Apollo Guidance Computer Activities

AGC Conference 3 - Ed Duggan's Introduction

Apollo Guidance Computer History Project

Third conference

November 30, 2001

Ed Duggan.

Ed Duggan's Introduction

ED DUGGAN: My name is Ed Duggan. I graduated from Yale, in engineering, and then came to the Sloan School. The choice was to stay at Yale and be an architect for a year, which is about the budget I'd had, or come to Sloan and get a masters, or at least kill myself trying. And the latter happened and I didn't finish my degree but I left to work for what was then Thompson Products, in Cleveland, and they became TRW, down the road. Thompson Ramo Wooldridge and then pretty soon the tail started to wag the dog, eventually. I wasn't there for that transition, but Dan Tost was.

DAVE BATES: They had to work with them on the Air Force side of the house, for Jupiter vs Thor.

ED DUGGAN: And this was my factory training. They insisted, at that time, that any new engineering hire started on the factory floor for what turned out to be--the requirement was, until they started complaining so hard they were running the risk of losing us. I worked second shift for a year. Now, Cleveland is the worst place in the world to be as a young bachelor. They keep you occupied from three in the afternoon till midnight. The only thing that saves you is the bars stay open till three. So it was kind of a grim first year. But I complained and I ended up getting assigned to a project group which basically, in that time, and this is circa 1952, '53, was under a fellow named John Cofill.  His task was to move hardware from the Research & Development group, whose motto was, "WE never had that problem," and into production, which gets nothing but problems. And he was looking for a young and energetic kid that didn't know any better, and thought this could be done. And I got involved in it.

And it was a marvelous place to--it was ...(inaudible) the first motors that we used on some of these birds, and as a matter of fact I saw the San Asu (?) mountains get blown up one afternoon. But anyway, after a while--this was a facility that was owned by the navy during the war and quite a few inspectors carried over, and there was a navy culture there, and the inspection was tough. The misery, if you were involved with the factory, was very high, and I decided life had to be better somewhere, and maybe if I finished that thesis, because the fifth year was coming up and that was it.

So I quit and went to Vermont and I got involved with magnetic drums up there, for three months, and then the crickets started to make me nervous, then I moved back to Cleveland with another raise - they moved me back and forth, from time to time. But anyway, I came back to MIT with the dedicated purpose of finishing that thesis, and I needed a part-time job for the summer, and John Newton who ran the mechanical design part of Draper Laboratory, who was involved in the navy program, hired me to assist in his production control activity and whatever, run errands and that type of thing.

And when I walked into the place, I got there at 7:30 in the morning and Newton showed up, I think it was around 9:30, and I watched all the secretaries arriving, and this was in April, May, and they all had shorts on, and I was in tie and dark suit and the whole thing, and I couldn't believe it. And I finally got involved in this. And, to make a long story short, when I finally realized that I'd died and gone to Heaven was when I sent a requisition down to purchasing and the guy came back with six numbers and he said, "quit bothering me and use these as you need them." If I’d done that at TRW, my career would have been a lot shorter than five years.

But anyway, the six months got parlayed into 34 years, and I think, for probably all of us, the Apollo program was the highwater mark of our careers. It came at a period of time when there was a national purpose, it came well-funded, and it came, in the things that I do, namely, the packaging area, at a time where we had in depth experience with another well-inspired and well-run program, the navy missiles, which preceded it.

And, as far as packaging goes, most of what we brought to Apollo, with the exception of ICs and things that were developed along the way, and I'm not minimizing those, they were very important, the welded technology was the one we came to the Apollo with. The development of that was done through the laboratory, based on, as I recall it, early work. There were modules in the Mark 1 guidance system for the Polaris, which were basically cold soldered joints. While that problem was being worked on with Schenectady, in terms of quality control, the secondary effort, through Sam Francis, of Francis Associates, who was doing thermal work across the street, for the Air Force, came up with - I don't know whether it was unsolicited or not - a small proposal to look at the possibility of using the technique in the vacuum tube industry for making vacuum tubes, that is, welding.

And he basically came in with a way of arranging the parts involved in a NOR gate so that you could weld them together. And the program somehow fell under Steve Cudlitz, who was looking for somebody with vast welding experience, and of course when he found out that I knew a lot about welding from TRW, he never bothered to ask what kind of welding or anything else, sufficient to meet his requirements, so that there were 100 kVA welders doing turbine blades. I mean, he just transferred that into two, fifteen-thousandths wires, went out and bought me welding machines which blew those joints out faster than you could put them in.

