HD-1955NV

North American Aviation – Navaho to Minuteman – 1955 - 1958

Process Engineer with GM KCKS to Test Engineer with North American Aviation Downey CA

The day I reported to George Keller at the Downey CA North American Aviation plant, he took me to meet his boss Don Williams and Williams assistant Elliott Buxton. {Don Williams went on to become head of Autonetics division of NAA which during that time became part of Rockwell International. Don helped start the Micro-Electronics division of Autonetics, but was eased out due to the Rockwell buyout, status unk.} George was a Group leader and assigned me to work for Paris Stafford to set up a High Temp test facility for the Navaho missile series. {Paris Stafford was a highly regarded engineering supervisor. Paris died of a heart attack about 1970}. A WWII bomb shelter had been hauled up Lakewood Blvd from the Douglas plant and test cells added to one side, it was an empty shell when I first saw it. We did work there on the B-70, as well as multi versions of Navaho. At that time Tom Schuler was a “knob twister” in the simulation lab. {Tom Shuler went on to become Chief Engineer for the Minuteman program. I was told by Paris Stafford that Tom was selected because he’d received the highest grades in a survey exam. Tom died 2001}. When the Navaho program crashed they laid off 4000 people in one day. I was shocked to find I was still there, kept because they intended to make use of the High Temp “Bomb Shelter” facility for follow on work. I was the only one of 8 left at the facility. Fortunately a short time later all the technicians were called back. A shortly after the severe layoffs, I felt embarrassed by receiving two pay increases! One was standard improvement, the other a salary adjustment to bring me up to what new hires were receiving. This was after I was told when leaving GM that I’d received more pay increases than any other in our dept. When I hired in to GM I was given a pay increase relative to what I received as a Civil Engineer. These double increases revealed a big difference in pay scales for civil service and auto industry vs aerospace research; though some of the difference as due to inflation.

Belly Shakes During Large Rocket Engine Firings:

For the first three weeks I was assigned a desk in Keller’s office while people were being relocated. Plant engineering was still moving the bomb shelter up Lakewood Blvd for our new test lab, my eventual assignment, and I was assigned to work with Don Calkins on a Jet Vane test at the Rocketdyne rocket firing site at Canoga Park CA. Technician Bob Logan, ex Navy, had taken a test console and hydraulic pump there, to be used to operate the fiberglass Jet Vanes used for attitude control of the Navaho liquid rocket booster. At the time I was unaware this was Americas first large booster engine after WW II, an important step in evolving the huge booster engines to put man on the moon. The Vanes poked into the rocket blast serving as airfoils and were eaten away during the process. The Germans had used carbon and our engineers were experimenting with fiber glass, knowing they would ablate away, but hoping they would do a better job. Movies were taken to show how well the vanes held up in the 2000 deg F blast. We watched from a mile away across the valley on a hill – even at that distance the sound caused my belly to shake like jelly. This was an awesome setup of test stands and block houses, in a canyon once a famous location for making cowboy movies. I knew I was in the middle of leading edge stuff. All the fellows were young and speaking a technical language new to me. WW II B-29 take off time noise was quiet compared with this.

See NAVAHO file describing the missile and engines under being tested

My First Real Challenge – and Capability Test

I first met Howard Topp & Bill Stroble at Don Calkins when Don invited some of George Kellers unit to his place for a party. {Howard Topp stayed with Space division when Autonetics was spun off and I don’t know what happened to him.} { Bill Stobel married later than most, they were very happy building new house for pending baby then abruptly Bill died of a heart attack. Everyone liked Bill, his enthusiasm was contagious.} This gave me a chance to meet some of the people I would be working with. A couple weeks later I called Howard Topp late one night at home, he came to the plant to help.

The second week I was told to help Clair Harshberger, one of the engineers assigned to the assembly area for the G-26 booster. Clair showed me how to run a frequency response test. He said he was going to Edwards Air Base the next day on a problem, and for me to stand by – though there would be nothing to do. The next day I had no more than walked through the door when “Tiny”, a huge fellow, the production foreman, asked me to run a frequency response test on the boosters jet vane servos. At GM, production was king and engineering dropped everything until productions problem was solved – I assumed NAA was the same way. What I didn’t know was that several engineers had been asked to run that test, it failed and nobody knew why. Tiny was being held up, the missile wouldn’t pass the test, nobody could tell him why – or how to fix it? This was the first G-26 Booster, with more to follow.

