HD1975PDPref

Patent Disclosure Preface

Patent Disclosure material has little meaning without and understanding of the setting and circumstance. The patent disclosure files consist of scant material remaining from what I had submitted, perhaps this preface can shed light on the otherwise quite esoteric.

Everyone was encouraged to make Patent Disclosures of any ideas they thought might have possibilities. This was expected especially from Sr Research Engineers. For example at a staff meeting it was announce there had been a lull in patent disclosure submittals. I was brought out of non involvement, aware others were looking at me. I had been doing “internal IR&D, “playing” with new ideas while the others “worked”. In a polite way I was “told”, in the form of a suggestion, that it would be nice if I submitted patent disclosures for some, at least one, of the ideas I’d talk about at the coffee machine. I dreaded the time such consumed, I didn’t feel I had anything patent-able -- that I knew would work. Others quickly told me – you don’t need a working model or need to have proven that it works – in effect they wanted and expected me to get them off the hook and fill the departments submittal quota. As a result of that “assignment” I wrote up an idea I’d abandon for a “digital position transducer”. The need and value of such read well, and since I didn’t need to prove it worked, made the submittal and forgot it. I was called back to do some rewrite and fill in information by the company Patent Office, as they were submitting it to the AF. I figured that was the last I’d hear of it -- thus I was quite surprised to hear the AF had assigned a patent attorney to complete and submit the concept for patent. Some 10 years later I received a check for $1000 for the submittal – a standard amount given when the AF receives a patent on a submitted idea.

When hiring in we signed an agreement stating that any patent-able ideas derived as a result of our work, belonged to the company or the government as they were paying for our efforts. Many avoided making the effort of writing up a patent because the extra personal work exceeded any reward – except for those to whom recognition was of personal importance.

In college I had taken a course in patents and inventions, what I remembered was an idea was not patent-able if anyone skilled in the trade would do that to fill a need. Also that adding an erasure to a pencil was not patent-able unless somehow 1 plus 1 equaled 3, rather than 2. I considered I was someone skilled in the trade solving a new problem. What I was doing seemed too obvious. I felt many things submitted for patent were either public domain ideas or not worthy of patent. My mind set was not patent oriented.

File HD61PD: This idea was to step hydraulic porting in a way to cause a servo actuators position to step an increment forward or backward as commanded. The industry was trying to come up with “digital” controls hardware to interface with “digital” computer outputs. Transistorized digital control was a hot buzzword in hydraulic servo technology. The industry was expending much time and energy on how to do this. In retrospect they wasted much time because they had no concept of the inner workings of digital computers.

My hydraulic actuator derived from attempt to make a stepping electrical clock. Aunt Margaret and Uncle Francis had given us a beautiful electric “see through” clock run with an electric motor in the base. However in time dust increased friction overwhelming the electrical clock motor and it would stop. I had in mind walking magnetic pole pieces to move the clock mechanism -- I actually built the concept but never completed it.

I came up with a three element “walking” process when I concluded only three independently active teeth were needed to “command” a large spur gear with many teeth. The extended active tooth would hold the large gear in a fixed position while independent active teeth on each side would be poised to move the gear clockwise or counter clockwise depending on which was activated. Thus three independent solenoid driven teeth could turn the large gear forward or backward by walking the electrical commands clockwise or counterclockwise. If electric pole pieces were used in lieu of gear teeth, then a see through clock wheel could be step rotated as a clock face.

Metal working lathe “lead screws”, used to advance thread cutting tools, use square threads, a kind of linear gear tooth shape. Perhaps these could be used to port or block fluid – like hydraulic control fluid? By a flip of thought the ”threads” could be cut on the inside diameter of a shaft. These thoughts led to the design of an actuator with hydraulic extend and retract pressure ports locking or enabling piston motion – by “walking” hydraulic control port flow.

Jet pipe servo valves, a newly available invention, was applied to step port pressure and return fluid. The submitted diagram shows how they were to be connected.

I abandon the idea because stall loads would ruin position indexing, and such a design would be very vulnerable to contamination – also positioning resolution would be dependent on mechanical steps.

Years later I saw a watch on sale that advertised the use of a three bit stepping method almost identical in principle to my uncompleted clock in the garage.

