Oral History Transcript — Drs. Edward Byram and Robert Kreplin
This transcript may not be quoted, reproduced or redistributed in whole or in part by any means except with the written permission of the American Institute of Physics.
This transcript is based on a tape-recorded interview deposited at the Center for History of Physics of the American Institute of Physics. The AIP's interviews have generally been transcribed from tape, edited by the interviewer for clarity, and then further edited by the interviewee. If this interview is important to you, you should consult earlier versions of the transcript or listen to the original tape. For many interviews, the AIP retains substantial files with further information about the interviewee and the interview itself. Please contact us for information about accessing these materials.
Please bear in mind that: 1) This material is a
transcript of the spoken word rather than a literary product; 2) An interview
must be read with the awareness that different people's memories about an event
will often differ, and that memories can change with time for many reasons
including subsequent experiences, interactions with others, and one's feelings
about an event. Disclaimer: This transcript was scanned from
a typescript, introducing occasional spelling errors. The original typescript
Access form | Project support | How to cite | Print this page
See the catalog record for this interview and search for other interviews in our collection
Edward Byram and Robert Kremlin; July 31, 1987
ABSTRACT: Development of early detectors and how they were used, including the BS-1, BS-212, BS-2; development of a gas-gain ionization chamber; Solrad series detectors, solar radiation research; ionization chamber detectors; collimators; development of stellar ultraviolet detectors; discovery of Scorpius X-1 and Crab Nebula X-ray sources; HEAO (High Energy Astronomy Observatory) detector; determining the aspect of a rocket, computers used.
DeVorkin: We're here this morning with Mr. Robert [W.] Kreplin and Mr. E[dward] T[aylor] Byram, both of the X-ray astronomy group at Naval Research Laboratory. We're here to examine some of the detectors that are representative of the detectors that have been used in the X-ray astronomy group over the last twenty, twenty-five years. What I'd like to do, starting out, is bring some of the earliest detectors onto the table and ask you each to identify what they were used for, approximately when they were used, how they were built, and
what kinds of observations you made with them.
The first one would be what we call the original BS-1, which I understand is representative of the earliest types of counters that were developed by the X-ray astronomy group in the early 1950s. It's marked
with a "3" on the side. I was wondering Mr. Byram, could you possibly lead off and give us a feeling for how this was made and what you used it for?
Byram: Really, Kreplin used those a lot more than I. I wasn't too intimately connected with the radiac problems.
DeVorkin: The radiac problems. Possibly you could —
Kreplin:When I came to the laboratory in 1948, '49, I worked on these. These were built by Amperex, I think, at the time, then Anton [Electronics Company] later. This BS-1 was a mica window Geiger counter.
DeVorkin: If we could possibly hold it about here. Is that all right? So they can see the inner part. Just hold it there.
Kreplin:And the filling was chlorine and argon, I believe.
Kreplin:They were used in instruments called radiacs, which were radiation-monitoring instruments for use aboard ship. The window is mica, and it was transparent to beta particles, as well as gamma and X-ray. But I don't think it was very sensitive to alphas; the alphas were stopped by it. What we did with these, mostly, is just to measure their plateaus and measure the drift with time, just set up and measured one tube after another. That's what you do as a summer student.
But about the same time, Dr. [Herbert] Friedman was experimenting with these and determined that they were sensitive, of course, to X-rays and to ultraviolet radiation. I think it was Dr. [Talbot Albert] Chubb
that began working with different mixtures of gases in order to define specific wave length bands in the X-ray and ultraviolet region.
DeVorkin: What was your contribution to this project with the radiacs?
Kreplin:Oh, I was just running the machinery to measure the plateaus and things like that.
This is while you were a summer student?
Kreplin:Yes, and developing some little circuits for pulse operation of these, to try to get them to operate in very high radiation fields.
I'd be interested. This was a commercially produced tube, as you said, by Anton.
DeVorkin: And looking back at the mica window, if we can get a close shot of that — that's the right angle for it — what is the seal used for sealing the mica tube? What is the substance of the shell? Is it copper?
Kreplin:No, this is stainless steel. I guess one of the reasons for using stainless steel was not only was it inert to any chemical reaction with the chlorine filling, but one could also use the soft glass to seal the mica window to it.
DeVorkin: I'm interested in the procedure for sealing or for the construction of tubes like this. When Dr. Friedman and his group built tubes as prototypes, did you follow any different procedures for sealing, for constructing the devices that didn't translate well into the actual production of the devices? Did Anton have to alter the seal or alter the techniques in making these things to transfer them? Because I understand thousands of these were made when they went into production.
Byram: I don't think there were any alterations necessary. I think everything worked just the way it was designed the first time.
DeVorkin: Did the tubes work as well in production? Did production tubes work as consistently as the tubes that you made in your labs?
Byram: Yes, I think so. We all worked in the vacuum-filling facility that we had. We all filled tubes and tested them while they were still on the vacuum system. What I'm not sure of is I don't remember whether Anton filled these or whether we filled them all at NRL.
Kreplin:I think they came filled. Certainly in later days they came filled. As I recall, you know, if a tube was good, you know, it stayed good, but if it had a leak or something else was happening to it, it went bad fairly rapidly, and I think we tried often to repair them.
Byram: Yes, I think that's right. And we would rerun the plateau curve once a week, for a while, anyway.
Byram: And especially these that we used in the rockets. We'd run those daily.
DeVorkin: Yes. In your testing program, in the radiac program, were you responsible for testing production tubes that Anton made?
DeVorkin: And you found them to be of similar quality to what you were making in the lab?
Kreplin:Well, when I started, I wasn't really making any of these tubes in the laboratory; I was just running these plateaus. Most of the things I was doing, measurements I was making, were on the commercially built tubes. Later on, we began building tubes for the rocket, and there I worked on the filling systems as well.
This tube was not very adaptable to the rocket launch, because it has a long anode, which is anchored at this end and is free at the other end. It's a long thick wire that has a little beaded glass on it. Of course, that would not survive very well under a high vibration environment. So we went to this type of structure for most of the rocket Geiger counters.
DeVorkin: That's the side.
Kreplin:Yes, to a side window.
DeVorkin: Okay. We'll get to that. Now you've just answered the next question, of course, and that is what kind of modifications did you have to go through in making these things adaptable for rocket research. So certainly they look different, but functionally they're just the same; they're just adapted to the rocket. Could you take us through, Mr. Kreplin, a description of what happens physically inside a side-fire tube?
Kreplin:Okay. All right.
DeVorkin: If that's the right way to call it. I can give you a pointer.
Kreplin:This is a section viewed through this tube, and the window through which the radiation, X-ray or ultraviolet, enters is right here, mounted in this large flange. The anode wire is connected between the two insulating caps here, using a spring to keep it under tension. Because of the cylindrical structure and a voltage placed on the wire here, you have a non-linear field. That is, the electric field becomes much stronger near the wire. The way this works is that a charged particle or X-ray comes in, [and] produces an electron-ion pair. The ion is carried back toward the negative shell; the electron is accelerated toward the anode wire. Because the field is changing very rapidly, increasing very rapidly toward the wire, soon that electron gains enough energy in the field to produce, by collision with other ions, additional electrons. So you have an avalanche formed.
In a Geiger counter, that avalanche propagates along the whole wire by virtue of the fact that those collisions also produce ultraviolet radiation, which, in turn, ionize other parts of the gas species here in
the tube. That's about the way they work.
DeVorkin: Now, this clearly has the wire supported at both ends.
DeVorkin: It's been ruggedized. It's also smaller overall than the original tube. Going back to the artifacts now, is there a reason that the chamber is smaller? Did you find that you didn't need as much gas? What governs the actual size of this chamber?
Kreplin:I think there's some relationship between the diameter of the outer shell, the wire diameter, and the length of the tube. I think what you're looking for is a uniform field along the whole wire, so that [with] any particle forming, an ion pair here sees the same field throughout the length. So one wants to guard the ends here so you don't have high field concentrations at the ends. I don't recall exactly how you arrived at that, but those, I think, were some of the general considerations.
Byram: I think it was pretty arbitrary, and I think one reason for having it this long was the avalanche would be longer and you'd get a bigger pulse. They're made of stainless steel partly because of the high work function, because you don't want visible light to come through and hit the cathode and give you photoelectrons from the cathode. All the action you want should take place in the gas.
DeVorkin: We're looking at two of the early side window detectors here, and they certainly look different. This lighter one here, which is, I guess, directly modified from an Anton design with these — are these ceramic end-caps here?
DeVorkin: Looks quite a bit different than this fellow over here. Why do they look different? Is this one made in the lab here, and this one adapted more from an Anton tube? What's the history of these two?
Byram: They made many, many, many different models of the same kind of tube. I don't recall the reasons for having a copper cathode. It doesn't seem like the right thing to do at the moment. They must have had to passivate the cathode.
Kreplin:The stainless steel was passivated, but yet I think for some of the fillings it was necessary to have an oxygen-free cathode, and for those fillings, they used the copper.
DeVorkin: This one has a lithium fluoride window, I understand.
DeVorkin: The copper one. And is more representative of the ones that you flew in [V-2] 49 in the V-2 era, I understand.
Byram: Well, on V-2 49 we had aluminum windows, beryllium windows, and lithium fluoride and calcium fluoride.
DeVorkin: So you had a whole array.
Byram: Yes. In fact, we even had a quartz window, so that we spanned all the way from 3000 angstroms to a few kilovolts.
DeVorkin: Yes, and you were using different combinations of the windows and quenching gases to get different band passes. Let me ask you about the construction of these two tubes, if I can just move them
carefully this way so the camera can pick up the connections here between the wires and the tube. Now, I take it these have to be insulating connections in both cases. They're very different. This one looks like a glass plug over here on the copper one, and this one is a much more standard-looking ceramic. Now, how would you say these differences came about? This one was definitely made in the lab?
Byram: I'm sure it was.
Byram: The glass and metal seals like that are standard, a standard product that our shop carried. We had a wide variety of those, so we could make up many different configurations of tubes. We didn't have any stockpile of ceramic seals.
Kreplin:These were Kovar glass seals, and they could be soft soldered into a shell without destroying the seal by heating them in a very uniform way. So it was easier in the laboratory to build a tube with this type of seal than it was to buy the specially made ceramic, and put it into the tubes. So just for testing and development, we'd usually do that sort of thing.
DeVorkin: Was there any difference in performance that you would have to account for in trying to predict what the production tube would do?
Kreplin:No, I don't believe so. I think they operated pretty much the same.
Byram: I think so, too.
DeVorkin: How much contact did you have with the people who were actually producing those seals and building the tubes? Were these all done in shops that were independent of your group, or were they within your group?
Byram: They were in our group. We had our own glass shop.
Kreplin:In the Laboratory.
Byram: And the filling system was there.
Kreplin:But some of us also, you know, would build some tubes.
Byram: Yes, everybody was an expert in those days.
DeVorkin: How did you learn to be an expert? Because you didn't learn how to make these sorts of seals, I take it, in class somewhere.
Well, I never tried to build a tube, actually, but I used to try different fillings, and...
Kreplin:Well, in the laboratory, in those days, anyway, you could always find somebody in the laboratory who knew how to do something you wanted done, and generally, you'd go and talk to them, and they'd tell you how they did it or give you some advice and some help. So you'd go back and try it yourself and fail miserably, and then decide to take it to someone who knew how to do it.
Byram: We had our own glass-blower. He was strictly a technician. He'd do anything you asked him to, but he didn't know how to make Geiger tubes, really, he just knew how to put things together.
DeVorkin: Is it fair to say that you learned these shop practices by watching these specialists?
DeVorkin: Let me ask one other element of the design here. As I look closer here, I can see that this is your filling tube here, not the anode.
DeVorkin: And that's pretty clear. This is the anode — this is the filling tube that I'm familiar with from most of the tubes we've been looking at today. But where is the filling tube here, actually, on the production model?
Kreplin:It's right here. It's a little glass tubulation.
DeVorkin: Oh, I see. What was the advantage of having a glass-filled tube over copper? I know that a number of yours were done from glass.
Kreplin:Well, I think in this case, I don't think you wanted any copper in this particular tube. It would have been impossible to solder a copper tube in there and seal it that way.
Byram: In the early days, I don't think we had any pinch-off tools. I think all of the early tubes were glass.
Kreplin:Yes. Right. The way this was sealed was with a rivet tool that was adapted with special hardened jaws so that you could squeeze it right down and actually pinch that tube down, and the end would just fall off.
Byram: It would sort of weld it.
Kreplin:Weld the end together. Then, you see, the solder was put on not to seal it, but just to reinforce that feather edge.
DeVorkin: Okay. Is there anything special about the nature of the seals here? I see they both use glyptal. One uses a lot more than the other, and one has a square — that's lithium fluoride, I take it, a square piece, whereas this aluminum is just a very small piece.