Setting Servo Actuators to Null

Two days before, I’d helped “null” jet vane servos, to set the position sensor to the center of the actuators stroke. To get into the “secret” area I was sent to have a “football” sticker placed on my badge – George’s signature was enough to authorize my having this special sticker. (They had already verified my military service record and engineering degree) For this task I was introduced to a VTVM (Vacuum Tube Voltmeter), used to determine when the position transducer (position sensor) was located at it’s center, which they called it’s Null (center of stoke). WW II aircraft were primarily mechanical, however the B-29 did have a C-1 Autopilot, that used variable resistors (potentiometers) for position sensors. WW II electrical specialists mechanics used DC voltmeters to measure the position of those. Missiles experienced extreme vibration and a moving wiper on a potentiometer was inadequate, as it would fret and loose contact. This rotary missile position sensors did not have rubbing parts. A rotating armature magnetically coupled transformer field windings. When centered the slug equally coupled opposing plus vs minus secondary winding voltages causing the output to go to a minimum, not to zero, but to it’s null. Transformers require AC excitation to work and output AC voltage which increases in voltage when moved from center. The output was demodulated the AC to DC, to go plus dc in one direction and minus in the other direction. {a demodulator is a full wave rectifier, where this one uses vacuum tubes (later missiles used transistors) to open or block signal flow, diodes (electrical check valves) have a voltage threshold and not rectify low level signals.} I was reading the transducers AC output direct with a VTVM. Even this small part of the system was new to me.

Left: The Navaho Missile was an Intercontinental Ramjet with forward kinard wings and aft swept delta wing who’s stainless steel body flew at mach 3. It rode piggy back to launch on a booster powered by the first US liquid rocket engine which was built by Rocketdyne division in Canoga Park and assembled at the Downey plant site which would later become Space Division. The booster, not seen in this view, is attitude controlled by jet vanes that poke into the rocket exhaust serving as airfoils.

Rightt: Typical frequency response test console mounted in portable electronic rack.

Rapid Changing Technology

I was constantly being exposed to new words and equipment. Technology was moving at a very rapid pace. I’d jumped from my first Model-T in 1939 to Model-A in 1941, to college, to B-29s in 1945, back to college, to civil engineering in 1949 to process engineering in 1951 to intercontinental missiles in 1955. Transistors had just come out and would be replacing vacuum tubes. Prior experience had not prepared me for what I was encountering.

Mechanical engineering books, immediately after WW II, were those written before WW II, the technology advancements of WW II had not yet entered into the school system while I was still in school. Engineers graduating after me were taught subjects we did not have, such as controls systems synthesis. I found myself in the midst of a select collection of highly skilled engineers with many specialties. It was assumed I knew what I was doing or I would not have been hired. Everyone was learning at a rapid pace, many were going to school at night. After four months on the job George had my boss Paris Stafford and I attend night classes at UCLA. I was not the only one being pressed to keep up.

The Setting—Final Assembly of the G-26 missile

Before continuing let me give you an appreciation of what I was trying to digest in a very short time. It was awesome to be in the final assembly area with the imposing Ram Jet powered missile, still disconnected but towering above the large booster on which it was to ride piggy back. The booster was a large fuel tank in front of a large liquid rocket engine housed in a structure with legs poking out the back. The material was not aluminum but stainless steel! The extensions that looked like feet contained actuator operated fiberglass vanes that extended into the rocket engine exhaust – these were to control flight during booster. The booster was to carry the Ram Jet powered transcontinental section to a proper altitude and speed so it’s ram jets could ignite. The part I was supposed to be testing was the servo actuator assembly, out of sight, inside the foot – the part that controlled vane position.

This would be the First of Five

This was the first of five B-26 missiles that were completed and launched from Cape Canaveral in FL. The first one launched but could not ignite the ram jet. This was fantastic film to watch, beautiful clear sky as the two, still joined together cycling as if a huge airborne Propose, increasing in amplitude, striving to ignite then – breaking apart – falling into the nearby ocean, a failure. As we watched the film I’m sure all of us were applying body English trying to help it succeed. The last of the five was an absolute success – destroyed by the range officer when he thought it was not responding to his command to turn from a Cuba bound trajectory. It was turning as commanded, but looking down range it was going so fast it was making too big an arc to detect it’s turn. Shortly after the program was cancelled.