File HD62IL: It was well known in solid propellant rocket business that much energy goes to waste. A nozzle throat at the bottom of the combustion chamber flow limits at sonic velocity, thus maintaining pressure in the combustion chamber. Combustion chamber temperature reach as 8000 deg F and cool due to expansion down the expanding exhaust nozzle. Hot gases bouncing off the expanding nozzle walls produce “forward” thrust. Idealistically the nozzle is expanded until the pressure at the exit plane is equal to atmospheric pressure. Ideal the exit diameter should be increased as a missile goes from sea level to higher altitude to optimize efficiency. Experiments had shown that squirting a liquid into the nozzle about 1/3 of the way down from the throat, a side force could be achieved. From this was born the idea of attitude control by fluid injection. Minuteman II stage II was designed to use secondary injection. At a meeting sponsored by Vickers in Detroit, I visited with a fellow from Naval Ordnance Test Station, wondering what the effect would be if ram air were injected uniformly – would that result in an increase in total thrust? What if water was injected, flashing it to steam, would that increase of mass and reduction in nozzle temperature result in more thrust? Typically the temperature at the exit plane was about 2000 deg F. I presented my the idea without delving into the math. Real assignments soon diverted my attention.

File HD70PD: Persons studying electric motors and their use appreciate the tremendous reliability and economy of the 60 cycle squirrel cage motor. There are no brushes, no moving parts except the armature delivering power on it’s shaft. All households use these extensively. A power plant generates a three phase output at 60 cycles per second. Squirrel cage motors lock onto this frequency and run at a fixed speed based on the power frequency. When a load become too heavy the motor will slip and not keep up. Diesel powered locomotives generate electricity and harness it in the same way. However large gear reductions and clutches were needed to generate full pulling power at low speeds. Signal transistors were developed more rapidly than power transistors. Good power transistors were difficult to obtain for Minuteman I. However in time power transistors began to catch up. When our Post Boost Propulsion System program came to an end, I looked into what might be possible with the improved power transistors. While experimenting at home with Silicon Controlled Rectifiers, a power diode rectifier with control lead; I also experimented with using power transistors in either a full on or full off mode – they tended to burn out when operated in throttling mode which causes them to over heated. (I would latter use this full on or full off mode to make it possible for a tiny CMOS transistor to drive the coil of a hydraulic servo valve.) While doing such experiments it occurred to me that you could use these power transistors to create alternating current of any frequency desired. We had been making alternating current by chopping battery power to excite MMI position transducer. I had been using silicon controlled rectifiers to make variable DC power from 60 cycle AC. Why not put the ideas together and make a three phase motor who’s speed was set by the frequency of command to transistors at each end of their Y wiring connections. A transistor only flows one way – however since they made PNP and NPN kinds of transistors where one flowed one way and the other the opposite it was possible to make a switching array that would walk three phase current flow at a speed determined by the operator.

With parts purchase out of my own pocket I had spent a considerable amount of time and money setting up the switches and even building and winding my own three phase field piece. When the mail came I laid it on the work bench and continued with my project. Then looking for something, I found it on top of the new Controls magazine – there, on the front page, was the full layout of the concept I was trying to build. The controls industry was already hard at work developing the idea. I stopped work immediately and read the article with much interest. What I’d come up with on my own was almost identical to what the article presented. Making the case that it might not be patent-able if it could be done by anyone skilled in the trade.

Years later a feature article in Union Pacific annual report told of using the new semiconductor frequency controlled electric motors on freight train engines to generate full power at very low speeds, in fact at any speed.

File HD75PD This is the one in which I was given a $1000 dollar award, as the idea was patented by the AF.