Kreplin:Yes. A pinhole. The lithium fluoride and calcium fluoride we usually bought in chunks and cleaved it ourselves. Since it cleaves in a square configuration, that's why the windows go on that way, and they can be ground, turned and ground in some cases. Later on, that was done, but this is a much simpler way, quicker to put the window on.
Byram: At least in the early days, it was considered a disaster to get any moisture onto a lithium fluoride window.
Kreplin:So you used a freshly cleaved surface.
Byram: Yes, freshly cleaved and never touched it. Never tried to wash it.
DeVorkin: Lithium fluoride was soluble in water?
Byram: Very slightly, and it spoiled its UV transmission characteristics.
DeVorkin: Okay. Keeping those there just for a minute to give another contrast, I'd like to move to another early tube that was adapted from an Anton design, and this is, I guess, what was called a BS-212, Anton counter shell that was converted to a UV corona lamp, as you've told me before.
Kreplin:Ted, you know more about those than I did.
Byram: I don't know much about them. I don't remember.
DeVorkin: They're quite small.
Byram: But we made all different sizes and shapes of tubes.
Kreplin:These were made primarily for checking out the instrumentation on the rocket. We needed some source of ultraviolet. In the laboratory, I think we had little mercury arc lamps in quartz shells, and that would go down to about 1800 or so angstroms. I think that's where quartz cuts off. To get lower into the wave length regions where calcium fluoride and lithium fluoride transmit down to about 1000 angstroms, we had to build our own stimulating source, and that was one of the things this was used for. It had a lithium fluoride window on it, and an anode. High voltage was placed on that and, I believe, hydrogen was used in the tube.
Kreplin:And produced a glow, a discharge in the hydrogen, and then you could see the radiation from that discharge through the lithium fluoride window.
Is it safe to assume, since this is an end-fire tube and I can see the very tiny anode wire in there — which is unsupported, of course, on this end — that this was not made for any high- acceleration or flight, and it was, indeed, as you say, a laboratory device?
DeVorkin: Okay. I know in a previous interview, one of your groups did say that tubes like this were flown later on in Solrad or something. I don't recall exactly which one.
Kreplin:I think that [James] Van Allen flew some tubes like that.
DeVorkin: Okay. Let's move on, then, to the experimental Geiger tube that looks to be far more rugged, even though it is an end-fire tube, but it is shaped like a pill. Now, from what we know of it already, I understand that it was flown on Viking, on the early Viking series, and it was designed as a pill because that allowed it to be nested together, a bunch of them to be nested together and flown as a bank. Were you involved to any degree in this particular project?
Byram: Yes, we all were, of course. When we went out to the field, we took several hundred of these tubes with us, and they all had to have their thresholds measured daily and plateaus measured. They were going bad daily. We also took our glass-blower with us so that he could refill them as fast as they went bad.
DeVorkin: I see.
Byram: At the time that the Viking was launched, I think we even had to use some dummies in position of some of those tubes. I don't know why they were so unreliable. They had another characteristic that was not good. There's a ring anode in there.
DeVorkin: A ring anode. So there's no wire here?
Byram: No wire, but it's ring-shaped.
DeVorkin: I see.
Byram: And it had the peculiar characteristic that light coming in diagonally gave a lot more response than something coming straight in. So it had a very peculiar angular response. We had some others that were windowless, and they had a similar problem for a different reason. There the gas came out the pinhole — we used a pinhole window — and they were free-flow. They weren't sealed off like this one. Then the gas coming out absorbed the radiation before it got into the detector for radiation coming straight in.
Kreplin:I point out that this angular response characteristic was quite important, because the attitude of the rocket — Ted got into this later in having to determine aspect; and fly aspect sensors in the rocket — but if your detector wasn't fairly well behaved, you had some uncertainty built into the actual calculation of X-ray or
DeVorkin: This is, I take it, the anode here, the anode contact. What is the other protrusion over here that's just smooth?
Byram: That's another support for that ring-shaped anode.
DeVorkin: I see. The other end of the anode is connected inside there. Okay. Is there anything else we need to know to understand this device? It sounds like it was one that was not too successful.
Byram: Well, the Viking wasn't too successful. So as far as results from a flight go, we didn't get any good data on this from any flight that I know of. The Viking went up, and it was supposed to be a controlled spin with a spin period of something like six seconds, and instead of being controlled, the control jets stayed on continuously, and the rocket just kept spinning faster and faster, and finally it was spinning so fast it threw vacuum tubes out of their sockets and everything else. Everything quit before we really got out of the atmosphere, even.
DeVorkin: Okay. The next counter is identified as a BS-2 counter shell with a special window. It is the first attempt, I understand, to adapt a large area crystalline window to a metal shell — a lithium fluoride window in this case. This was made circa 1952-53. Our record says the seal was made by you, Mr. Kreplin, using silver spinning with silver chloride seal.
DeVorkin: Could you discuss, in some detail, how you went about doing that?
Kreplin:I think, as I recall, the reason for wanting a large window is that we wanted to fly some experiments at night and look at aurora and air glow in the ultraviolet. The small windows in the Geiger counters did not give us high enough sensitivity, but the problem was that the coefficients of expansion of the lithium fluoride and the stainless steel or copper were very, very different, one from another. So if you cemented a lithium fluoride window of one-inch diameter like that, with a few temperature cycles the seal would fail. And in a rocket flight, of course, you have fairly wide temperature extremes, due to the skin heating and the coldness of the high atmosphere.
So I found a reference to some work that some people had done in putting large area windows — I think rock salt windows — in infrared devices that had to be cooled to liquid nitrogen temperatures, and so I adapted their design for this particular purpose. So this end is a silver spinning made from about 10/1000ths silver. It comes out and then is folded back. On the inside, a crystalline lithium fluoride window is placed. This had been ground to the appropriate diameter in a billet, then cleaved off, sliced off one window at a time. But before slicing, that billet was coated on the outside with liquid bright platinum and baked to give a platinum surface. That allowed us, then, to put a small ring of silver chloride right at the junction between the window and the silver, and then heating it in the furnace in an upright position like that. That silver chloride would flow and seal the window to the silver, and in this particular case — this was just experimental — we just took that window assembly and cemented it to the tube with epoxy resin. It worked, and we were able to put it into liquid nitrogen, and it did not leak.
These were flown in, I think — a version of this was flown in 1953 up in a balloon-rocket combination on a trip up to Thule, Greenland.
DeVorkin: Up in Greenland. I am curious that this does have the anode that is always supported at one end.
Kreplin:Yes, that's right.
DeVorkin: So had you solved the problem of ruggedness? Is it a thicker anode?
Kreplin:The anode's fairly thick in this, and, of course, it's much shorter, too.
DeVorkin: Possibly you can get another shot of the anode. Is that all right? Okay. I'd like to stop now just for a few minutes, okay?
DeVorkin: This must have been a successful design for a counter.
Kreplin:Yes. Yes, it was.
DeVorkin: Can you identify any specific flights or types of information that were gathered from counters like this?
Kreplin:This particular counter, in another form, actually, geometrically it was the same, but it had a flange, so it was mounted in the side of a rocket, a small rocket, launched from a balloon in 1953, I think it was. It was looking out at the night sky. But I think the tube went into saturation for some reason. Either it was too sensitive and saw the aurora or there was some electrical malfunction. So as far as I recall, there was no data actually acquired with this particular type of tube, but it formed the design basis from which other types of detectors were built.
DeVorkin: If it didn't gather any data, then how do you know it was successful? How do you classify it as a successful tube?
Oh, just in our laboratory testing. The seal was the successful thing.
DeVorkin: Now, what did you have to change to make this a working tube?
Kreplin:Well, to determine the wave length response of the detector here, we would use the crystalline window. Lithium fluoride would pass Lyman alpha, but then it was necessary to use another type of filling; chlorine argon wouldn't have worked in this tube, because the chlorine would have attacked the silver, so it was necessary to use another type of quench agent. I think that one of the things that was used was nitric oxide, was it not?
Kreplin:We used nitric oxide as the quench [agent], because that could be ionized by Lyman alpha quite readily. Nitric oxide, unfortunately, reacted with, as I recall, the oxygen on the surface of the stainless steel. So it was necessary to build some detectors with oxygen-free copper. Also, because of that reactivity to traces of impurity in the detector body, the stability of these tubes was not really very good. We had to do a lot of filling to get a few tubes, and their lifetimes were not all that great.
DeVorkin: Another thing I notice about the tube is that it does have what looks like a factory ceramic insulator there. Of course, it has the glass filler plug. Was this modified from a production Anton counter, just chopped off, or what?
Kreplin:Yes, I think they were.
Byram: I wouldn't know.
Kreplin:They just used whatever parts were readily available insofar as the seals and all and the terminals and tubulation. I think the shell itself was probably made according to the laboratory design. But I think that short one was kind of a standard shell.
Byram: It might have been. We made so many different sizes and shapes, and I don't remember.
DeVorkin: Do you recall tubes like this one, you’re working with it? Did you make the seals also?
Byram: No, I didn't.
DeVorkin: How did you convince everybody that that seal was a good seal?
Kreplin:I just went ahead and did it, and that was one of the rather nice things about the laboratory. We were pretty free to follow whatever ideas that we had. We just pretty much took a problem and found a solution to it.
DeVorkin: This is a counter. As I understand it from the record we have now, this is an ionization chamber that was derived in its design from that prototype counter. This has a lithium fluoride window as well, with a silver chloride seal. Is this, then, an adaptation of the first seal?
Kreplin:Yes. That used the same technique. In this case, the silver piece was simply stamped. We didn't find that it was necessary to use the spinning. It made it much simpler to manufacture. The silver chloride seal was used in the same way it was here, and everything was assembled to a copper body. This was a Lyman alpha ionization chamber, and it was designed to be used with an NO (nitric oxide) filling, to be operated as an ionization chamber.
DeVorkin: What does that mean, the difference between a counter and an ionization chamber?
Kreplin:Well, in the ionization chamber, you have no gas amplification. For one electron-ion pair produced, you collect one electron, and so it's necessary with that type of device to have a very sensitive electrometer amplifier. The ionization chamber, though, is considerably more stable in its characteristics than the Geiger counter. A Geiger counter is very sensitive to impurities, to changes in the composition of the gas filling, due to any chemical changes going on in the tube. The ionization chamber is not so sensitive to those things.
DeVorkin: These were both used for solar work, I imagine.
Kreplin:This was a night sky one. This was used for solar.
DeVorkin: So the ionization chamber is fine for solar work, but you need a Geiger counter when you do stellar work.
Kreplin:Or a gas-gain ionization chamber, and those were what Ted developed.
DeVorkin: We'll get to some of those pretty soon. Now, this one is even shorter than your prototype. How did you finally decide upon an operational length?
Kreplin:Well, it's a function of the gas pressure running inside and the wave length of ultraviolet that you wish to absorb. You'd make it only as deep as you need to absorb the radiation. It wouldn't really be necessary to build a tube any deeper than that.
DeVorkin: As you say, this was a prototype. This is labeled as an operational model.
DeVorkin: Were they still made in the NRL shops, or did you ever contract out to have some company make this?
Kreplin:We did contract to a company called Melpar, to manufacture those detectors.
DeVorkin: Where is that company, and how did you get to know them?
Kreplin:Well, they were over — in fact, the building still exists. I don't know whether it's still in business or not, but they were over at about the intersection of the beltway and Route 50 [Falls Church, Virginia]. I don't remember exactly how they got into this.
Byram: Some of those companies got into it because they hired people from NRL that we knew. I know in one case there was somebody from the vacuum tube model shop, a glass-blower went to one of the companies, and we followed him.
Kreplin:Yes, that's a possibility.
DeVorkin: Could it have been the case in the Melpar company?
Kreplin:I don't recall anybody that moved over there from that, but my recollection is that one of their people was looking for something to build — I don't know. I just don't recall well enough who it was or when it happened. But they expressed an interest in building and developing the techniques.
DeVorkin: The first silver chloride seal you did in about '53, did you say?
DeVorkin: And this is an ionization chamber based on that design, and that was then representative of what was flying in the mid-fifties? Is that an accurate period of time, approximately?
Kreplin:Well, late fifties. I think, as I recall, we flew that in Solrad 1 in 1960.
Byram: We flew it earlier than that, though, because we flew them — I know that I used to put them in [Donald MacGregor] Packer's rockets.
Byram: I used to put one ion chamber in each one of his to monitor Lyman alpha.
DeVorkin: Now, the next chamber is a gas-gain ionization chamber, as it's identified, and it is labeled also as flying on Aerobees in the late 1950s and early 1960s; that puts it pretty much in that time frame that you mentioned. It looks different than the ionization chamber in a number of ways. I take it this has the same seal here, same silver chloride seal. Is that correct?
DeVorkin: So everything in the front end is the same.