The Frequency Response Test

That first day with Clair the booster jet vane control system had not passed the frequency response test – but I was hardly aware, being overwhelmed by the new terms, the equipment and procedure I was dealing with for the first time. The following day after Tiny’s request I kept running the test, rechecking the manual, trying and trying to get the test to pass, convinced it was because I wasn’t running the test right.

For performing this test there was an electronic console, about the size of a standard work bench, equipped with oscilloscope, recorders, and panels with labeled knobs and switches. While off to one side was another console that delivered hydraulic power to the vehicle for ground check out. During the first day my attention was focused on the test console, the names of the players, what they did and how to operate them in a procedural sequence. That was the easy part, it was the terms and descriptions used when talking about the part I couldn’t see, which was the part I was to test – the hydraulic servo actuator in the vehicle was commanded by control electronics inside the test console.

Most of the terms were new to me – trying to master things I could not see. Thankfully I was not afraid to ask questions, and during my first two weeks I was asking questions and reading almost continuously. This is some of what I was learning.

This Minuteman I servo-actuator, though different, shows the integration of servo-actuator parts.

The earlier model Navaho had a bolt on servo valve and a side mounted rotary position transducer.

The above early Minuteman design is an advance over its Navaho predecessor. The linear position transducer, buried inside the actuator, was invented by Gary Collins while still an NAA engineer. The idea was based on IF radio transformers which used a movable slug to tune flux coupling. The transducer is a transformer, who’s output varies according to the position of it’s inner slug – null (minimum) at center, increasing plus or minus when moved.

A hydraulic servo actuator consists of a vacuum tube control hydraulic servo valve, which on command ports fluid to extend or retract the hydraulic actuator. A position transducer {Transducers are devices that convert conditions as pressure, position, velocity, etc; to an electrical signal.} on the actuator sends a position feedback signal to the control electronics – where feedback is compared with the position command. {The term feedback is where the output of something is brought back and compared with an intended command, to determining their difference; i.e., the error between actual and intended. This is called negative feedback because it’s connected to cancel a command.} The command enters through a resistor to a summing junction. The feedback signal also enters through a resistor connected to this same summing junction. The voltage at the summing junction point is indeed the “arithmetic sum” of the two analog voltages; the difference between the commanded position and actual position. This error is amplified by the electronics amplifier. {The amplifier, sometimes called the valve driver, converts a small electrical input signals to output signals sufficient to command the hydraulic valve.} The signal is amplified by an adjustable amount called amplifier gain. {Amplifier gain control is similar to the volume control on a TV.} The output of this amplifier goes to the servo valve. When the electronics and servo actuator are connected it is called a closed loop. If the gain is set too low performance is too sluggish and if too high it becomes unstable. A frequency response test is used to properly set loop gain and prove all parts are connected and working properly. A sine wave signal generator is used to send commands to the servo loop and the amplitude command is set for about half stroke. The command and feedback signals are fed into the X and Y coordinates of an Oscilloscope (with TV like display tube). When the command and feedback signals are in phase, an hour glass figure is created on the Oscilloscope. As commands to the servo actuator increase in frequency, the actuator amplitude drops off and there is a time delay, a phase shift, between command and feedback signals. The control test panel had a phase shift dial, which was adjusted to achieve an hour glass figure on the oscilloscope. An operator records performance at various frequencies and plotts amplitude ratio & phase shift vs command frequency. {A plot of amplitude ratio began as being 1.00 (like 100%) at 1 cycle then as a ratio (%) of 1.00 at higher frequencies – many of terms came from radio electronics.} A sine wave command causes a sine wave feedback, with amplitude in sync at low frequency. As the frequency increases the device cannot keep up and its amount of lag is measured by it’s shift in phase, a measure of how much it’s out of sync.

The amplifier gain can be increase to keep amplitude ratio from falling off too fast, but if set too high it would peak above 1 and become unstable.

Our test console included a Sanborn Recorder, with hot wire on paper, similar to those use in Dr’s offices at that time used for running EKG tests. The prior day while watching Clair Hershbarger run the test I was doing well just to learn the names of some of the instruments. On this second day I was reading and rereading about each item learning the terms on my own. I thought perhaps some of the electronics in the console had not been set right, knowing that Howard Topp had designed and built the test equipment we were using I called him at home to come help.

