrease in the amount of material, this type is taking the place of the original coil condenser and quartz crystal filter.
Polkinghorn:
My earliest recollection of you was when you were working with F. R. Lack although I think I knew you earlier.
Mason:
As a result of this filter work, and due to the promotion of F. R. Lack to head of the vacuum tube department, I was made head of the crystal research department, which is a branch of the Radio Research Department under Ralph Baum. We worked on many quartz crystals, such as the AT., GT., and X-cut crystal, all of which were used in filters and in the control of oscillator frequencies for use in radio systems. Even without temperature control, the frequency of an oscillator could be controlled to one part in a million. This corresponds to thirty seconds a year. But with a watch driven by crystal control and mounted on the wrist, a rudimentary temperature control, accuracies in the order of one second a year were claimed. With good temperature control, frequencies on the order of a few parts in ten to the tenth power are realized. The only superior clocks are the atomic clocks but these need quartz crystal oscillators to interpolate to the desired frequency.
Polkinghorn:
I recalled that you worked on filters with coaxial elements about this time.
Mason:
That's right. At the request of the radio department of which Frank Polkinghorn here was a member, another study resulting from this early work on acoustic filters was the application of coaxial conductors and wave guides to the construction of wave filters. These elements have distributor constants and, as in the acoustic filters, it was necessary to consider wave propagation in the elements in calculating the wave correlated characteristics. An early application was the use of such a filter in the Green Harbor-Provincetown long-wave radio link.
Polkinghorn:
You also worked on other types of crystals about that time?
Mason:
Yes, we studied theoretically and experimentally ferro-electric crystals. These are crystals that have a spontaneous electrical polarization analogous to spontaneous magnetic polarization occurring in such ferro-magnetic materials as iron and nickel. Rossell salt was the first such material. I extended the investigations of Professor J. Vallesec, who was the first one to investigate their mechanical properties as well as to their dialectric properties and who produced a theory connected with the motion in hydrogen bombs which explains both the lower and upper cut off peaks, that is the temperatures that separate the polarized from the unpolarized region. This theory is accepted today as the most likely explanation for the properties of ______salt.
World War II
Polkinghorn:
What work did you do during World War II?
Mason:
I continued with other magnetic ferroelectric crystals such as KDP and ADP. KDP was a ferroelectric, but ADP was an anti-ferroelectric. It was more useful since it had a higher electromechanical coupling constant. It was used in underwater sound transducers during World War II. During the war work in 1941, work shifted from research on crystals and mechanical systems, principally to work on the ADP crystals for sonar transducers. The department was shifted over to the chemical research group, who were at that time growing the crystals. At the same time the demand for quartz crystals was increased very rapidly, as 100 separate crystals were required for each tank. Three of the men in the department were lent to the Western Electric Company at the Hawthorne division in Chicago. They instructed the manufacturing division on how to manufacture these crystals, what tests were to be made, and how the crystals should be used in practice. The Western Electric Company produced over ten million crystals for this work. The rest of us worked on underwater sound transducers and radar systems. In addition to checking the performance of the ADP crystals, I worked on a design of the transducer, which resulted in the OJA, that's a code name, transducer that was used in submarine warfare. Another use for ADP transducers was as the ears in acoustic torpedoes. In collaboration with the apparatus development department headed by A. C. Keller, these transducers were developed and adapted by the U.S. Navy. Another type of war work was delay lines for use in moving target indicators for the MIT radar systems. In this system, the first pulse returned from the surrounding terrain was sent through a delay line having the same delay as the delay time between pulses. Phase of the delay line was reversed, and stationary targets were cancelled out to a residual 40db. Any moving target did not give the same echo from frame to frame and hence would show up on the radar scope. The original delay line was a mercury delay line, which was not too useful because it would corrode, so our group designed a multi-path fused quartz delay line that was much more satisfactory. This was used extensively in the war.
Gun Silencer
Polkinghorn:
Did you also work on a gun silencer?
Mason:
Yes. Another project I got on account of my work on acoustic filters was the design of silenced pistols for immobilizing sentries. The final design employed cooling of the gases as well as filtering. The reduction of sound output was in the order of 40db, and the design was adopted by the U.S. Army. It was also applied to forty-five caliber machine guns. During the war period, there was only one paper produced, but I got twenty patents on various things.
Solid-State Division
Polkinghorn:
What did you do after World War II?