DeVorkin: There seems to be a glassiness, though, right outside of the window itself. Is that part of the crystal?
Kreplin:No, that's silver chloride. And oftentimes we just painted over that silver chloride a very thin coating of epoxy, because silver chloride is somewhat reactive to the atmosphere.
DeVorkin: So what is the very shiny, actual silvery surface above it? Is that part of the paint?
Kreplin:The darker surface?
DeVorkin: Yes, both the lighter silver and then the darker surface.
Kreplin:Oh, around here?
Kreplin:Oh, that dark is just tarnish, tarnish on the silver.
DeVorkin: So is that tarnished silver chloride?
Kreplin:No. Out here it was soldered.
DeVorkin: That's the solder.
Kreplin:But I think that's just tarnished silver there.
DeVorkin: How is this as a gas-gain ionization chamber? How are they different than straight ionization chambers? Maybe Mr. Byram, since you apparently worked with these?
Byram: Well, they acted a little bit like a Geiger tube. It was more of a proportional counter than a Geiger tube, but there would be gas gain as the electrons were attracted to the anode. You would get collisions and further ionization. We operated them at a point where they had a gain of about 5,000. Generally, we operated the shell with a negative voltage. This was insulated from the frame of the rocket, but because there was a high voltage difference, we had to have a ceramic insulator at the rear here. This actually acted as a guard ring, also, to collect leakage current.
DeVorkin: This is coax inside here?
Byram: No. This shell acted as a guard for leakage from this to anode.
DeVorkin: I see. Did you use these solely in Aerobees, or did you also use these in the Solrads?
Byram: I don't recall using them in Solrad.
Kreplin:No, I don't think so.
Byram: They may very well have been used in something other than Aerobees, but we always used them in Aerobees, in our ultraviolet Aerobees at night.
Kreplin:There was a telescope on a Ranger, too, where you used those.
Yes, that's right. That was the Ranger that was built — the instrumentation was built by Cal Tech [California Institute of Technology], and they did something wrong. They didn't like measuring a negative current, so they reversed the voltage on the cathode in order to make their electrometer easier to build. The result was that there was no gas gain. It acted as a straight ion chamber.
DeVorkin: They did this because it was easier for them to design, actually, and then they didn't tell you about it?
Byram: No, they never told us until two days before it was to launch. Actually, that was a day after they tried to launch, but the launch failed. So they called us on a Friday and said we'd better get down there and find out what was wrong; our detectors had all gone bad. That's all we supplied for that experiment was the detector.
DeVorkin: So they did the collecting surfaces?
Byram: Yes, they did all the electronics and the mechanical apparatus. It was an experiment that involved a collector mirror of about ten inches in diameter, and it was going to go part way to the moon. Maybe it was going to orbit the moon and look back at the earth and scan across the earth to look at the Lyman alpha glow around the earth that we knew it was there. But I went down there to see what I could do, see why the tubes were going bad, and Dr. Chubb stayed at NRL and made new tubes, filled them. And halfway through the morning after I got there Saturday morning, I was accidentally poking around with the volt meter, and all of a sudden the volt meter jumps off scale in the wrong direction. I asked them why was that. They gave me the sad story about being easier to work with a positive voltage than a negative voltage.
DeVorkin: So the instrument that you're talking about is this very shiny and expensive-looking device here, I take it.
DeVorkin: And this detector here went right at the prime focus?
DeVorkin: Right there. And it was pointing down?
DeVorkin: To gather the collected radiation for that geocorona experiment. That's right. Okay. Now that we're talking about probes and moving into that era, we're beginning to look at non-solar detectors, but also larger solar detectors. What I'd like to do is move on to a Solrad series at this moment that is, I guess, another larger gas-gain system for what you call the free-flow X-ray detector. It will just take us a second to regroup here, put these back so we don't get them mixed up. Now we'll bring on these components.
Now, this is a detector of a totally different order, and it would be very hard for me, at least, to relate the technology that we have been talking about, the small, closed systems, to what we see here in Solrad 10. First of all, I would like to ask you how did you come to decide upon producing these free-flow gas-gain detectors, and is this a typical example of what one looked like? Were they always this large, or did you have smaller ones in the beginning?
Byram: We had larger ones in the beginning, actually. This was small because of Solrad. But after our first night X-ray experiment, where we used modifications of those gas-gain detectors, the gas-gain ion chamber type detector, we had an assembly of those. I don't know, maybe 21 detectors arranged in an array, all in parallel.
DeVorkin: You're saying these little ones.
Yes. We had a whole array of those, just looking outside of the rocket in our first X-ray experiment. We had two different arrangements of those. One of them worked, but they were very difficult to prepare, and soon after that flight, Dr. Chubb decided that the way to do it was to build a counter more or less on this shape, with square-shaped cells, each one inch by one inch, with gas flowing through the counter, with a window of this material to retain the gas and to allow the X-rays to come in. The detectors were built in two sections, a front section to look at the X-ray data, and a back section to collect cosmic rays to use in an anti-coincidence circuit, to cut down our background. And in the first ones, these ground planes here were solid, but built in such a way that the gas would flow in a labyrinthian pattern, and eventually flush the whole counter out. The filling was —
DeVorkin: Go ahead.
Byram: I don't know what the filling was. I've forgotten what that is. Oh, I do know, too. It was P10; it was helium and 10 percent methane. We also had argon and methane, and xenon and methane to cover various X-ray band widths. The wires were strung across here from each insulator, and there was a 5 mil stainless steel wire used. The wires aren't installed in the bottom, but there were identical wires in the lower tray and in the upper tray. This arrangement, then, was used in all of our X-ray experiments.
DeVorkin: On Solrad, or even beyond that?
Byram: On our Aerobees, also.
DeVorkin: On the Aerobees as well. Could you identify the different pieces now? This is the tray that has this thin film on the top. Was there a collimator on top of that?
Byram: Yes, there was a honeycomb collimator, in general.
DeVorkin: It would have looked something like this?
Byram: Yes. This really is the final flight configuration. This didn't fly; this did. The gas would flow in here
and out through here, through a fixed leak.
DeVorkin: So this is the input for the gas and the output for the gas.
Byram: Before takeoff, we had a solenoid valve that would allow us to flush the counter out so it was filled with fresh gas at liftoff. This is the gas tank that went with this experiment, and that held enough of the counting mixture to last something like three years.
DeVorkin: What are we really looking at in this gas tank? Clearly, the sphere is the part that holds the gas, but what is this entire apparatus over here, a valving system?
Byram: Well, it's a valve system and it's a pressure regulator. This is the pressure regulator — the initial pressure in the tank was 500 psi, but the counter operated at close to an atmosphere. [Interruption] Then the solenoid valves were used to initially flush the gas and finally to just maintain the pressure.
DeVorkin: Was that a standard system that you procured, or is that something that you designed within the shops?
Byram: It was designed by one of our mechanical engineers and was built in our shop.
DeVorkin: Okay. I think we'll end this tape right here, just to be safe, and we'll take a quick break and reconnoiter about the Solrad and discuss it.
DeVorkin: We've been talking about the gas-gain ionization chambers, the free-flow chambers for Solrad 10. I would just like to have both of you recall a little more about the experience of having the gas-flow system designed and built and tested and actually flown. Was it a simple procedure to make these systems available, or did you find that it was quite complex and there were problems associated with it?
Kreplin:It was pretty complex. I don't recall whether you did a free-flow experiment before OSO-B, the satellites, or not.
Byram: I don't think so.
Kreplin:But the OSO-B experiment was a free-flow Geiger counter experiment with a manifold of about five counters in a chain with a calibrated leak and gas supply. It was designed to operate about six months in orbit. In order to do that, it was necessary to store the gas at a relatively high pressure and then reduce the pressure to about one atmosphere to go into the Geiger counters. Before that, a pressure regulator was required which would work in a vacuum. We first bought some specially built regulators from I think it was Aero
Equipment Company, and found that they did not work satisfactorily. We took them apart and found considerable problems with the manufacturing technique and cleaning. We spent a lot of time cleaning them and trying to get them to work. We finally gave up and went to another company, Air Reduction, who had another type of regulator, which we bought and adapted to use in the OSO spacecraft.
DeVorkin: Where is the regulator on this device?
Kreplin:This is the regulator here. Let me take this off and turn this around. This is the pressure regulator here. It's a two-stage regulator.
DeVorkin: This is it right here?
Kreplin:This, I believe, is the high pressure end; this is the low pressure end. And of course, the way they were furnished, we could not — didn't use them. What we did was to use the body. We made a metal bellows instead of a rubber one, and made some other changes in them also, especially, I think, we replaced O-rings with Viton and did some other things like that.
DeVorkin: What is Viton?
Kreplin:It's a rubber compound, but it's especially impervious to any kind of chemicals. It has better operating properties over wide temperature ranges. So I guess that was the last experience I had with regulators, and then Ted was responsible for building this experiment.
DeVorkin: What was your experience, then, with the regulators here as they were built for Solrad 10? Did they work all right?
Byram: They worked fine, as long as we rebuilt them before we used them.
DeVorkin: What does rebuilding actually require?
Byram: Changing a rubber diaphragm here and cleaning.
DeVorkin: How highly do they have to be cleaned?
Byram: Just ordinary techniques.
Byram: Alcohol or one of the other — trichloroethylene, one of those solids.
DeVorkin: Couldn't you specify that from the manufacturer, ask the manufacturer to do that? Or you couldn't trust them?
Byram: We just never bothered to do it. We wanted to make sure that everything was okay, and we were going to take them apart and look anyway, so we took the responsibility of cleaning and put no requirement on the manufacturer. And they worked fine. This one had a modification that some didn't. This one has a — you can see a little hole at that end in the very center, and that's a vent that converts it from an absolute regulator to a relative regulator. That allowed me to operate this counter on the ground. It might be over an atmosphere, but after it was in orbit, the pressure dropped to about half an atmosphere.
Kreplin:Why did you have to go so low?
Byram: Just to get a good voltage operating range. I didn't want to have too much high voltage. Lowering the gas pressure lowered the voltage. I think it operated around 1200 volts.
Kreplin:This was a particularly difficult problem in doing space research from rockets and satellites, the use of high voltage in spacecraft. We had lots and lots of problems with that over the years.
DeVorkin: Were you able to test this system in a vacuum here in the lab? Just as we saw in one of our earlier interviews, you had that large chamber for the HEAO. Did you have chambers to test the membrane in here against bursting, the entire procedure?
Byram: Yes. Yes, we had a small vacuum chamber we could put that in. It allowed us to actually fill the tube after it was in the vacuum.
DeVorkin: Did you go through a complete operational cycle with it in the vacuum chamber, feeding the gas in and then sensing the characteristics of the detector?
Byram: Yes, I think so. I think we used that gas tank and supplied the gas and operated the solenoid valves,
flushed the counter out, filled it. And we had a radioactive source in the vacuum so we could see that we were getting a good signal from it.
In integrating this into a Solrad satellite — Solrad, of course, was an NRL satellite — was everything done then in NRL? Within building 209, did you actually bring the satellite in and integrate the equipment yourself?
Byram: No, we integrated over in Satellite Techniques area.
DeVorkin: Where is that?
Kreplin:That was on the south side of the mall, in one of the buildings. Fifty-nine?
Byram: Fifty-nine, I think, was the one.
DeVorkin: Okay. Is there anything else about this particular — oh, yes, I know. The collimator itself. We should talk about that collimator. What is the purpose of that?
Byram: It serves two purposes. In the first place, the thin film has to be supported, and this collimator accomplishes that. In addition, it defines the field of view of the detector as it scans across the sky. It has about a 7 degree by 7 degree field of view.
Kreplin:Where did you get the materials for the collimator?
DeVorkin: Is that this material, the mylar again that we're talking about?
Byram: Well, that's the mylar. The mylar is 1/10th mil, which comes from Du Pont [E.I. duPont de Nemours and Company].
DeVorkin: Is there special treatment on there? It doesn't look like it's quite clear.
It's metallized with about 50 angstroms of nichrome, and that's to provide a conducting film on the inner surface to complete the geometry of the counting circuits in there. I don't know who supplies the honeycomb.
Kreplin:The honeycomb, though, is another example of using materials which are originally designed for something else, but adapted to use for our purposes. A lot of the spacecraft structure is built on honeycomb, which is essentially a panel made up of this honeycomb material between two sheets of aluminum, and they're bonded together. But the honeycomb itself, Ted's been using that as a collimator for quite a few years.
DeVorkin: So there's another honeycomb deeper in?
Byram: No, no, it's just this honeycomb.
What actually determines, then — oh, I see. The honeycomb is about a quarter of an inch deep. There are walls that are about a quarter of an inch thick there? From here it looks like a wire, but in fact, are these a bunch of little cells?
Byram: They're a bunch of cells, and the cells are maybe an inch — the honeycomb comes down to this surface right here.