Early servo valve torque motors were “wet”, immersed in the oil. While testing servo valves for the Navaho I found metal would collect and inhibit control flapper motion. I showed these to Jim Hager of Moog valve. Jim initiated design of an isolation tube to keep their motors “dry”. This was an industry first and made it possible for Moog to make bi-propellant valves that operated from one torque motor, a significant advantage – winning them the Minuteman III Post Boost contract for 13 of their valves in each vehicle.

Howard a terrific young engineer was soon there and worked with me for several hours not finding anything wrong with the equipment or how I was running the test. A bit before midnight the night shift lead engineer, who had stopped by and was working with us said, go home, worry about it tomorrow.

The next morning I learned more about what a frequency response really measured and gathered complete details of the test console, the servo actuator and the vehicle hydraulic system. A diagram of the hydraulic plumbing revealed there were two check valves in the lines feeding the servo – they were to block loss of fluid when the foot structure with servo actuators and control vane were blown off. I closely inspected the air frame structure and extended foot, they were made of stainless steel and had been hand pounded to shape. Ah, I felt on familiar ground and became immediately convinced the problem was with the check valves not properly opening and were throttling fluid flow. From experience at GM building airframe I knew the importance of tools, fixtures and drawing tolerances to assure there are no miss fits due to tolerance build ups. Using drawing dimensions I made a tolerance build up analysis. That was it ! The check valves were only cracked open, enough to permit operation at low frequency but not enough for rapid motion. Keller was impressed and gave me full support and more responsibility after that.

I was hired to help set up this facility at Downey CA plant

Lee Atwood Chief Engineer of NAA visits the Bomb Shelter

The Bomb Shelter had been operational for about a year, and others happened to be gone, when I looked up from my desk in the innards of the Bomb Shelter and saw Lee Atwood standing there! He said he’d heard of our facility and would like a tour. { Lee Atwood followed Dutch Kindleberger, as Chief Engineer NAA then CEO of NAA when Dutch died. Lee was active up till the time he died at age 95.} Lee had a habit of coming to Downey on the late shift and visit with those in the Simmulation Lab. (Sets of analog computers used to make flight control studies) There he teamed with Mal Johnson to submit a patent on ideas Atwood had on gust alleviation, using a B-25 as the model. They were awarded the patent. {Malcom Johnson and I shared an office for several years, during which Mal’s services were constantly in demand to run flight simulations or the equivalent on IBM main frames. He was supervisor of such studies before retiring & attends our annual Christmas party.}

I spend the next hour giving Atwood a tour and the reasoning behind what we had done. The facility had four main test cells with ovens and was equipped to remotely control tests on equipments operating up to 1100 F and 4000 psi. I had designed and built special heat exchanger systems to pump up to 8 gpm with a room temperature pump delivering oil at 600 F 3000 psi. We were using OS-45 a silicon based synthetic hydraulic fluid made by Monsanto.

The tasks required much innovation and I was able to describe each item to Atwood, often telling the source of the idea – sometimes referring to how a similar thing was done on aircraft systems – on which I knew he was an expert. Many of the remote controls were Palley Supply, LA surplus sales, aircraft parts. I had gone through aircraft mechanic school, taught in such a school and graduated from Yale University in a cadet program for aircraft maintenance – thus I was widely versed in things I knew that he knew. When we completed the tour we paused at where we started. He looked at me for a long long time, saying nothing, I could almost see his thoughts at work. Then with a very nice smile he said thank you and left. I’ve often wondered what was going through his thoughts. The NAA division had nothing to compare with what we had, which is why they sent some of their B-70 work to Downey for testing.

Buck (Elliott Buxton) had mailed Atwood a one page description of North American Aviation I’d written as preamble to a scanned copy of an NAA 1945 Aeronautics Handbook loaned to me by Fritz Gardner. {Fritz Gardner was head of our internal IR&D during early Minuteman. He’s now retired we still have lunch together. Fritz was in the Signal Corps out of ROTC Ohio State, reassigned to AF at Wright Field. He was assigned to determine what was wrong with the B-17 supercharger control & found Honeywell forgot to include a critical bias resistor on a main amplifier tube. Fritz attends annual Christmas party.} Lee Atwood wrote back, at perhaps age 93, saying that was the most condensed complete description of NAA he’d ever read -- then made a correction about where in Tennessee the Minuteman III Post Boost system had been tested. It was nice to know his mind remained sharp as ever.