Mason:
After the war it was felt that our work on crystals was more closely related to solid-state physics, and the group was transferred into physical research in the solid-state division headed by William Shockley and Stanley Morgan. Our work was connected with piezoelectric crystals and dielectric properties. One of the first problems arose because the supply of quartz was limited. Although the chemists were studying methods for growing quartz, they were not near enough to success to expect to solve this problem in time to replace Brazilian quartz. Hence work was started on a water-soluble crystal. The chemists succeeded in producing a number of crystals. We measured the properties of these, and two monocloic crystals, that is crystals with low symmetry, were found which had zero temperature coefficient. Most promising was ethelyne diomene titrate with the trade name EDT. This was produced and developed by the apparatus development department. Although the crystal produced usable filters, it was abandoned when quartz crystal growing was accomplished. Other work in this department had to do with the electrostrictive effect in borium titenate ceramics.
This material was first shown to be a ferroelectric material by Professor von Hippel of MIT in 1945. We found that by polarizing the ceramic it could be made to act as a piezoelectric crystal. This is the analog of the magnetostrictive material and hence has been given electrostriction. The advantage of the electrostrivtive material is that the displacement can be much larger than that for a magnetostrictive material, and the ferroelectric type ceramics have become the most-used transducer materials for all frequencies 60 Hertz, that is the electrical power frequencies. My part in developing this material was in deriving the theoretical equations governing the actions of these electrostrictive materials. I also showed that, by introducing lead titenate and calcium titenate, the Curie temperature could be raised while the second transition temperature could be lowered, which resulted in a smoother temperature characteristic and a high coercive force. This allowed higher voltages to be applied to the ceramic with a higher power output resulting. During my stay in the solid-state division between 1945 and 1948, I published 17 papers and obtained 21 patents. I also started my second book, Piezo-Electric Crystals and Their Applications in Ultrasonics, which was published in 1950 by Van Nostrand.
Mechanics Research
Polkinghorn:
What did you do after you left the solid-state physics department?
Mason:
In 1948 I was raised in position to department head and put in charge of a new department called Mechanics Research. Some of the work on Piezoelectric crystals, particularly the work on transducers and barium titenate ceramics, remained, but the principal objective was to investigate the mechanical properties of materials and structures as they applied to Bell System uses.
Polkinghorn:
You also studied sound transmission in metals, I believe.
Mason:
Yes, I did. This was a study on the effect of metal grain sizes on the transmission of sound waves. The scattering of sound waves was found to vary as the fourth power of the frequency when the wave length approached the grain size. The result was of interest for metal delay lines used in storing information for radar systems or telephone switching systems.
Polkinghorn:
What did you do after that?
Mason:
One of the principal things that we did was to use torsional vibrating quartz crystals for the measurement of shear viscosity and elasticity of liquids. This was probably the first instrument to demonstrate that viscous liquids, such as castor oil, had a shear stiffness as well as a shear viscosity. This work attracted the attention of the polymer chemist W.O. Baker, who was president of Bell Laboratories, and a joint program was set up for measuring the shear properties of polymer liquids and solutions. First, measurements were made of the viscosity and elasticity of a solution and also polymer liquids — as a function of frequency. A very high frequency device was constructed by Ace J. Miskimman of my department, and with these two methods we were able to cover the entire frequency and temperature ranges. By analyzing the results, the various types of motion possible for a polymer chain could be determined. This type of work was extended to solid polymers and was useful in determining the mechanical properties of polymer insulators, plastics, and polymer materials used in the telephone system.
Polkinghorn:
About this time you did some work with Professor Minlen of Columbia?
Mason:
Yes, we did. Minlen had derived theoretical formula on the effect of tangential forces on the wear of contact surfaces of elastic spheres. He developed the formula and showed that if two spherical surfaces were pressed together with a normal force and then a tangential force was applied, the equation showed that the shearing stress reached infinity at the edge of the circle of contact. Since infinite stresses are not possible, he made the assumption that sliding would occur in a circular ring until the shearing stress was equal to the product of normal stress times the coefficient of friction.
Spring Relays
Polkinghorn:
Did you work on Sarnoff's contacts about that time?
Mason:
That's right. First, however, we did some work on wire spring relays at the request of the apparatus development department. They had the job of devising a relay that had forty contacts and was supposed to last forty years, which is about a billion operations. The first ones that they devised didn't do this. After about a million operations there was so much wear on the contacts that they ceased to mate at the proper time. We used our barium titenate high-amptitude cylinder to study this effect. This would go at the rate of twenty thousand cycles a minute, and you could get amplitudes as high as ______[]. By using a normal force, we showed that there was no wear there; it was all due to tangential force. We also showed that some of the plastics would produce a much better wear system than the metals or glasses or anything else. Also, if you were able to cut the tangential motion down to a very small amount, say two-tenths of a million, there wouldn't be any appreciable wear. By using the right plastics and by putting a double bend in the wires of the relay, we were able to pull the tangential motion under the normal amount, and the design objective of one billion operations was met. This relay has been used in all the mechanical switching systems of the Bell System.