And when I got the arc to be about this long (2 inches) I decided that maybe there was something wrong with these electrodes, we ought to look elsewhere. So I got involved with Gunnar Johnson, at Raytheon, in the old Chapel Street, and that was an odyssey in itself. There was one thing, in that phase of that program on the Polaris, that I think is a significant difference in the structure of the way the management, the development, the funding and how it was controlled. That is, the navy had us send money from our contract to whoever we bought from on a major development like that, and we could control it. In other words, if they weren't doing what we wanted, then we would say, "Well, we'll get somebody else," and shut them off. And it's amazing the attention you can develop with that kind of a clout. And I had never had that experience before. Whenever I went near a vendor before, at TRW, I had two purchasing guys with me, and they wouldn't let me stay overnight by myself.

DAVE BATES: I remember those first contracts says, build it like they tell you to.

ED DUGGAN: Yeah. I mean, it was open-ended, but it was great, because it--But, so we were able to come and make a prototype, which was a DDA, I think.

ELDON HALL: Polaris computer, at the time, which was an DDA computer.

__: The Mark 1.

ED DUGGAN: Yeah, it was a Mark 1, and it counted down the trajectory and came up with an answer for a given set of equations.  We had this out on the display bench that was worked up, and I remember it got the wrong answer one time but it was the same solution, only backwards. And fortunately the guy was demonstrating to some Navy brass. That's a whole other story.

DAVE BATES: However, Raytheon was not in on the Mark 1 program.

ELDON HALL: No, they came--

DAVE BATES: They came in on Mark 2, on a competitive basis. Jack never believed that, but it was a competitive basis.

ELDON HALL: Competitive basis. Since we had had this ...(inaudible) with you, ...(inaudible)

DAVE BATES: The welded design.

ED DUGGAN: Well, Chapel Street, I raised so much hell down there one day, we started getting people from the Sudbury facility that were getting interested in what was going on and what was this guy making so much noise about? I could yell a lot louder in those days than I can now.

DAVE BATES: Actually, from GE’s standpoint, we were happy to get Raytheon in doing the computer, because we had so much trouble with Light Military, trying to get them to do something.  I was the Polaris program manager at GE. It was called the ordnance department under Jerry Hoyt. And the thing was that when they broke away from the army in Huntsville, it went to GE in Pittsfield, the ordnance department, and the ordnance department did the IMU. The computer stuff was done by Light Military, in Utica.

DAVID MINDELL: Was this the same part of GE that was doing the Mark 56?


DAVID MINDELL: No, the fire control.

DAVE BATES: Yes. They were doing that in OP-1, losing money on every one they delivered.

ELDON HALL: Cline hasn't finished his story. With this breadboard there's one story. What about Captain Harold?

ED DUGGAN: Captain Harold?

ELDON HALL: You don't remember that one?

ED DUGGAN: No, you tell it. I remember Captain Harold.

ELDON HALL: Well, Captain Harold who was the navy's boss on this whole program, he was the big guy, and Ed had put this welded computer out here, on a bench like this, and they brought Captain Harold in to look at this thing, because we were trying to sell it to the navy, this approach for Mark II. And when you brought him in and showed it to him, the thing blew up in his face. It got him his pension.

ED DUGGAN: That's right, he was leaning over a module. The plastic blew, because we'd shorted it out.

DAVE BATES: Captain Harold was the guy in the navy who came down to Huntsville and the first head of SP, Special Projects office, and one of the guys that worked for him was a commander, Bill Kurtz, and he is the same guy that worked on Apollo. So you can see, there was an awful lot of interlacing with people. And that's what made this thing go. It wasn't an organization, and it wasn't contracts or this, it was people working together.

DAVID MINDELL: I think Harold was the guy who brought the Bureau of Ordnance to Draper, during the war.

DAVE BATES: That's right, he was.

ED DUGGAN: I remember coming back from an early meeting at Raytheon where we were discussing the type of things we were building and what we wanted to do with them. It was close to that Chapel Street experience. And he says, "well, we got their money noses to itch." Because now we started to see more and more people coming from down the road.

Well, anyway, some of the things I brought with me, I guess, may be shifting into electronics. First of all, it was a mechanical problem, and, secondly, the mule, which I think you've seen. We ended up with this little ULB, the Universal Logic Box is what I called it, and basically could build the fixture into this injection molded piece, and once we had a fairly defined length of this thing we could make injection moldings for either end of it. Now, it had to be two injection moldings, simply because the hold-down screw was not dead in the middle of anything, so we had to off-set it one way or the other.