Secret to P-51 Performance

After I retired I gave Buxton a CD on which I’d documented the Bomb Shelter facility. Buck had not been part of what we did there during the Navaho, but was interested and aware of what we did for early Minuteman, such as experiments with fly wheels for running Minuteman I hydraulic pumps. In return Buck mailed me a paper written by Atwood on the design of the P-51, where they had salvaged heat energy from the coolant to provide extra thrust. That’s when I appreciated how Lee was tracking everything I’d been saying about our handling of heat transfer problems – he was an expert in the field, and without elaboration had appreciated our problems and solutions. Atwood’s article tells how he applied British research on salvaging heat from coolant to convince the British to fund a new plane rather than produce more P-40’s. Later Messershmit told a North American Engineer they probably ran more wind tunnel test than we had on the P-51 trying to determine how it achieved superior performance. Even he had not taken note of what was buried under the pilot, us mechanics were told superior performance was due to it’s new Laminar Flow wing. Of the energy in a gallon of gas, 1/3 is applied to the crank shaft, 1/3 goes out the exhaust, and 1/3 is carried away in the coolant, thus even if only 10% is recovered as thrust, it provides a significant advantage.

Remote Controlled Mirrors: These captured the attention of all visitors. The system was based on a periscope method of seeing around an oven with two fixed mirrors to view a third overhead operator controlled mirror. They worked very well. I got the idea from a WW I book of my fathers showing soldiers looking out from the protection of a trench.

Dealing with concrete walls: I often visited with Bill Stobel as he was a bright capable likable guy. One day he showed up at the Bomb Shelter with things he wanted tested. He began trying to tell me how to set up the test and became hung up by the fact that the walls, on which he wanted to mount things, were all concrete. I said, Bill don’t waste time telling me how to make the set up, just tell me what your wanting to do and achieve. He told me and I said come back in a few hours and we will have it ready for you. He came back and was delighted and intrigued by what we had done. We had mounted his equipment on the concrete walls by drilling holes in the concrete with carbide tip drill bit, hammered in lead devices called TampIns and mounted the thing with screws. Bill was highly respected by all as being bright and skilled in the domain of mathematics. Bill was too young for WW II, had graduated from Berkley, going right into aerospace. He’d never been exposed to “barn yard mechanics” on how to do things, which he mastered later by his own initiative.

Open Charge Account: We had the most advanced high temp test facility in the US. It was common to conduct tours of engineers from other aerospace companies through our facilities, tours set up by George Keller. In fact George arranged for me to have a Gray (Specialist) badge so I could escort visitors entering via the back gate. An item that intrigued visitors was our system of remote controlled mirrors rigged to replace our closed circuit TV for “seeing” equipment on the other side of the ovens. I used war surplus camera control motors to remote control mirror position. I was authorized an unlimited charge number to the shop to build whatever I sketched on my desktop my large quadrille pad. These were the same craftsmen working on inertial platforms. Compared to a quick adequate job at GM, these were often overkill masterpieces – they didn’t “glue” weld legs to a plate, they machined them to fit and attach with countersunk machine screws!

Heat Exchangers & ASV Valve: We did work for Johns Hopkins University, testing their ASV Acceleration Switching Valve – George hoped to pick up Navy work this way. Later the manager of Johns Hopkins controls systems studies came out for a visit as they wanted to build such a facility. I declined a job offer, but did offer to help them specify a heat exchanger system they intended to buy – as there was nothing off the shelf to do such things. I provided requirements in the form of a spec plus response to a request for quote. I gave it a high cost and plenty of time after design approval, convinced that was the last of it. I was shocked to find they called Stafford saying never mind design approval, just build it! It was then I got out my McAdams Heat transfer text book and began design of the thing – I never did well in math nor well in our Heat Transfer class. It was right after WWII and we didn’t have a text book until half way through the semester – the instructor was an ex Army officer and was giving us tests prepared by another instructor. All of us in the class were ex GI’s recently returned and went about our studies as if killing snakes – we applied brute force to master that class. I’d specified conditions the heat exchanger was to meet so I could not just hit and miss, modify to make work, as with our own equipment. This resulted in my first contact with Tom Schuler, Tom reviewed my heat transfer equations before their delivery to Johns Hopkins as part of the design submittal.