Bonding Wires to Silicon Wafers
Polkinghorn:
You carried some of that work over into bonding wires to silicon wafers, did you not?
Mason:
Yes, I did that. I might also say that our high-amplitude device was used in studying fatigue in metals. Previously it had been studied by bending things at the rate of twenty cycles a second, but we could speed this up a thousand times and get a complete fatigue run in the order of a half day when previously it would take six months to determine the fatigue property. In regard to the use of bonding, another investigation was under taken at the request of the switching apparatus department, which was to determine what held the type of connection called the "solderless wrapped connection" together — to obtain some indication of whether it would hold together for the design objective of forty years. This connection was made by wrapping a copper wire around a rectangular binding post which had square corners, under tension produced by the wrapping tool. When the tool was taken off, the wire gripped the terminal and made a good contact. However, it was well known that stress will relax in a period of time particularly at a high temperature, and we were given the job to find out whether it was likely that this connection would hold together for forty years. The stress relaxation was investigated by winding the wire on a spring whose angular displacement was determined by the tension of the wire. By observing the angular displacement as a function of time and temperature, we found that at least half the tension would be present at the end of forty years. Furthermore, by observing the force required to strip the wire off the terminal, it was found that there was a solid-state join between the wire and the terminal which produced a thermal compression bond. Hence it was concluded that the terminal was safe. There are now over a billion terminals a year made in this way by the Bell System and another billion produced by IBM.
Mason:
What you learned there carried over to some other fields, did it not?
Polkinghorn:
That's true. Due originally to this work, two people, O. L. Anderson in my department and C. J. Christenson in the transistor department, studied the joining of metal wires to silicon wafers used in transistors. By pressing gold wire onto the transistor at high pressures and temperatures, we showed that a good bond was made between the wire and the silicon wafer. This process has become a standard operation in producing transistors. Furthermore it is used in sealing transistor cans without using the high heat necessary for welding.
Transatlantic Cable
Polkinghorn:
What other phenomena were you interested in at that time?
Mason:
This compressing band also supplied the answer to another problem plaguing the underwater cables. The first transatlantic cable with the repeaters had an acceptable transmission characteristic when it was made, but it was found that as a function of time the attenuation decreased, and this played havoc with the gain adjustment. While there were some adjustable features, the question arose as to whether the cable would have to be abandoned. Our department was called in to investigate this phenomenon, which we found to be due to the fact that the outer conductor was made of copper pieces laid in the form of a helix around the outside insulation. This was a method that had been used in constructing telegraph cables and was also used in the New York-Havana telephone cable. When the cable was made, there was a fair insulation between adjacent members of the helix. As a result the current path was larger in the outside layers, and this added about 100db to the total attenuation of about 1100db. At the large hydrostatic pressure at the sea bottom, the helical copper pieces were pressed together and a join between them occurred as a function of time. They tended to become a complete layer with a reduced attentuation. The solution to this problem was to do away with the helical strips and make the outer conductor a complete concentric layer. This has been done in later cables and no aging is observed.
Magnetostrictive Phenomenon
Polkinghorn:
You did some work on magnetostrictive phenomenon about that time, too?
Mason:
We investigated the mechanical properties of nickel in cooperation with a magnetics group headed by Bozar. We found that due to the motion of the domain wall in nickel a drop in the elastic constant occurred. This is known as a "Del E. effect" [?] and increased the attenuation. It was shown for the first time that this was a relaxation process, which reached the maximum at a critical frequency that depends on the size of the magnetic domain. At higher frequencies, the effect disappears.
Sound Attenuation in Metals
Polkinghorn:
Didn't you do work on the sound attenuation in metals, also?
Mason:
Yes, we studied the attenuation of sound in superconductors. In a normal conductor, as you go down in temperature, the attenuation increases very markedly due to the driving of the electrons by the sound waves. This is larger at low temperatures because you have a much longer path length for the electron. In the superconductor however, the electrons become paired, and this type of attenuation disappears. These measurements were the first to show this effect, and they supplied evidence in agreement with the Bardeen-Cooper-Shrinker theory, for which the latest Nobel Prize has been given.
Polkinghorn:
I understand you developed a strain gauge as a result of some of this work?