DeVorkin: Oh, it's much deeper. I see. It's quite thick. Yes, I see. It's very hard to tell. If I move it around, can you see? Can you pick up from the camera that there's a depth in there?
Byram: That may have been coated with soot or something.
DeVorkin: It's coated with soot?
Byram: It's coated with benzene soot.
DeVorkin: Is that on purpose?
Byram: Just to make it optically look better. It allows it to radiate heat out better.
DeVorkin: These are straight mechanical collimators and the length of the collimator gives you the angle of acceptance.
DeVorkin: Okay. What are the electronics over on the side? What do they represent?
Byram: They just supply high voltage to the wires, and the high voltage comes in here and is distributed through it. Then the output is through a condensor and a resistor combination down here. This is a high voltage filter down at this end.
DeVorkin: Why is it encased in that amber-looking material? What is that material?
Byram: It's C-7. It's an epoxy that just seals it against high voltage breakdown.
DeVorkin: So that's a potting, a clear potting compound.
Byram: Yes, clear potting.
DeVorkin: It makes it a lot prettier than the opaque potting material. What was the reason — was it esthetic for going from one to the other, or is this a better compound?
Byram: Partly it's esthetic; partly it's a good idea. You can look in there and see the quality of the potting operation. You can look for voids.
Kreplin:Yes, that was not only important that there be no voids in it, but also that the material itself didn't outgas and produce contamination.
Byram: The epoxy was prepared in a vacuum chamber. It was pumped on to get rid of air that was dissolved in the epoxy.
Byram: That was quite a long and delicate procedure.
DeVorkin: Did the idea of using this clear epoxy come to you internally in the group, or was this something that was a regulation substance that was decided upon, let's say, by the satellite people who were integrating the satellites and building them, worrying about them? Do you recall?
Byram: I don't remember.
Kreplin:I think that, you know, the material was really very desirable from our point of view, and as I recall, we didn't have much trouble here at the Laboratory. [But] but when we began working with NASA [National Aeronautics and Space Administration] and other people, they had a lot of materials people that insisted on telling us which epoxies were acceptable and which were not, and there were very few that were acceptable. This Armstrong C-7, with the activator W that we used, was one of the ones that was acceptable. Before you could use one that they didn't know about, you had to do a lot of tests, vacuum tests on it, testing weight loss and deposition of contaminants and that sort of thing. But I think we've always used epoxies.
Byram: Some of our very early rockets, we used some sort of foam potting.
Byram: That stuff was not very good.
Kreplin:Yes, yes, I remember that.
Byram: That gave us all kinds of trouble.
Kreplin:I remember that. That was light, though.
Byram: Yes, it was light.
DeVorkin: One final question about the gas chamber itself. Did you have to go through explosive tests or burst tests on this chamber, even in the early unmanned Solrad series? Did they want these chambers to be tested?
Byram: No. Any tests that we wanted to do was fine, but they put no requirements on us.
DeVorkin: How did that change?
Byram: That was entirely NRL. It was internal to NRL entirely. It was a much relaxed system compared to NASA.
DeVorkin: Did NASA require the burst tests?
Kreplin:Currently, I think, to —
You had to design things four times their actual operating pressure and sometimes they had to survive — in fact, they sometimes burst the equipment deliberately.
DeVorkin: We'll talk about that soon enough when we get to HEAO [High Energy Astronomy Observatory], where we have a burst chamber here. Okay.
I think it would be a good idea to move backward in time to discuss an earlier series in Solrad. Solrad was a very long series. It started in 1960. The first flight was in '60. And when was the last flight, the approximate number?
Kreplin:1976, and it was turned off — Solrad 11 was turned off in October of 1979.
DeVorkin: The first Solrads, of course, were quite small, and Solrad 10 was not one of the original 20-inch NRL spheres.
Kreplin:That's correct, yes.
DeVorkin: But just to get a comparison, these are representative detectors flown on early Solrads; I take it, some of the ones we've seen already. Now, why did you scale up from these early detectors to the larger detectors in the later Solrads? What was the need for that?
Kreplin:Well, these detectors are really for different purposes. The original small detectors were designed for doing solar studies, and the larger detectors were night-sky stellar work.
DeVorkin: Oh, I see. I assumed that most of what Solrad was doing was always solar, but then you're saying that you — started doing stellar work as well.
Byram: Yes. It was called Stelrad, at least in our branch.
DeVorkin: It was called many things — Solrad, Solwind.
Kreplin:No, Solwind was the experiment that Don Michaels flew on P78-1.
DeVorkin: Oh, okay.
Kreplin:And that was quite a different experiment.
DeVorkin: Okay. Then why don't we then change the menu here. I'll move these out of the way. We'll talk about the small detectors and how they were used, what they were used for, and we'll go, also, to the photographs.
Kreplin:Well, we attempted, in the mid- and late 1950s, to measure solar flare X-ray radiation with small rockets. Our first attempt was to fly a rocket hanging from a balloon, and then allow it to drift along until a flare was observed on the sun, and then fire the rocket. That was just partly successful. Most of the time we lost the rocket because the balloon drifted out of range and the rocket had to be fired.
Later on, we went to use a Nike-ASP, a Nike small rocket combination. That way we could keep the rocket on the ground and fire it whenever we saw a flare occurring. But it always took a minute or so
to get to the altitude where it could begin measuring, making measurements, and so we were never able to see the beginnings of a flare. We always saw the tail end, or the flare that lasted ten or fifteen minutes, we always lost the first part of it.
So when satellites became available, it was a natural thing to put detectors aboard satellites, and we put a set of detectors on Vanguard 3 and on Explorer 7. Neither of those experiments was successful. Although the detectors worked all right, what we saw were the Van Allen Belt particle radiation, and it swamped everything.
DeVorkin: This is the beryllium window that was used on Explorer 7 and Solrad 1 as well, is that correct?
DeVorkin: Okay. So that's the one you're talking about.
Kreplin:And on Vanguard 3.
Kreplin:But the difference between Solrad 1 and the reason it was successful was because we put a large magnet in front of this detector, which swept out the soft electrons of the Van Allen Belts. This is a photograph of Solrad 1 with some of the group from, I guess — they changed their name several times, but I think at that time they were called Satellite Techniques Branch.
DeVorkin: Satellite Techniques Branch?
Kreplin:Something like that. Anyway, Marty Votaw was the head of that branch at the time. This is Ed Dix, who took over after Marty left. Paul Lester was one of our technicians. Arny Unsinger was one of the scientists working on the experiment. The other people, I think, are technicians and engineers from the Satellite Techniques group.
DeVorkin: Was this group within Herb Friedman's group or separate from it?
Kreplin:No, this was separate. This was a separate group.
DeVorkin: This group is not in the [E.O.] Hulbert Center.
Kreplin:That's correct, yes.
DeVorkin: Who did they report to, and why was it separate, do you know?
Kreplin:Who did they report to? I've forgotten at the time who it was.
DeVorkin: Was it considered more of an NRL-wide facility?
Kreplin:Yes, I think they were under Code 7000. We were also under Code 7000, as I recall, so we were two separate divisions. But you better not press me on the details of that organization because it's changed so much.
DeVorkin: But these are people you worked with.
Kreplin:Yes, that's right. This is the 20-inch diameter Solrad satellite which was derived from the Vanguard design. The X-ray detector is right here in the belly-band of the satellite, and it has this little square opening, which is the opening in the magnet. The antennas fold up and then come down at launch. Now, I just want to show another shot of the satellite there. That's another one.
DeVorkin: Which one would you like?
Kreplin:This one. It's kind of in order. This is a shot, a little larger, showing the Solrad 1, and you see the technician here is irradiating that detector with an iron-55 source to provide a signal which is being read out here on a strip chart recorder. Now, this was all analog circuitry and analog telemetry, analog recording. Things aren't done that way anymore. That's a later one. Let me have the other one that was coming next there. I'll keep going there. That's Solrad 3. That's it.
I just wanted to show you the way in which Solrad was launched. It was generally not launched as a single satellite, but in this case it was part of a package, and I believe that was a transit satellite which was an early Navy navigation satellite. We were mounted up on top and in the nose cone of a rocket. They were different rockets at different times. I think Thor-Able-Star was one that was used, and I don't know that that was the one used for Solrad 1. Let's see. Let me show Solrad 3. We have some pictures of Solrad 3, which came afterward.
This shows a better shot of the magnet, which protected the detector right here. The satellite was launched and spun up, and there was no active control of the orientation of the satellite, so it spun freely and it precessed slowly. We would inject it so that the spin axis was perpendicular to the direction to the sun, so that the detectors which were on the belly-band always saw the sun once per rotation. But because it precessed, it was necessary to put in aspect detectors, and I think that we used aspect detectors that Ted had developed for rockets, and they also were located here. They can't be seen in this particular shot, but it was necessary to telemeter both the aspect information, as well as the information from the X-ray detectors.
Byram: I think I went down to the Cape one day and calibrated one of those aspect detectors.
Kreplin:Let me see. Here's another shot of some people. Yes, I just wanted to show one. This is a picture of the young Peter Wilhelm, who is now currently head of Code 8000, now the Spacecraft Technology group. Anyway, things have grown a bit. Some of the people are the same. I think that's Bob Beal, who is still working
at the lab. I think that's all. We can discuss the detectors next, if you wanted to, although this one is kind of interesting. Yes, there's kind of an interesting story in Solrad 3. Solrad 1 and 3 were pretty much the same, with an exception of changing the detectors somewhat. But Solrad 3 was launched, again, as part of one of these stacks of, I think, a transit satellite. This is the satellite built by the state university of Iowa [Iowa State University] called Injun. This is Solrad 3.
DeVorkin: There are three piggybacks.
Kreplin:Yes. Solrad 3, and the Injun never separated, so that made the moments of inertia wrong, and it had a very complicated spinning and precessing motion. At that time, Loren Acton was working at the laboratory as a graduate student, and as part of his thesis work, he was going to use all the data from Solrad 3 to write his thesis. Well, when it didn't spin, he had a great deal of trouble. I remember one time when Loren cracked the code. He had hung up all the telemetry records, the analog telemetry records, down one hall, across the end and all the way up the other hall, and he'd walk back and forth in front of these, trying to figure out what the satellite was doing to produce the aspect signals, and he finally deciphered it and was able to write his thesis.
DeVorkin: Did he use only the solar signal or could he find other signals?
Kreplin:Yes, he used the solar signal and the positions of the X-ray signals, too.
DeVorkin: This was Solrad 3. That was still before you really had too many stellar sources. Did you have Sco X-1 by then?
Kreplin:None of our detectors on these early Solrads were sufficiently sensitive to see any of the sources.
DeVorkin: So what other source would he have seen other than the sun?
Kreplin:Just the sun.
DeVorkin: Just the sun. Okay. Did you help him with this aspect problem?
DeVorkin: Was that one you didn't want to attack yourself?
Byram: I had other things to do, I suspect. I don't think I knew Loren Acton that well.
DeVorkin: Okay. Let's move on to talking about the detectors themselves. We'll start with the beryllium detector. Here's a blow-up of it.
Kreplin:These are ionization chamber detectors, and I think we've discussed what the ionization chamber was. This is the earliest version of the ionization chamber built here at NRL. We actually designed this here at the laboratory and had our machinists cut the aluminum. We ordered the beryllium from Brush Beryllium Company. That was cut to proper size, since we did not — we were aware at the time of the hazards of beryllium, so knowingly we didn't do any cutting or grinding of beryllium. I think earlier, though, we probably had cut some, broken it.
So it's very simple, just a simple diode, [that] maintains a voltage difference between the anode and the cathode of about 45 volts, and then measures the very small current produced when ionizing radiation produces the electron-ion pairs inside the detector. This is a larger picture of it. The beryllium carried one atmosphere of argon. And because the beryllium was fairly thin, I think this was 5 mil beryllium, 5/1,000ths of an inch, and it was somewhat fragile, we put this strongback along here just to support the internal pressure. This was sealed with epoxy, but you know, because it was operated as an ionization chamber, it didn't really matter that you got a little admixture in the argon of some gas evolving from the epoxy. It worked just as well. The wave-length sensitivity of the detector was determined by the transmission of the window and the absorption of the gas. And this particular beryllium transmits out to about 8 angstroms and then falls off and becomes opaque after that. The argon begins absorbing radiation strongly at about 2 angstroms, so gradually it then peaks around — what is the K edge of argon, 3.8?
Byram: That sounds right.
So it peaks up about 3, drops off 3.8, and there's another peak at the K edge, and then falls. No, it goes on up then, but the product of the two, the window transmission and the gas absorption give you a curve which peaks at about, oh, 4 or 5 angstroms. So it becomes a band-sensitive detector. We used the same technique for a number of bands. This is the only X-ray detector that was flown in Solrad 1, but in Solrad 3, we included also a detector which had an aluminum window. [This window] is a thin film of aluminum, which is a quarter of a mil thick supported by a mesh. That would transmit out to about 16 angstroms, and we used, in this one — gosh, what was that? I've forgotten the filling we used. It could have been argon at a lower pressure. But anyway, the transmission was the same, defined in the same way. It's probably in one of those publications.