TMV then ASV Valves: North American Aviation had developed a TMV, Time Modulated Valve, a single stage valve, a porting spool dithered between two solenoid valves. Ray Curci { Ray Curci was in the project office when I hired in and was a lead project engineer when he retired. I met Ray during my first month and we continued to have work contact and visit ever since. He once canceled a team of people going to a sub contractor and sent me in their place. We had often proven reason to have confidence in each other. Having health problems.} had helped design this valve before I hired in. The TMV valve single stage was a fore runner to a two stage valve developed by Moog. The ASW valve was an adaptation to the Moog two stage valve. I later applied things learned about the ASV Valve, when evolving a way to command an analog servo valve coil with digital electronics. The Johns Hopkins engineers had removed the centering springs from the second stage spool and superimposed a dither frequency on the valve command with the idea being that a dithering valve would not hang up with contamination. All such valves were designed to be driven by vacuum tubes, with an output of about 8 ma.

Remote Control of Hydraulic & Electrical Power: Necessity caused me to determine how to remotely control and instrument primary facility equipment as well as avionic devices. I took the hydraulic pump control valve apart and found I could plumb a garden variety needle valve in the control room to adjust hydraulic power. Similarly I determined how to remotely control three phase 440 volt heavy duty power boxes by rewiring then so their control wires fed through small aircraft switches on the control room panel. I was constantly in trouble with the electrical union because of things like that. The facility used about 700 amps of three phase 440 volt power. This power arrangement was a mess, often modified and patched. After repeated tries plant engineering gave up on how to fix it. So I made a design and took it to the head of plant engineering. He was motivated to have this fixed because main plant power passed through there and was often knocked out if we blew a circuit. He implemented the change via a private contractor. I was amazed to find the contractors work order directing them to built per my large quadrille pad layout!

One reason for the confused maze was that it was very difficult to run electrical conduits, or similar pipe through the 15 inch steel reinforced bomb shelter walls. When I sent a request to plant engineering for an opening from control room to test cell, they sent a crew of three who jack hammered for two days. They never did hammer through, encountering multiple 1 ½ inch reinforcing bars.

When the outside electrical contractor arrived I was apprehensive of their ability to implement the layout I’d made. No sweat, they came in a day in advance, drilled four holes into the concrete, pounded in lead tampins and mounted a 12 inch diameter carbide tipped hole saw driven by a large electric motor they advanced as if a drill press. In less than an hour, using water cooling, they had a clean port cut through concrete and steel. I knew immediately these guys were professionals who would do a first class job, and they did.

We emulated a kids squirt gun to start a fire in this oven to plot % Oxygen vs Temp to ignite silicon oil.

It was a challenge to determine how and to implement this test. We potted electrical wires using glass like plastic in a hydraulic fitting in order to hold vacuum. We bought Saturable Reactor control system, with full instrumentation to control the heaters. It was a delight to work with the two bright young engineers who provided the equipment. I had to get special transformers from our plant engineering to interface our plant power to their equipment.

Pre Fill Evacuating Hydraulic System: I later used this vacuum pump to evacuate air from MM I Nozzle control units – air was sucked out and oil let in. Oil itself was vacuum treated to remove entrapped air. Doing experiments with actuators made of clear plastic we were amazed to watch bubbles in the oil come out of solution. Thus I made it a requirement to be sure all air was removed from drilled hydraulic passages as well as the oil. Such things contributed to the success and reliability of our equipment – though few people knew of this.

Electrical Isolation with Transformers: Transformers are flux coupled, with input and output capable of being at very different voltage potentials. However transformers only operate from alternating current. The voltage ratios between the two sides is dependent upon the ratio of turns between primary in and secondary out. To achieve a continuous, direct current output, it’s necessary to use diodes to rectify the AC into DC. I needed to isolate cold temperature from hot temperature. A primary and secondary actuator coupled with a shaft, could provide thermal isolation and transmit power by push pull motion – AC hydraulics??