Happy ending for abandon idea

Ideas not submitted for patent

Aerodynamic Bump Control: This idea began with MIT’s Diamond Ordnance Lab’s Doffle Men shown at right in the following figure. The AF knew they could be faced with nuclear survivability problems and were seeking some way to compute and perform logic functions without dependence on electronics. Someone found a way to make fluid elements which would perform flip flop logic functions and the AF funded MIT’s lab to develop the concept. The flow paths looked like little men with source head, two control arms and two output legs. By adding or removing pressure at a control arm, the output could be flipped from one leg to the other. Corning glass produced this in logic arrays by etching glass. Those in controls technology had their fling at adapting these, but nothing became of it. At the time I was engrossed in rocket engine attitude controls and had expansion nozzles and secondary injection on my mind. I was also reading a book on Boundary Layer flow, by Schlictin a German reporting work they had done. I sketched the contours of a trust nozzle then folded it out to where the nozzle contour was on the outside. I knew the resultant ‘”bump” would disturb air flow, using Doffleman methods, perhaps control air could be used to convert stagnant air to laminar flow – changing pressure on the downstream surface. By adding such a bump to forward and back end I could create a control force couple to do attitude control. Ram air could be used as the source of air and the outside of the “vehicle” could be a “flying Doffleman”. I proceeded with the idea of testing it out in the form of a plastic garbage can converted into a supply drop package – with built in passages and contours for guiding supply drops to a target. I made a number of sketches of the idea and began to pursue the problem mathematically. I was trying to use boundary layer equations from the book but became frustrated when what I took as multiply “dots” were the European way of showing a decimal point. Math was not my strong suit and work demands soon diverted my attention. I felt certain the idea would work, but how well and at what cost I did not pursue.

Work on the following was link to doffleman application.

CMOS chip servo valve driver: While working on the idea of replacing analog electronics with digital electronics, the first stumbling block was how to drive a hydraulic servo valve torque motor using digital electronics only. Over and over I was told it could not be done. Emil Koeler, who had been assigned to work with me, gave me excellent guidance in how to think of controlling a transistor. He said you “source” the base to turn in on and “sink” the base to turn it off. The base is like the valve stem on a faucet. Let current in to source and drain current away to sink. It was a simple and effective way to think – thus avoiding “electron” and “hole” flow descriptions of physics hand books. He taught me another principle. Transistors burn out when they over heat. If a transistor if full ON it has almost no resistance and can carry lots of current without over heating. If it’s full OFF there is no current flow and no heating. Remembering this I made an H switch to drive the servo valve.

To appreciate what I was trying to do it would help to understand what had been done before.

Experimental Digital Valve to port Digital Bits Typical proven Analog Servo valve

The Digital Hydraulic Bit valve shown above was part of an R&D task to develop a Digital Valve. In the meantime vacuum tube driven analog servo valves were being used and by the start of Minuteman had become very reliable. These were at first driven with vacuum tubes, then using transistors for the Minuteman program.

Vacuum tube analog valve driver left transistor analog valve driver right

When browsing literature of new parts I found where RCA had come out with CMOS transistors, then a new technology. They sold a 4016 device for experimenters that included four independent CMOS transistors. I ordered some samples and used them to make an H switch to drive a servo valve. It worked!!! I brought people into the lab to show them the tiny chip in action controlling a servo valve – disbelievers were impressed, placing their hand on the device which remained at room temperature while operating the servo valve. Their response was, yeah but how are you going to command that thing with a binary digital number?

At the lower left above are the four transistors connected as an H switch, with the valve coil being the cross bar and switches in each of the four legs. Turning diagonally opposite pairs on or off will cause current to flow “extend” or “retract” through the coil.

I recalled what I’d learned from Bob Kelley and knew the valve would average high speed commands. I parallel loaded a shift register with all 1’s for full on, all 0’s for full off and some % in between to emulate an analog proportional control. I cycled the shift register at 6 khz, the same frequency as used to excite the position transducers. I used an electronic look up table to convert a binary command number to it’s equivalent % on. I’d strobe in the new command value at each command iteration update. This is shown in upper left above.

I then used the sign bit to vector the command as extend or retract. I strobed the sign bit into a flip flop latch which had dual complementary outputs. I the applied that to a pair of AND gates with truth table logic as shown above. This way the sign bit directed 1 “ON” commands to extend or retract and the 0 “OFF” commands simple remained off.

To the best of my knowledge this had never been done before and I was repeatedly describing how it worked to others within the company, other divisions and to other aerospace companies. It was also presented to the AF and TRW as a way to command pending MX missile servo valves. By the time the MX contract was activated such a method was considered state of the art – as if everyone in this field knew how to do that. But I spent hours and hours trying to figure how to actually built it and make it work. I initially did the look up table conversion with logic – a very tedious process. The next figure is from reports to TRW and AF on how the method worked.

What had been down stage Flight Control Electronic, was moved up stage into the Flight Control Computer where they implemented digital servo controls for MX program. Other people took over the work I’d been doing and I was assigned to help restart the B-1B program; during which I wore many hats and worked in a variety of technologies. Being adaptable was part of sustaining a pay check.