The purpose of Solrad 1, which was the first successful satellite we had, was to fly a Lyman alpha ionization chamber, which we discussed. Yes, that's the one right here.
DeVorkin: We can set this up on the chart. Can you focus in close enough on this?
DeVorkin: The lower right.
DeVorkin: Take it in as far as it will go.
Kreplin:It's this one right here. This portion of the sensitivity curve is determined by the absorption of the argon, and then this portion of it is the transmission of the beryllium.
DeVorkin: And that's how the combination of the two would give you your band pass.
Kreplin:Yes. In this particular one, we only flew that one detector, anyway. But just to finish the story on Solrad 1, we flew one Lyman alpha ionization chamber.
DeVorkin: That's the copper one.
Kreplin:That's the copper one, and one X-ray ionization chamber protected by a magnet. The satellite lasted for about six months in all, operated okay, and that was the first time we were able to see the beginning of a solar flare in X-ray and in Lyman alpha. And we found that the enhanced atmospheric effects were due to the X-ray radiation and not the Lyman alpha radiation.
DeVorkin: Let me ask about the aluminum face. It's got a grid on it.
DeVorkin: What is that grid for?
Kreplin:The grid [was required] because the aluminum is so thin, you'd rupture the aluminum foil with even, you know, small pressure inside, so the grid is electroformed nickel mesh, [and] supports the thin foil of aluminum.
DeVorkin: That seems to be a continuing problem, supporting these very, very delicate, thin films. Okay. I think we can move, then, from the Solrad, since already when we looked at Solrad 10, we looked at a collimator, a crude one, to be sure, the hexagonal honeycomb. I think we should move now to some of the collimators that you developed at about that time. In fact, you had other collimators that you flew on earlier flights, on Aerobees, as well. So why don't we stop now and talk about collimators for the next discussion. We'll take a short break.
DeVorkin: We've been talking about collimators for different X-ray detectors that have been used in the past by the X-ray group at NRL. What we've set up now is a short demonstration of how these collimators work. I'd like to first look at an X-ray detector with a collimator that's in the center of the table, that is typical of ones flown by the group in, I understand, the late sixties and early 1970s on Aerobees. These are typical large X-ray detectors for stellar work, for non-solar work. If you could lift up this first one over here that's right in the middle, we could take a look at the honeycomb structure.
DeVorkin: This is, as I understand it, the original type of collimator that you used that also acted as a support system for the mylar, is that correct?
Byram: That's right.
DeVorkin: Okay. We'll look at photographs of these later, but I understand now, as you continued to fly these detectors; you put collimators, more precise collimators in front of them.
Byram: For this type of detector, we did that. We would use the collimator and we'd use a pointing control with the rocket and cause the rocket to scan across a small section of the sky. So we would look at a band of the sky 15 degrees broad and maybe 20 or 30 degrees long.
DeVorkin: Now you said that the resolution of the honeycomb was on the order of about 7 degrees, but you wanted better resolution than that, better angular resolution?
DeVorkin: Let's put this back here and move now to this other collimator. While I hold it up here, before we actually go through the demonstration, it's hard to see, but it's got —
Kreplin:You forgot to turn your light on.
DeVorkin: We won't turn it on yet. You can see some lines that are coming down this way just vaguely, but there are very fine lines that are going perpendicular to it. What is that kind of a structure?
Byram: They're on the bias, because there were two detectors, one above the other, and this had the bias set that direction, the one below it had the bias set in the opposite direction. So you would scan down through the sky, one detector sweeping out a path across the sky that was tilted like that, the other one would be tilted like that, and so you could tell the position relative to the center line of the scan where the source is you're looking at.
DeVorkin: When we turn on this back illumination here, we're going to be taking a look at diffuse light streaming through this collimator. Could you tell us actually what we're seeing?
Byram: Well, you're seeing the transmission window of the collimator. The light — if you were back in infinity, you would find that this whole collimator would be illuminated, but since you're so close, it looks like you're seeing only a very narrow piece of the light. But it's because of the finite size of the aperture over there and the distance and the band width of the collimator. The collimator has about a 1-degree field of view, so you see a 1-degree band of light.
DeVorkin: But how does this demonstrate the geometry of the collimator?
Byram: Well, I've just described it, really. You can see it has a one-degree field of view. Since your eye is at a distance, then you will see a band of light that's just one degree wide, and at the limits, it's attenuated compared to what it is at the center. But if you were looking at this at a great distance, like a couple of hundred feet, you would see that this band of light fills the whole collimator.
DeVorkin: But if I were to tilt the collimator back and forth at infinity, then the band of light would switch on and off?
Byram: It would fade on and off.
DeVorkin: It would fade on and off. Now what kind of geometry of wires in there does this kind of function?
Byram: Well, the sheets are photo-etched with very regularly spaced lines, and each line allows the opening to be about 80 percent of the total spacing. Then there may be a dozen sheets in there, and they're spaced in a geometrical progression such that you do not get light leaks, and it makes a very lightweight assembly.
DeVorkin: Did you produce these from the photo-etching right to the mounting of these plates in your lab?
Byram: No, we bought it. We sent the pattern to Buckby Mirrors, and they did the photo engraving, and we would assemble them in a fixture.
Kreplin:How did you assemble them to get them to register perfectly?
Byram: Well, there are alignment notches. There's one alignment notch here, and then on the opposite side, there are two alignment notches.
Kreplin:Oh, I see.
Byram: So that we had three knife edges that fit into those notches. That way they would all come out and register almost automatically.
DeVorkin: I see these bands here. Are these the layers, the actual different layers?
Byram: Yes. You can see that it goes up in a geometric ratio.
DeVorkin: That's right, the separations change.
Kreplin:Why, again, did you say that is in that geometrical ratio? Why aren't they equally spaced?
Byram: Well, if they were equally spaced, you'd get light leaks between the grids.
Byram: So each grid serves the purpose of blocking out one set of light leaks.
Byram: In addition to defining the field of view. And at least one of the solar experiments had a similar collimator, but it was 50 percent transparent, and this one is more like 80 percent transparent. With the 50 percent transparent arrangement, spacing the first grid was at half of the distance, and the next grid was a quarter and went on down like that.
Byram: Here the first grid is 20 percent of the total spacing; the next one is 20 percent of that, and so on.
DeVorkin: What are the electrical connections here, the very, very fine copper wires?
Byram: That was just to measure the strength of the whole assembly. There's a little stress-measuring gadget right here.
DeVorkin: Very tiny little strips.
Byram: Yes, it's a little photo-etched device that's a strain gauge. And you measure that stress with a Wheatstone bridge.
DeVorkin: This whole collimator, then, went on top of the detector in this fashion?
DeVorkin: Was there anything critical about the spacing here?
Byram: Nothing real critical.
DeVorkin: Okay. And then what is the nature of the detector? Was this a gas flow design again?
Byram: Yes, a free-flow proportional counter.
DeVorkin: But this has a box underneath it, and there was a place for what looks like another counter tube which is not there. What was that set for?
Byram: Well, on this particular assembly, that other counter was a photomultiplier, and this whole bath was surrounded with a quonset-shaped plastic box. The plastic box was material that scintillated when it was radiated, and that was used to get the counts for the anti-coincidence circuitry to cut down on the cosmic ray background.
DeVorkin: Okay. Is there anything else about this particular collimator that we should discuss before we move on to look at some earlier examples of collimators on photographs?
Byram: No, I don't think there's anything more.
DeVorkin: Okay. Why don't we turn to the photographs for a few minutes. Okay, we're finished.
DeVorkin: We've been looking at collimators, and we examined the Aerobee collimator in front of a light box. Here, if we now take a look at a photograph of the same type of collimator, we'll see, as Mr. Byram had indicated, what the illumination is like if viewed from a much greater distance. Possibly, Mr. Byram, you could describe what is in the photograph there?
Byram: Well, the camera is much farther back, and you're still looking at a one-degree field of view, but you can see that the light is spread over a much larger area of the collimator.
DeVorkin: Was this a typical test for a collimator like that?
Byram: Yes. We always did it for everyone.
DeVorkin: I'm curious why you took a photograph or a photograph was taken of such a large area in the entire laboratory and not, say, a telephoto on the collimator itself. Did that focal length of the camera lens have something to do with it?
Well, the actual test didn't — the photograph was just a matter of recording the fact that we had done the test. We didn't use anything from the photograph as part of the test. The test is just to get up there with your eyeball and look around and look for light leaks.
DeVorkin: I see. So this was a purely visual test.
DeVorkin: You didn't have some form of X-ray detector sitting back there looking for anything.
Byram: No. We have taken X-ray pictures that look similar to this, but mostly it was a lot easier to just use the eyeball and look up and down. You can scan your eye over the whole collimator.
DeVorkin: This is, as we said, a rather sophisticated multi-layer collimator. Possibly what we could look at now are two pictures of earlier types of counters and collimators that your group worked on. So if you could identify a little bit about this first one. What is the period that this is representative of?
Byram: Well, this is really the second generation of our stellar X-ray work. The first experiment was an assembly of cylindrical-shaped detectors, and one of them worked fine, but one of them didn't. These were Geiger tubes made similarly to the small ones that we used on Stelrad. They're just 1-inch by 1-inch cells, 6 or 8 inches long, each one with a 5 mil wire in the center of the cell, with gas freely circulating, starting at one corner and working its way up to an opposite corner.
DeVorkin: Was this representative of it?
Byram: Yes, that's representative.
Byram: But in this first version, really our second version, the ground planes were really solid instead of being photo-etched sheets.
DeVorkin: These are the ground planes here.
Byram: They defined the cells. But essentially, it's the same thing.
Well, the detector is divided up into a number of 1-inch by 1-inch cells, each one something like 6 inches long, with a 5 mil wire stretched at the center of each of these 1-inch by 1-inch cells. There were two layers. The two layers were isolated from one another with the ground plane. It's not in here.
DeVorkin: Are these the wires you're talking about, though?
DeVorkin: These stretched —
Byram: Well, no, no. The wires go between insulator and insulator. This assembly is not anywhere near complete. There's just a few pieces in here representative of what the whole thing was like.
DeVorkin: So the wire would be stretched between this insulator and this one, these two brass pins.
DeVorkin: What are these devices below, then?
Byram: They're just ground planes that define the rectangular area of the counter. And in this case, it was a proportional counter, but the only difference between the proportional counter and the Geiger tubes is the gas filling, the voltage at which it operates, and the amount of electronic gain you need to look at the data.
DeVorkin: Now, this is similar, then, to the early counters you have in the Aerobee there in the photograph?
DeVorkin: Okay. Let's move back to the photograph, then.
Byram: This rocket was used in the mode where we would allow the rocket to spin and precess and scan across the whole sky. We mapped the whole sky on several occasions.
DeVorkin: You mentioned that this was the laboratory at White Sands [Missile Range].
Byram: Yes, this picture was taken at White Sands, and it's a picture of the rocket after recovery. It's been up and down 100 miles or so.
DeVorkin: Could you point out some of the other devices there that you mentioned before?
Byram: Well, at the top is the magnetometer assembly, which we used for our aspect determination. The next two boxes are telemeters. Each of the telemeters was capable of displaying 15 channels of data. Battery boxes below that, some other miscellaneous electronics that I can't identify. This little circular dark area was an aspect detector, a photomultiplier that was sensitive enough to see the stars as we passed.
DeVorkin: Do you want to point to it? Is it that one right there?
Byram: Yes, that's right.
DeVorkin: Maybe you could point to it.
Byram: It's okay. Down there, that's a gravity-operated switch that has to do with activating the separation after — it's a G switch. I don't remember exactly what that — oh, I guess it activated the parachute assembly. No, the parachute assembly was activated with a pressure switch somewhere up in here. So I don't know. This little black cylindrical object is a rate gyro, which was also used for measuring the roll rate of the rocket.
Kreplin:You could point out the covers that were used to protect the counters on takeoff.
Byram: Yes. These doors would swing open on the action of a timer, it would open after burnout. The rocket would burn out after about 50 seconds, and then it would coast for approximately four minutes. That four minutes is where we got all our data. Four minutes of data would last us about six months, and we would go back and do it all over. The cover caused a little bit of damage. You see these dents in the honeycomb. The honeycomb was the total collimation on this experiment.
DeVorkin: These were all down after San Diego Hi and the other missions such as that, I imagine. This is already late fifties, early sixties.
Byram: These were in the early sixties.
DeVorkin: So you had detected some of the first X-ray sources with this type of equipment.
Byram: Yes. The rocket that went just before this one is the one that discovered Sco XR-1 and the Crab Nebula and a few others.