And so, the concept which I'd been educated to for most of my packaging and design experience for that matter, is, if you can build the tubing into the part, then you've got a way of getting reproducibility that's fairly good. You've got other issues, obviously, but you certainly have the fixturing is there for you, and it's repeatable and so forth. And with electronics you can experience a lot of value changes, but you could usually, as a mechanical engineer, raise enough hell, so the sizes tried to stay the same, so that it gave a place to have a fairly stable base to execute the design.

Hal Laning got interested in this and we made a module. Hal had an interconnection program which would do the interconnection from the to/from list, which is the signal list, and that signal list had all the connections required for that module. And he could put it in his computer and come out with a matrix, if he knew how many lines along the wires, and how many vertical, he could make the pre-established points where they made an exit from this X/Y matrix.

And then it was a matter of cutting out the blanks in between appropriately. And this was fully automated. One thing he didn't have in that program is where those circuits were functional, so that when we fired it up, half a flip-flop would be in one module and the other one would be down the street, and when you got--what was it, the one--the width of that pulse was enormous, four micro-seconds, wasn't it?

ELDON HALL: I think it was one micro-second.

ED DUGGAN: Yeah. It was a ...(inaudible) mile.

DAVID MINDELL: What year was that?

ED DUGGAN: That was about 1961, because I came in '58 and I got involved in this in the early sixties. The first one I made, all the wires came out this way, and I had the problem, and it was designed around a Mark I guidance. And Cutliss, who was in charge--and I figured this was my last day at the instrumentation lab --we got hypodermic needles. I was the biggest buyer of needles, until the hippies showed up, and we put them down in and you could weld through them on the bottom, and then squeeze them on the top and put solder on them. And then came the problem of what do you do about those wires that kept jumping out?

So we put RTV on either side and then just poured epoxy down it, which immediately made all of those tubes insulators. And we then went to a dentist's office and got an abrader and abraded off the epoxy so we could make connections to the computer. And the last time I saw this, Steve Cutliss was pounding it back into the back end of one of these Mark I cases, and they'd had a photographer from the internal media photographing that we had installed it in the Mark I case, therefore the funding would be given for it.

And all I wanted to do was die because we'd stacked all the memory together and it looked like a Dow Jones average, going up and down across the top of it. It was the worst mess I ever saw. At that point, we had to have a better way, and that's when I discovered Malco? connectors, because they were being used on the fire control, and that was a gift from Heaven. And when we got to the Apollo we were able to get those cut in half in their size so we could shrink them, too.

So Laning made a major contribution there, and then he got more and more interested in how you do wiring like this and he soon was able to contribute dramatically. Another person that came to MIT by virtue of flunking out twice, I think, was one of the brightest guys I ever met, was Bob Morse. And he had the ability to program these matrices such that we could run automation in the test equipment--in other words, we could do continuity testing and so forth. I remember we went out to the wire wrap people out in Wisconsin and then could give them a deck which we validated ourselves for these techniques, and they could take and wire wrap.

BARD TURNER: It was Michigan. I've got all the features of that program that Morse wrote, which is a very significant--

ED DUGGAN: It was all on punch cards, which--

DAVE BATES: You should get Morse to resurrect one for you.

ED DUGGAN: So we had the automation for accuracy, that was one of the things. We knew that, with all this complexity, we had to have something other than people checking it out. And secondly, the wire wrap reproducibility that we could replicate, and pretty soon we found that we could drop signal nameplates over the back of these planes and pop the internal part, to keep the wires from moving around.

We got into trouble with wire wrap once, and that was in the factory, and it was at Raytheon, and it turned out to be, basically, fumes in the air were attacking the polymer coating on the wire, and they just broke up into rings, and I don't know if you remember that problem or not, but it was the cleaning fluid that was near--

DAVE BATES: Broke up into rings?

ED DUGGAN: Yeah, little ringlets, along the polyester that was extruded

ELDON HALL: Was that Apollo?

ED DUGGAN: It wasn't in the production, but they were doing the wrapping for us, because we didn't have any capability within Draper. Part of that, we were using the source in Michigan. I remember having a debate in that early program, too, about why welding would someday make soldering obsolete, and as soon as we saw that was wrong, we ended up with multi layer boards, and we put the appropriate metalization on the top layers so that we could make weld joints. And the same thing we had to do with resistors. So we did have to modify some of the parts. Fortunately, there were only, what, three resistors, four resistors in that switch, and a diode, and the lead switch, were already--that was ...(inaudible)

__: Covar (?).

ED DUGGAN: Which allowed us--I remember getting the president of one of the companies--because I only ordered three types of resistors and just wanted them forever, and he came in, and we realized what program it was, and again I'm blocking on the name of the company.