Heat Isolation with Hydraulic Transformers

The NAA division was still working on the B-70 and I had an idea on how to do away with the costly heat exchanger systems. Having become comfortable with a mix of Electrical-Hydraulic-Electronic disciplines it occurred to me that we could use a transformer {An electric transformer has an independent input winding (primary) and an independent output winding (secondary) connected by magnetic flux – the two sides are otherwise isolated from each other.}to isolate heat chambers, room temp outside and high temp inside an oven, as if IF transformers separating vacuum tube voltage stages. (IF Transformer story follows) By then the Minuteman program was under way and I’d been called in from the Bomb Shelter to work on Minuteman Power elements testing. Still thinking of how to solve the heat exchanger problems for the B-70 need, I used load cylinders set up to test early Minuteman servos as a means of proving my idea. I set up a Minuteman actuator to be the “primary” (cold) winding and the shaft attached (isolated) load cylinder to be the “secondary” (hot) winding. I rigged a full wave rectifier using check valves as diodes to rectify the AC hydraulics to be DC output hydraulics. I used hydraulic accumulators as capacitors to make a pressure (voltage) “filter” to smooth out the DC. I connected a hydrauliscope to measure the quality of the DC. I used a servo driver test console to drive the primary, using a “pressure control” servo valve in lieu of a flow control valve. I set up a switches at the end of the stoke to reverse current flow through the valve coil so it would produce a defined pressure in either direction and auto reverse. I used a pressure transducer on the DC output as feed back to the commanded pressure control. I was told by many experts that it would not work, at least not well enough. So I set up a third actuator to operate off the DC output – and ran a frequency response test to prove it would work in response to a variable load demand. It took a technician and I one day to set it up and run the test. I called in the key people and skeptics to see proof it worked – then made a quadrille pad design of it to be sent to NAA division for use on the B-70 program. However they were shutting down the B-70 program, and plans for a Super Sonic Transport – all need for high temperature hydraulics evaporated. The Europeans continued and built the Concord. Due to our work on the Navaho and B-70 we were considerably ahead of the Europeans, they were biting off more than they realized.

Operational Amplifiers in Hydraulic Test console

Art Greer asked me to come up with a servo control test console for use in developing Minuteman servos. He had in mind a console with a full set of variable resistors, capacitors to simulate various shaping networks. We also discussed the need to be able to program a load cylinder to produce various loads. These loads were to be a function of position, velocity and acceleration. At first I thought I’d need position, velocity and acceleration transducers, then I realized we could electronically convert position x to velocity v = dx/dt and acceleration a = dv/dt. I started to design such a thing myself and Stafford said; define it then buy it, you don’t have time to be designing test equipment. So purchasing called in an outfit that made computer simulation labs equipment. This was an education to me, and I was soon learning about operational amplifiers and their use. The system was built and tested but it was never used. By then a Minuteman Simulation Lab was in operation and we found no need to do complex filtering on the hydraulic servo loops for the Minuteman. Experience with the Hound Dog Missile had required building in notch filters to solve a body bending mode had led us to believe we might need such a simulator. (This was before digital computers and it was necessary to use analog operational amplifier computers, to do such things as block out body bending notch frequencies while passing lesser and greater flight control frequencies.) Clarence Ashe {Clarence Ashe later went to work for Honeywell, eventually becoming a Director, he was very bright and capable. We still exchange Christmas cards.} evolved a notch filter for that servo and was sent into the field to check vehicles in the inventory. I was very pleased to have Clarence work for me on Minuteman III before being moved to other assignments – eventually responding to an excellent offer from Honeywell.

Isolation with IF Radio Transformers

During a train ride coast to coast in WW II I bought and read a book on Superhetrodyne Radio, and was introduced to oscillators, carrier waves & detectors. The feeble incoming signal is imposed on an IF, Intermediate Frequency, generated by an oscillator which “carries” the signal through amplification stages well above signal ground – as if to a higher altitude so the desired signal cannot be clipped off by the tree top noise close to the ground. I soon found I was revisiting this method when learning about Operational Amplifiers as applied to Analog Computers. Operational Amplifiers handle signals at a carrier frequency, so signal swings modified to simulate an event can do so without the distortion of ground effects. Resistor/Capacitor network methods, for applying LaPlace Transforms, were made obsolete by digital computer sample data methods. Control systems synthesis, evolved in those days, is being applied to molecular biology today.

The Bomb Shelter Test Lab opened the door to many worlds

It was a fantastic learning experience