DeVorkin: Here's a picture of that one, I take it. Here's a good picture of the detectors. Possibly you could put that in there.
Kreplin:When you first did this, didn't you use bundles of hypodermic needles as the collimator at one time?
Byram: No, not on these.
DeVorkin: I think this is the best picture of the honeycomb.
Byram: Yes, let's use that. Well, I've got too many pictures here, I think.
DeVorkin: We can take those out. Now, this is a detector that would be one generation earlier, I take it.
Byram: Yes. This was our first attempt at measuring X-rays at night from the sky, and the detector was an assembly of these small cylindrical detectors that were derived in part from the gas-gain ion chambers.
Kreplin:That one, yes. That was kind of the aft end of this.
Byram: Yes, yes, that's all it was, a bundle of these all connected in parallel with a common window and a common gas filling, which doesn't show up well on here, and the honeycomb collimator in front.
Byram: Our first generation X-ray detector for looking at stellar objects at night was an assembly of cylindrical detectors all connected in parallel with a common gas filling and with a common plastic window — a beryllium window, in this case. The cylindrical parts were identical to the steel parts of this detector. It was operated as a Geiger tube. We had two of these. The other detector was slightly different, and the other detector, the one that didn't work, each detector was sealed individually, had its own individual window.
DeVorkin: Why don't we put up a picture of both of them behind the honeycomb?
DeVorkin: We'd now like to look at a series of photographs that identify the detectors that were used for two very famous discoveries, confirmation discoveries here, as well. That is that of the detection of the X-1 radiation source in the constellation of Scorpius, known as Sco X-1, and then in the X-ray detection of the Crab Nebula that was done first here by NRL people. Let's turn, first, to the Scorpius or Sco X-1 flight. Mr. Byram, Mr. Kreplin, possibly you could take us through how the detectors were used to locate that particular source.
Byram: Well, the rocket was allowed to precess and roll. The fins of the rocket were deliberately set to a specified angle to give it a clockwise roll of maybe a 5 or 6-second period. The rocket always turned counterclockwise, but other than that, it worked fine, because we always got a good roll rate. But after we were out of the atmosphere, the doors opened. You can see the doors here. The doors opened by a timer, and our high voltage went on about the same time. No, we don't need to point anything out very much. Well, I can point out these. These are gas-gain ion chambers at the focus of four-inch mirrors.
DeVorkin: We'll be talking about those next, after we finish this.
Byram: We can look at the data that resulted. This is what typical scans look like as we scan through the Sco XR-1. This is a section of a star map showing the location, which is roughly 16 hours and 20 minutes of right ascension and roughly 15 degrees declination. We scanned through it several times. Sometimes — these two are where the scan was closest to the source. We just more or less found the geometric center of that pattern of scans and put the "X" there.
DeVorkin: Was there a way to adjust the attitude of the spinning rocket so that you knew you were going to that source?
Byram: No, no, it was all accidental. But the precession, the rocket would go up and it would spin, and while it was spinning, it would precess in quite a large cone. In that way you'd scan across nearly the whole sky.
DeVorkin: You mentioned before, when you mentioned that the rocket always spins counterclockwise, Mr. Kreplin, I noticed that you sort of seemed to have a recollection of that.
Kreplin:No, I was just chuckling because Ted had said that the fins were always set to turn it clockwise, but it always ended up going counterclockwise.
DeVorkin: Why is that the case?
Byram: Who knows?
Kreplin:I don't know. [Laughs]
Byram: The spin was really determined more by the irregularities in the fins than anything else.
Kreplin:I know on one of the Aerobees that I was on, it was designed to spin, and turned out it went through its whole flight rotating only 5 degrees in spin.
DeVorkin: Virtually no spin at all.
Kreplin:None at all. Straight, just like an arrow. And for that reason, the flight was a failure, because we weren't able, in this particular time, to look at the sun as it spun.
Byram: Was this one of those where we were shooting a series of three at a time?
Kreplin:Yes, [NRL Aerobee flights] eight, nine, ten.
Byram: Yes. Well, there was always one of those when we would shoot three rockets at a time, and one of them always was very much better than the rest.
DeVorkin: Was there a reason for that?
Byram: It was all accidental.
DeVorkin: After you flew the original honeycomb detector that detected Sco X-1, your group decided, I guess, to produce better collimated detectors. Was that the intent?
Byram: The intent was more stable detectors with a larger field of view.
DeVorkin: With a larger field of view. So less spatial resolution?
Byram: No. Perhaps better spatial resolution. I do not recall the collimation on that first experiment, the angular resolution.
DeVorkin: But that was with the straight honeycombs that we saw with a lot of the little ionization detectors behind it. This had a field of view of 10 degrees, it says.
Byram: Okay. The subsequent ones were more like 7 degrees field of view, and we had trouble with every one of them. There were Geiger tubes that were very touchy. They had aluminum cathodes, and the cathode had to be passivated. The passivation was not permanent, and you'd get to the field, ready to shoot, and you'd find that the passivation had disappeared and you had to redo it, which mostly was just flowing nitric oxide through the counter. But it was a violent treatment, and it would leave the detector somewhat unstable for a time. Then there would be a period when it was very nice and well behaved, and then the passivation would be disappearing again, and then you'd go through the whole sequence of events again, if your rocket was delayed, for instance, which frequently happened.
DeVorkin: What is going on in this scene now? This is the Crab Nebula?
Byram: This is the Crab Nebula, and this is the vertical interference test — I mean the horizontal interference test.
DeVorkin: What are you testing for?
Testing our detectors to make sure they're working, testing the housekeeping functions to see that they're operating properly, testing the timers and testing the telemetry, testing for interference between our detectors and the telemetry, testing for cross-talk from channel to channel. The bottom section is the parachute for recovery, I believe. The next one up is a pointing control to point this — these are the roll jets that control the roll position of the rocket during the pointing maneuver. This I'm not sure of. It maybe is a beacon. That is an RF antenna of some sort. These two things are the telemetry antennas. Then our two proportional counters — not proportional counters, they're Geiger tubes again, with the same difficulties that I just described. And then the nose cone that covers the housekeeping instrumentation and batteries and telemetry.
This little circular thing is an aspect detector to see if we're pointing at the moon at the time that we encounter the occultation. The time of the occultation was calculated for us by the Greenwich Observatory in England, and it was calculated for the latitude and longitude of the position that we assumed the rocket would have at the peak of its flight, and it was calculated for the altitude of that at that time, which was really beyond our capabilities.
DeVorkin: What was beyond your capabilities?
Byram: The calculation of the point of where our rocket was going to intercept the occultation.
DeVorkin: So you worked with Greenwich.
DeVorkin: I'm curious why you used Greenwich and not the Naval Observatory.
Byram: At the time, Greenwich did all the occultations for everybody in the world.
DeVorkin: How does this detector differ from the earlier one? Is this a single detector, or is this still your array of small gas-gain tubes?
Byram: Its internal construction was very similar to the previous ones.
DeVorkin: So that there were a series of small gas-gain cylinders.
DeVorkin: Okay. So you hadn't moved on to the trays yet, the single trays.
DeVorkin: Okay. Possibly you can get a better picture of the collimator by looking at the next photograph.
Byram: Yes, you can even see the shadow of the ground planes that separate the cells.
DeVorkin: So now between the ground planes you would have each one of the detectors?
Byram: Well, yes. The wires are all really connected in parallel. They're individual cells, just to ensure that the gas is circulating.
DeVorkin: So again, that is the same as the open aluminum tray that we looked at before.
DeVorkin: So it is not a collection of different detectors; it's one detector.
Byram: It's one detector with a gas system that flows gas through everything.
Byram: And it's a thin plastic window. The rocket was fired on schedule as required by the occultations, but it rained that morning when we were in the tower, and water vapor goes through mylar without any difficulty at all. So we had to go through our passivation procedure, and things didn't look good, but in the end they were working, and we shot it. It went through the occultation just as it was supposed to, almost. It did scan past the sun, which they were trying to avoid, but somehow, the sun — we looked directly at the sun with our detectors during the maneuver.
DeVorkin: And it didn't blow out the electronics?
Byram: No, it didn't hurt anything. The only thing it might have done is to alter the composition of the gas in the counters by evaporating some of our passivation material or some such thing. But it was a short enough interval that nothing happened.
DeVorkin: The Crab had already been detected, is my understanding, but you were getting a lower limit on the angular size of the source from the occultation experiment, is that correct?
Byram: That's correct. We were looking to see if it was a point source in the middle of the nebula, or whether we were seeing just diffuse X-rays from the whole galaxy.
DeVorkin: Right. What I'm interested in, actually, is the following kind of question; you knew approximately what the sensitivity of your detector had to be as a result of that, but previously, for the Sco X-1 observation, and even prior to the AS&E [American Science and Engineering] detections, you had no idea what your sensitivity had to be in order to pick up non-solar sources. That you were going ahead and designing these detectors and looking for non-solar sources is certainly a fact. Did you simply try to make those detectors as sensitive as possible and still avoid the sun? This is what prompted this question; because you were trying to avoid the sun, you obviously had very high sensitivity. What was your thinking on that in, say, the late fifties, early sixties?
Byram: Well, we shot them at night, so we didn't worry about the sun.
DeVorkin: But in this case —
Byram: — except here, where during the occultation — the occultation occurred at something like two or three o'clock in the afternoon. We had no choice there, except that it was a controlled rocket, and we thought they could program it to miss the sun.
DeVorkin: So normally, you would shoot these all at night to avoid the sun.
DeVorkin: And that solved the sensitivity range problem. Okay, fine. I think we've finished with this sequence. What we'll do now is move on to a light-collection problem, which is similar but yet different from a collimating problem.
Byram: I wanted to say one more thing about this.
DeVorkin: Certainly. About the Sco X-1?
Byram: Yes. This same flight saw the Crab.
DeVorkin: Oh, I see. In '63.
Yes. It wasn't announced in this particular paper, but we did see it.
DeVorkin: So you detected a number of sources on that same flight.
Byram: Yes. There were sources in Cygnus and maybe in Ophiuchus, I'm not certain.
DeVorkin: Okay. Now we have to be a bit careful with how we move this concave mirror around so that we don't reflect light into the cameras, talking about sensitivity. What I'd like to turn to now is the collimated—or I should say, well, it's not proper to call this a collimated experiment, but in a sense —
Byram: It really is.
DeVorkin: This is a collimated experiment. I'd like to talk about the origin of your interest in looking at the ultraviolet sky, other than the sun, and how you came to producing this particular type of telescope operation. We should start by looking at these and then talking about their precursors, which are the tiny tubes, the hypodermic tubes that you used. But what is the main purpose here of both the tubes and of the mirrors? Are they primarily for producing a collimated beam or in collecting more light?
Byram: Well, we were interested in collecting more light. We were interested in having a focused image scan across our detector and get better definition of it.
DeVorkin: Starting with the hypodermic needles let me ask how did you decide, first, to use hypodermic needles behind your ionization chambers, and then make the transition to the larger collecting areas. Was this simply a stop-gap measure, something you could work on or produce very quickly while you were building the larger and more sophisticated systems?
Byram: Well, we didn't know what to expect, I think. We started off with something that we could do, and when we found that we could see individual sources, then we wanted to repeat it with higher sensitivity and with better resolution.
Kreplin:These were ultraviolet experiments or X-rays?
Byram: No, these were ultraviolet.
Byram: Anywhere from the cutoff of lithium fluoride up to 1500 or 1600 angstroms, with various windows, calcium fluoride, strontium fluoride.
Kreplin:I presume the needles, the hypodermic needles, were used because it was a ready source of a very small diameter tube made in a very uniform fashion. I think this is typical of using the kinds of things that are available and adapting them to our particular use, rather than having something manufactured specifically for that purpose.
DeVorkin: The same with the honeycomb layer that was used for constructing pieces of satellites and rockets.
Kreplin:Yes, and the gas regulator.
DeVorkin: In this particular experiment, then, your goal was to find non-stellar sources of UV radiation.
Byram: No, we were interested in stellar sources.
DeVorkin: I mean stellar, of course. Sorry. Who was the person in the group most interested in that? Was that you or Dr. Friedman?
Byram: Dr. Friedman, Dr. Chubb, and [James Edward] Kupperian, I suspect, at this time.
DeVorkin: Why don't you describe, then, how you moved from the hypodermic, what they showed you, how that helped you design the mirror-collecting system, [and] how you knew that 4 inches would be sufficient for what you wanted to do.
Byram: The reason we went to mirrors was that people at NRL had developed reflecting coatings that were reflective in the higher ultraviolet and made it practical to use a mirror. At the time we used the hypodermic needles, those coatings were not available.
DeVorkin: What kind of coatings are they, actually?
Byram: Multi-layer aluminum and — I'm not certain what the other — it was a combination of aluminum and insulators.