ELDON HALL: It was IRC, wasn't it?

ED DUGGAN: No, it was a big resistor company, but anyway, it doesn't make any difference. But when he found out what program we worked with--

DAVE BATES: Bradley?

ED DUGGAN: Bradley, Allen Bradley. I got to find out what you guys take for these memory blocks.

DAVE BATES: They were in Huntsville.

ED DUGGAN: And he came in and he was so impressed on what we were working on, for whom, in this garage, basically, when we were in Building 5, he said he’d make sure we got them. Once he spoke, we got all the resistors we wanted. We then could dictate to you people, we don't care what you use for logic as long as it includes this parts list, and that's all you get, because we can't afford to develop the other one.

So it was that type of merging with the mechanical things. I also remember the incident of the core logic, and at that point in time Apollo had gone to the lab. I remember going to these ...(inaudible) and thinking that these guys are crazy but I'll go along with it because in case that ever happens. And they couldn't have it all housed in any place, so we moved to what was called the Apollo Hilton. I'm sure somebody remembers that name. And the building was empty, it had just been vacated, but they sent, naturally, the mechanical group down there, so it could be by itself and not bother anybody anymore. And we started to lay that out. And I remember I was one of the people that complained vigorously about how large this computer would be if we had to do this, because I frankly didn't know how to handle it. And that, plus the logic issues, we ended up with NOR logic, which basically made it very much an upgraded derivative of work done for Trident. That covers it broadly, at least the area that I was involved in.

DAVID MINDELL: Can you say a few words about your work after Apollo?

ED DUGGAN: After Apollo? By that time, the group that I had, the group had grown quite large, and I think at the time when Apollo was in full bloom, it was about 140 people, in that range. We had test labs and things like this, potting labs and lots of draftsman, but the strategy for going to that size and then shrinking was always to have residents from all of the participating companies send man power in and then we could send it back.

I had a similar program that I left--the navy program went on, and I left my senior engineer in charge of that, and that program was running into a problem at that point in time, and I got called back and asked to work on it because I was basically not doing very much on the Apollo side. Things were done on Apollo, basically. And so I moved back onto navy programs and was there until the Pershing and the ballistic missile bases. And after that we became the group that was the supplier for packaging on all of the -- I can't say all of the equipment, but we certainly did all of the navy equipment, at that point, for ...(inaudible) and for the laboratory, and we did flight electronics with any of them, and then pretty soon it was for the entire laboratory, after we reorganized.

But then when Russia started to prove to be an "unreliable enemy" things started to slow down and it looked like the last four years I wanted to retire reasonably early anyway, but I'd spend that time laying off people that got overcome by the economics of the society. I retired in '92. But it was the greatest place in the world to work, in that timeframe. That's it.

DAVE BATES: With all the high technology you're talking about here, one of the things that always mystified me when I went down to Waltham was the Dixie-cup on the wire that came out of the veil, the wire going into the wire wrap thing. Can somebody explain to me why we had a Dixie cup, other than it kept the wire going in there in a straight way? And who came up with the idea about a Dixie cup?

JACK POUNDSTONE: It was probably a Raytheon person, I'm sure.

DAVE HANLEY: Ed, do you know why?


DAVE BATES: That was very interesting. I mean, every one of those ...(inaudible) wire wrap machines, and we had ten of them, finally, when we were through, and nine of them were for commercial types of application. We only use one for Polaris and Apollo, but every one of them had a Dixie-cup on the thing, and I asked the guys, why is it Dixie-cup? And the guy says, it works.

ED DUGGAN: That's a good reason.

DAVE BATES: I thought that was a good one, too. But who came up with it? We don't know. Maybe one of the girls did it, or somebody else.

BARD TURNER: I think that goes back to Grand Haven.

DAVE BATES: You think so?

BARD TURNER: Yeah, because I was involved with Grand Haven when that was the old--

DAVE BATES: And you didn't have a big one or a small one or a plastic one or -- what it was, it was just a Dixie-cup. But that's a high technology thing.

ED DUGGAN: It's like the plumber's code: it's what he knows and doesn't tell you.

DAVE BATES: That's right.

Next: What made Apollo a success?

Home: Conference 3