Mason:
Yes, we studied the use of the piezo-resistance effect in silicon, that is the change in resistance as you apply a strain to it. This is of course the same thing that occurs in an ordinary wire strain gauge, but the effect is much larger — on the order of a hundred times. So you could get a much more sensitive strain gauged element which could do away with the amplifiers needed with the ordinary strain gauge. This was so successful that I was awarded the Arnold L. Beckman Award of the Instrument Society of America in 1964.
Stress-Strain Linearity of Crystals
=
Polkinghorn:
You were also studying the stress-strain linearity of crystals?
Mason:
Yes. If you strain crystals hard enough, they become non-linear in their stress-strain relationship. This is known as a third-order moduli, and it is important because it's what causes the combination of the effect of photons on the attenuation of metals. A theory was evolved which showed that the attenuation in perfect crystals, that is those free from such defects as dislocations, was related to size and locations of these moduli. This has been applied to a number of crystals for which third-order moduli have been measured with good results.
Polkinghorn:
You mentioned earlier the effect of doping on acoustic attenuation in crystals. Did this have an application in solid-state devices?
Mason:
The final work of the mechanics department, when I was there, had to do with the attentuation of sound waves in silicon and germanium at low temperatures. It was shown that a very large attenuation could occur, which is a function of the doping, that is the amount of impurities you added which produced electrons and holes in these semiconductors. As you went down in temperature, you got an attenuation that was many times larger than you would get at room temperature. This work provided the basis for determining the scattering mechanisms present and relaxation times associated with the scattering.
Polkinghorn:
How long did you continue with this type of work?
Mason:
This was about at the end of my stay at Bell Laboratories, which ended in 1965. In the seventeen years I headed the department, I produced 89 papers, 111 patents and a book entitled Physical Acoustics and the Properties of Solids. I was also the editor of a series of books entitled Physical Acoustics published by the Academic Press. I've kept this work up since retiring, and there are now a total of thirteen volumes with more to come.
Return to Columbia University
Polkinghorn:
After you retired you went to work with Columbia University?
Mason:
Yes, after retiring from Bell Laboratories in 1965, I became a visiting professor at Columbia University in the department of Civil Engineering and Engineering Mechanics. I taught one seminar called Crystal Physics, for which I wrote a new book entitled Crystal Physics of Interaction Processes. I spent most of my time in the Institute of Fatigue and Reliability, studying fatigue mechanisms occurring in different materials. Also, considerable work was done in transmission of sound waves in different metal crystals, and in relating these to dislocation, motion and damping.
Polkinghorn:
You went to work for the Crumb School of Mines, I believe?
Mason:
Yes. In 1969 the Institute of Fatigue and Reliability moved to George Washington University in Washington, D.C., where they had a better financial arrangement. Since I did not wish to move to Washington, I compromised by spending about four days a month down there as an advisor with a young Ph.D. to do the experimental work. Some work was done with Professor W. A. Wood, who moved down from Columbia, and several papers were written on the application of ultrasonic methods to study of fatigue mechanisms. This work terminated in 1972.
In the meantime, another branch of Columbia, the Henry Crumb School of Mines decided to hire me as a consultant with the title of Senior Research Associate. Their interest was in accounting for the attentuation of sound waves in rocks and pure metals, and to study acoustic emission, that is the noise generated in metal or rock when it is strained. By measuring the attenuation of three rocks I found that there were two regimes for dislocation motion: the long-frequency one producing an attenuation directly proportional to frequency, while the high-frequency mode produced an increase at first and then a lowering of attenuation inversely proportional to frequency. This last term is similar to that found for other metals at high-frequency, but the low-frequency mode is new and accounts for the attenuation measured in seismic waves. This work has been covered by three papers presented in The Journal of Applied Physics, The Journal of Geophysical Research, and Nature. Acoustic emission work is just starting, but already the sound pickup corresponds to several well-known dislocation events.
Retirement and Patents
Polkinghorn:
Have you done other work since you've retired?
Mason:
Yes. After retiring I became a consultant for two years to the Allentown branch of Bell Laboratories on the theory and development of their monolithic crystal filter, in which all eight sections of the filter are obtained on a single crystal plate by evaporating a series of eight electrodes of controlled size and thickness. This construction, which has lowered the cost of producing crystal filters, is in the process of replacing coil and condenser and crystal types in all microwave carrier and underseas cables of the Bell System. As a consultant I produced two papers dealing with the monolithic time transducer and the method for calculating the attenuation of the structure. Also three patents resulted from this work.
Polkinghorn:
I have heard it said that you obtained more patents while at BTL than any other person. How many did you get?
Mason:
Well, counting the last three, I have 216 patents.
Polkinghorn:
Thank you very much for being interviewed.