Kreplin:Mag[nesium] fluoride was one.
Byram: Yes, mag[nesium] fluoride. I'm not sure. It was Bill [William Ray] Hunter that did a lot of that.
DeVorkin: Were these mirrors made at NRL?
Byram: Yes. No, the mirrors themselves were made by 3B Optical Company in Mars, Pennsylvania.
DeVorkin: I don't know them. Do you know anything about their history?
Byram: No, except that they made good optics. I don't remember now how we got onto them. I know they did all of our mirrors.
DeVorkin: The basic configuration, then, you might — I'll try to hold this up, and as I understand it, the detector was placed right at the prime focus like that.
Byram: Yes, and then this filter went in between on one set of experiments. The filter would go in there like that, and this pedestal went down to a thermoelectric cooler, and the silver parts, at least, were maintained at a fixed temperature, at a low temperature. We had several of these, and we covered a range from minus 15 degrees up to 20 degrees. The calcium fluoride filter, its wave-length cutoff is temperature sensitive, so we were looking to try to get better spectral resolution out of our telescope.
DeVorkin: What we're putting in the tray now is a — I guess that is you. [Interruption] Okay. Now, what are you doing here? It looks like you were adjusting the filter, the stem of the filter.
Byram: I'm probably just attaching the connector to the back of the detector.
DeVorkin: Possibly you could point to where the mirror —
Byram: The detector is a gas-gain ion chamber similar to this.
Byram: The primary mirror is at the back end, back in here. There's a secondary mirror. The mirror's an off-axis paraboloid, and then there was a secondary mirror just slightly inside that housing that supported the detector. So it acted more or less like a Newtonian telescope.
I'm confused then. These pictures seem to show, if we look at a close-up —
DeVorkin: Oh, these are different instruments.
Byram: Different instrument, different rocket.
Byram: This was probably in 1959. That was our first.
DeVorkin: Okay. Do we have any of the earlier ones here?
DeVorkin: These are all later.
DeVorkin: All of this is later.
Byram: That shows where it was being installed.
DeVorkin: Okay. Let's put that one in next.
DeVorkin: The first versions, then, of this mirror system were Newtonians.
Byram: Yes, it was a one-time deal. We only flew it once. It was a piggyback on another branch's rocket, Dr. [Richard] Tousey's branch.
DeVorkin: Why did you choose a Newtonian design in the beginning?
Byram: It seemed simpler to me in terms of construction.
DeVorkin: It allowed your detector to be outside of the tube?
Byram: One reason was that all the high voltages were in a pressure-tight compartment. It was pressurized to an atmosphere.
DeVorkin: And where would that be in that picture there?
Byram: It doesn't show in here. We'll have to show this picture later to illustrate that. I guess we're really almost through with this. There's a high-voltage power supply, and then an electrometer on the top there. That's the total amount of instrumentation that was required.
DeVorkin: Are you about to place the mirror inside the tube there? Is that what you have in your hand?
Byram: Yes. I've forgotten. I had a little assembly there. I'm not sure what I'm doing there, but it must have been the installation of the mirror. Just how that was done, I'm not sure.
DeVorkin: The next one here, this shows the power supply better and also the chamber within which everything resides?
Byram: Yes. This whole assembly fits into that cavity, and the cavity was pressurized to an atmosphere, so this Newtonian construction allowed me to do all the high-voltage work at an atmospheric pressure.
Byram: Eliminated a lot of problems. The front end of this was covered by a metal disc that was rolled to the shape of the rocket relative to the same curvature, and that [cover plate] was held on by pumping the air out of the cylindrical part of the experiment. Somewhere in there, there's a little valve and a connection for a vacuum pump. So before launch, I pumped the mirror out, and that held on the cover plate. As the rocket rolls through the atmosphere, the pressure finally fell low enough so that the spin of the rocket just threw the door off.
Byram: No latches, no timers.
Byram: No explosive bolts.
DeVorkin: It's real simple.
Byram: Yes, and it worked.
DeVorkin: That was used once, you said, but then you moved on to a prime-focus system.
DeVorkin: Why was that? Why did you switch?
Byram: Well, it was impractical to try to build this newer assembly in such a way that you could pressurize it, so there was no advantage at all to using the Newtonian.
DeVorkin: By an assembly, you mean an array of more than one telescope.
Byram: Many telescopes.
DeVorkin: You put that in to see all the different telescopes. Here we can see the spider that's holding the detector, is that correct?
DeVorkin: And there are four of them.
Byram: There are four of them. This arrangement [also] eliminated one reflection.
DeVorkin: Are they all the same band pass, or are they different band passes?
Byram: They're different band passes. I believe that the larger telescopes were at longer wave lengths, like 1500 angstroms or so. I think they used — I'm not sure what they used for a filter, something that cut off around 14, 15 angstroms.
DeVorkin: Here is a whole series of mirrors, and here's Dr. Friedman with another member of your team. I thought we would finish with that one in this collection of photographs. You can possibly describe who that is and what they're doing.
Byram: That's Dr. Friedman and Joe [John] Nemecek. Joe is installing a mirror.
DeVorkin: How many times were these flown?
Byram: Three or four times; I'm not certain.
DeVorkin: Was the object to make a map of the ultraviolet stellar sky?
Byram: Yes, make a catalog, which was done, of hot stars.
DeVorkin: This was done at about the same time as Ray Davis and some of the other people at the Harvard College Observatory were getting interested in the ultraviolet sky. Was there any contact with Davis? I'm not sure if it's Ray Davis.
Byram: No. The only Davis I know is Leo Davis, and he was at Goddard [Spaceflight Center].
DeVorkin: So you had no contact with the Harvard people about the ultraviolet colors.
Byram: Not that I know of. We had a graduate student that did a lot of the work.
DeVorkin: What was your plan once you took this and produced this ultraviolet catalog? What was the next step in your mind or in Dr. Friedman's mind to take after the ultraviolet catalog was produced?
Byram: Well, I don't think I know.
Byram: I think all that was driven out of our mind by the discovery of X-rays in the sky, the night sky.
DeVorkin: So you dropped the ultraviolet work completely and concentrated on X-ray. Okay. Well, we've already looked, then, at a wide range of X-ray detectors, and we've noticed that as you move into the sixties and seventies, and you move to the gas-flow detectors, they get bigger and bigger as you move into stellar astronomy.
I think what we should do now is move on to the largest detector, if that's all right, the HEAO detector, and how it came to pass, and what your roles were. Now, I understand, Mr. Kreplin, that you did not take a role.
Kreplin:No, the Solrad satellite program occupied most all of my time from 1960 on, so we were working [on] these programs more or less independently. But I'm sure that Drs. Friedman and Chubb were aware of what was going on in each one. I think that Dr. Friedman's interest was primarily in the X-ray observations of the celestial sources, rather than solar.
DeVorkin: Let's take a look at a few pictures of what HEAO is, and then ask you, Mr. Byram, what your role was in its production. If we could put this open behind the plate there, that'll do it. We'll start with the left-hand image of the satellite itself, if that's correct. Mr. Byram, when do you recall you got involved in the HEAO project here?
Byram: It was in the early seventies. This version of HEAO started in 1973, and there was a previous version of HEAO that was much larger, and I don't remember when that started — either in 1971 or '72. But NASA decided it was too expensive and they couldn't afford it.
DeVorkin: Even if that previous HEAO was much larger, you're still looking at one that's far larger than — each of the detectors is far larger than anything you had made previous to that time.
DeVorkin: What were your concerns about scaling up?
Byram: Well, what you just said is not quite true. We made a prototype counter that was roughly 3 feet by 3 feet. We knew that it was practical to build a large counter. The things that might have been of concern were can you have all those wires and make them all have the same characteristics so you can parallel them and have a reasonably good proportional counter? And that 3 by 3-foot detector did prove that, in fact, we could build a big detector.
Kreplin:How many anodes were there in the 3 by 3 detector?
Byram: Maybe a dozen. They were on 1-inch centers, so there were probably two dozen wires, maybe a little more. There's a lot of lost space in there. You can't just say there were 36 wires. It had to be less than that. Somewhere between two dozen and 30, maybe, and it was a double-layered counter, so the back layer could be used for anti-coincidence. We had window material that worked well over that large area; that was another area that might have been of concern. We were able to support it.
DeVorkin: Was it similar to this mylar here?
Byram: Yes. It may have been off that same roll. That one roll — we still have that same roll of mylar, and it's very much better than some of the other material we've had. This one roll is unique.
Kreplin:What energy X-rays were you able to observe with these windows?
Well, almost —
Kreplin:You could go pretty soft.
Byram: Very soft, yes. Certainly much softer than a kilovolt, like maybe we could certainly go to a half a kilovolt. With luck and a strong source, we could go to a quarter kilovolt, which would be maybe the carbon K-alpha line.
DeVorkin: Was there anything in the production of these detectors, knowing that you were going to use several of them at once, that was a unique problem that you hadn't faced before?
Byram: I don't think so. The only unique problem that we hadn't faced was assembling a large collimator and having it all come out in registers. We had to find a source that could make them that large, the grids for the collimator.
DeVorkin: Are these essentially the same types of etched grids that we saw a little while ago?
Byram: Yes, but Buckby Mirrors had a limit on size, and the size they could make was 18 by 18, or something like that. We found a company in Orlando, Florida, that had very much larger equipment, and they were able to make the full-size collimators for HEAO. There were stress problems of supporting that much area over an atmosphere.
DeVorkin: Now we can look at the six detectors in a row. Who had decided that they were going to have two rows of three like that, that six-detector configuration? Is that something that Dr. Friedman decided here or that NASA decided by determining the dimensions of the satellite?
Byram: NASA decided, essentially. They dictated to us what they would let us put in. They knew we wanted as many as possible. We wanted twice as much as they would let us. They gave us six detectors on one side and one detector on the opposite side.
DeVorkin: And what was the reasoning for that? Because the six worked together but not that one.
Byram: Actually, there was a pair that worked together, then there were four that worked together, and then the seventh one was an independent experiment. It had a larger field of view and was to serve a different purpose.
DeVorkin: What is going on in these pictures? Is there anything useful to discuss? I'd be interested to know if any of these people other than you are identifiable. I know you're in the rightmost picture yourself.
Byram: Well, I can identify some of the people. This is Al Ishikawa that works for TRW. He was a mechanical engineer. I'm not sure who that is; that might be Don Broseau. And this is an inspector; Fuchs was his name. No, not an inspector. He was in charge of that area of TRW, the area where we did all our testing,
which was in this large room.
DeVorkin: How did you get along with these fellows?
Byram: I got along well with them. [Interruption]
DeVorkin: Most of these are TRW people. How did you get along with them?
Byram: I got along well with them personally, but they didn't like the way I operated, really.
DeVorkin: What was that? How was that?
Byram: Well, I was always getting into somebody else's business, and I was always going around poking at my experiment, doing things that they'd think I shouldn't have been doing. I frequently had to go in and flush the counters out. We had an external gas supply that we could use, and an external supply of a special gas, so it was necessary for me to get at my experiment frequently. At least once every two weeks I had to go in and refill our module tanks, which are these tanks here, with a special mixture of gas that would operate at an atmosphere. I mentioned earlier that maybe one of our problems might be a tress problem supporting that much mylar like this over that much area. We conquered that problem by operating the counter at 2 psi when it's in orbit. In the laboratory, it operated at 17 psi.
DeVorkin: That was inside the vacuum chamber?
Byram: No, like it is here. We had to keep the positive pressure on those windows at all times for fear of what might happen if the window were relaxed. You might get a crease in there, a fold in it, and break one of these windows.
DeVorkin: These are the detectors, or at least the prototype of the detector that you had tested in the vacuum chamber that we saw in a previous session.
DeVorkin: Now, this reputation that you had gained for getting into things that people didn't wish you to get into, was that the way that you found it most effective to operate? Not getting in their way, but what were the rules? Why did you feel as if these rules were too restrictive on you?
Byram: Well, I knew what I had to do to keep my experiment working. I knew better than they, but they liked to impose their rules, which conflicted with my belief of what was best for the experiment. But one of the things that I did that annoyed them was that I complained about an instrument they used in this room. They felt they had to avoid static shocks, so they had a huge ion generator to ionize the air, and circulated this ionized air in the room. But what it really was circulating was ozone, and it was ozone that was powerful enough to poison people, I'm sure. So I made them bring over their safety officer, and it didn't take him long to agree with me. Another thing that might have happened from the ozone is that ozone attacks rubber, and I'm sure that there's an awful lot of rubber exposed in that payload.
DeVorkin: So this was the typical sort of place where you ran down a different path than the TRW engineers.
DeVorkin: What about NASA inspectors? What was your relationship to them?
Byram: Well, there was only one NASA inspector, and I got along all right with him, but he agreed with TRW most of the time. That was Joe Jones. I don't think his picture is here.
DeVorkin: A lot of these components were tested for, as you said, breakage, especially the gas chamber that we can see in the right-handmost photograph there. It's that silver sort of sausage-shaped object. Was that the source of the gas for the gas flow during the life of HEAO?
Byram: Yes. There were really two supplies of gas. There was a central supply, and then each module had its own individual supply.
DeVorkin: We have one here that seems to have gone through some rough times.
Byram: The requirement of NASA was that this be tested at four times its rated pressure. We were going to operate at 500 psi, so the test had to go to 2000. It went to 2000 successfully, but they didn't stop the test; they ran it on up and blew it up.
DeVorkin: Did they want to test it to destruction?
Byram: They wanted to, but if I had been there, they wouldn't have done it.
Kreplin:That illustrates the difference between our tradition growing up in the space business, running our own shows, and then getting involved in the very large programs that NASA was running. The management levels get piled one on top of another, so that now working with the shuttle, we find that in a case like this, it's necessary to build four of these to get one, because the first has to be tested and burst; the second one has to be tested, cycled, and then burst; and the third one has to be cycled. Again, this is two levels greater than flight. And the fourth one you can use.
DeVorkin: Assuming the fourth one has the characteristics of the first three.
Kreplin:Well, that's assumed, I guess.
DeVorkin: How comfortable do you feel in those types of assumptions, especially when you're dealing with contractors?
Byram: Well, I don't like the idea personally.
DeVorkin: You do prefer to fly one that you test?
DeVorkin: Did bursting this particular chamber cause any problems or delay in providing a full detector?
Byram: Well, yes. No, it delayed building our first prototype detector, but there were other delays. It also involved $5,000 that we could have better used for something else.
DeVorkin: So that chamber cost $5,000?
DeVorkin: Was that something that was built on contract?
Byram: Yes, it was built on contract by — I think the name is on here somewhere. I think it's — I don't see it now.
DeVorkin: I'm curious. It's a relatively simple-looking chamber. Wouldn't that be something that normally you'd be able to build at NRL?
Byram: No. This is very thin, and there was a cryogenic treatment of some sort during the annealing that gave it very, very high tensile strength, much higher than normal. It was well over 100,000 pounds per square inch.
DeVorkin: So it's far more sophisticated than this spherical —
Byram: Oh, yes. That was an off-the-shelf item that was used for probably floating buoys out in the [Chesapeake] Bay.
DeVorkin: [Laughter] So that's more typical of the instrumentation at NRL that you're familiar with.
DeVorkin: Something you could build. Almost like a ball cock. Well, let's now change the scene and go to the detector that we have here, the HEAO detector, and identify the different pieces of the detector, both on the sky end and on the internal end. So we'll stop tape right now, I assume, and then reset up just a minute after we walk over there.
DeVorkin: We're now looking at the HEAO detector, one of the seven detectors. I would like to have you, Mr. Byram, point out where the collimators are and give us some idea of how the whole instrument is put together.
Byram: Well, the collimator is this grid that you can see. There's not really one grid; there's a couple of dozen grids. All of them are identical, and they're stacked one behind the other. Here is an alignment notch, here is another alignment notch, and there's another alignment notch right here.
DeVorkin: Should we turn it around right now?
Byram: Yes, I think you could turn it now. There's one other thing. Buried inside of the support grid here are heater wires, and there's one piece of the experiment that's missing, and that's a heat shield that went out in front of everything. That was just, essentially, an aluminized piece of mylar to reflect the sunlight.
DeVorkin: Okay. Let's turn this around now so we can look at the layers.
Byram: This is really two collimators assembled back to back. The spacers proceed from here, from the center out and from the center in, in a geometrical progression with the ratio of the spacer being roughly 20 percent of the space from here to here, which is roughly the transmission of an individual grid. Behind the collimator, there is another — the equivalent of a honeycomb to support the mylar, but it's not a hex-shaped honeycomb. The cells are rectangular because the rectangular shape could be assembled with long sheets of foil that would be unbent and would provide the strength to support the window. It's a much stronger arrangement than the honeycomb. The honeycomb is inherently not a stiff item.
DeVorkin: What came next?
Byram: Then behind that came — the counter has three layers, and for this counter, because it's operating at a lower pressure, the cells are larger than our rocket experiments. The cells are 2-inch by 2-inch. The spacing, then, between layers is 2 inches.
DeVorkin: These are the layers here?
DeVorkin: There's one behind this bar.
Byram: Yes, one behind that. The outer wire at each end on the top layer is used in anti-coincidence, and there's also a wire running crosswise that also provides anti-coincidence in the front layer. The back layer reinforces the data collection of the front layer and also provides some anti-coincidence counting. And the third
layer is strictly for anti-coincidence.
DeVorkin: What is this material? It's very soapy.
Byram: That's RTV. RTV-11 is the name of that. It's rubber, synthetic rubber. It was necessary to pot it, because we had gas leaks. We didn't really want to put that potting in, but it was necessary because our wire seals that covered the ends of the wire were not leak-tight at low temperatures. We found that out during a test at the last minute here at NRL. Caused much grief.
DeVorkin: Let's now move to the back where all the amplifiers and the sources are. Tell us what is here and what is missing. We can start just from this end and come across. There are a lot of things that look familiar, like the —
Byram: Standard old regulator, but with a different bellows that would operate at 2 psi. These are pressure transducers that tell us what the pressures are in the tank and in the counter. These are solenoid valves that control the flow of the gas. These tanks had to be refilled from a central gas tank periodically, like once a week.
Byram: This tank could hold about a week's supply of gas and would then periodically be refilled from the central tanks.
DeVorkin: There were central tanks on HEAO?
Byram: Yes. And the gas mixture was 90 percent xenon and the remainder was carbon dioxide.
DeVorkin: What were some of these other boxes? I see they're labeled "command decoder" and "module junction box." These are all electronics here?
Byram: A lot of it was just housekeeping functions to monitor voltages, to supply high voltage. These were high-voltage supplies. There was a telemetry system, an encoder system that was on here. One of these boxes was the housekeeping output. This was just a junction box that connected all the other boxes together.
DeVorkin: What was this rather large — is this a gas-filled spigot here?
Byram: We were going to put the relief valve there, but we moved the relief valve over here. So this served no purpose in the end. This is a control counter that was used to regulate the high voltage.
DeVorkin: What are these large dark gold patches?
Byram: Those are heaters.
DeVorkin: And what were they meant to heat?
Byram: The counter, keep the counter warm.
DeVorkin: The counter itself.
Byram: We had to supply heat that would radiate out through the front end.
DeVorkin: Is there a particular reason why all of this electronics was all modular and associated with the HEAO detector itself, rather than sitting on some central bus and removed from this?
Byram: Well, a lot of it had to be on the counter itself. We didn't want to run high voltage wires all over; we didn't want to run signal wires. We wanted to be able to test the counters individually without outside support. This was completely independent and could be operated without having any of the other parts. There was a central electronics package, but this could operate independently of that.
DeVorkin: Was this a design that you wanted, that your group and Dr. Friedman wanted? Did NASA resist this idea?
Byram: No, they didn't mind. They let us do it any way we wanted, so long as we kept within their weight limits and within their power limits.
DeVorkin: Were there any problems with power and weight?
Byram: Both. [DeVorkin chuckles]
DeVorkin: You had trouble keeping within them.
Byram: Yes. And we had, in the end, to get permission to go over that limit, partly because of the potting material, partly because of inaccurate estimates on the weights of some components.
DeVorkin: Did you use RTV throughout?
DeVorkin: What is missing from here?
Byram: I don't know the items that are missing.
DeVorkin: Okay. Is there anything else?
Byram: There's some high voltage missing, I know, but I'm not sure what these other boxes are. These two look identical, whatever they were.
DeVorkin: Okay. It says "proto" here in various places. Is this the prototype from HEAO?
Byram: Yes, this is a prototype.
DeVorkin: So this is the one you…
Byram: This is the first one we put together.
DeVorkin: Is there anything else we need to know about this while we have it in front of us?
Byram: Well, there was one unique kind of construction.
DeVorkin: Where can we see that?
DeVorkin: Okay. Let's turn it back. This way?
Byram: Yes, keep going, because this front here shows one of the unique things that was done. This was a development that was done by Bendix in Ann Arbor. They had the contract to build the original HEAO that started in 1971 or so. They said the way to go about it was to have all these frames extruded, so this is an extrusion, and then it was mitered and welded together as the front support. These pieces here are extrusions also. So these things came in 20-foot lengths, and we'd cut them up and assemble it into a box.
DeVorkin: I see.
Byram: And it worked very well.
DeVorkin: This is aluminum.
DeVorkin: And what did the extrusion allow you? That it came in continuous sheets, that was the advantage?
Byram: It eliminated all the machine work that would have had to been done. All these welds here where the wiring goes would have had to been hogged out on a milling machine.
DeVorkin: And that was a tremendous reduction in labor?
DeVorkin: That's very good. Anything else about it, or have we pretty much covered it?
Byram: That's about all.
DeVorkin: Okay. Thank you very much for showing us HEAO. We'll break then.
DeVorkin: We're here with Mr. E.T. Byram, who will lead us through the way that one can determine the aspect of a rocket, especially the aspect of the rocket for the critical observations of Sco X-1. Mr. Byram.
Byram: Well, to start off with, I'd better describe the motion of the rocket that we're talking about. The rocket is launched vertically. While it's going up vertically, it's drifting down-range. That puts torque on the fins, which cause it to not only spin, but to precess just like a top. It scans around like that. If you think of this as being our detector, that's the way it will scan across the sky.
To get a start, you know what stars are at the zenith, so you can set up a star chart or a star map such that the zenith at White Sands is vertical, so you'll see all the stars that are above this horizontal plane. So you've got a good idea of what to look for. You know that there are only a few extremely bright stars, so you can start trying to identify them.
DeVorkin: You might tell us just a little bit, concentrate on the ball for a minute or two and tell us what it is and how you made it.
Byram: Well, the ball was made by a bowling-ball company, Brunswick. They make almost perfectly round balls. This happens to be made of nylon rather than their regular black material.
DeVorkin: Did you ask for that to be made specially?
Byram: Yes. It didn't cost a whole lot, about the same price as a regular bowling ball. Twenty years ago or thirty years ago, that wasn't very much.
DeVorkin: Now this represents the sky.
Byram: This represents the celestial sphere. Here is the North Pole. This arc along here is the equator. This marks the ecliptic, which is the plane that contains the sun, the moon, and all of the planets and the earth. This represents the zenith at White Sands.
DeVorkin: And that ball is free to move in the socket?
Byram: The ball can be moved around, and you can draw circles on it and actually chart the motion of the rocket as you figure it out. For instance, this could mark the center of the yaw cone, so that the nose of the rocket follows a path like that. Now, at any point on that path, our detector is looking out perpendicular to that, which describes the circle over here. So you look along over here. You know that initially the rocket is going to be nearly vertical, so if you have a bright star such as Beta Centaurus in the field of view near the early part of the flight, you can usually determine at least the position of the nose of the rocket. If you know it's looking at a star over here, then you know the nose of the rocket is on a great circle scanning across this way. So you let a little time go by and find another star that you can identify, and then you can get a second great circle that will intersect the first great circle, and that will give you a close approximation to where the nose is.
DeVorkin: How do you determine which of the bright stars are the ones that you're seeing? How do you determine their identifications?
Byram: Well, you've got to find two stars that are close enough so that you can measure their angular separation.
DeVorkin: I just wanted to know — we're going to move to the chart now?
Byram: There are two stars here, for instance, Alpha Sco, that's Antares, that's a very bright star. Beta Centaurus is one of the brightest stars in the sky. Now, this chart is moving—it takes 15 seconds to make a roll on this particular rocket. The other thing that helps is that this has an X-ray source in it, so that we know that we're near Sco XR-1, and that means that we're near the constellation of Sco, and the brightest star in it is Alpha. So that's a pretty easy identification.
Then we can take this scale. We know that it's a 15-second period, so we know that it's 24 degrees per second. Each one of these bars is at one-second of time. At this time we're 200 seconds from launch, which is about 180 seconds from burnout. Well, we can't span 24 degrees, but we can make it 48. So each division on here will be a half a degree of scan.
DeVorkin: This looks like a very special kind of a ruler. What is that spring there?
Byram: It's just a calibrated spring. It's adjustable. You can make it any length you want. You can adjust it so that 48 divisions is one second, so each division is half a degree. You can measure the distance between this thing that we identify as Alpha Sco, and it comes down here to Beta Centaurus, and it's 100 divisions, which is just 50 degrees. So if you have a star chart handy, you can check up on that, which I did a couple of days ago, and it turns out to be correct that Sco XR-1 is indeed 50 degrees from Beta Centaurus. Beta Centaurus is way down on the horizon, and Alpha Sco is over here somewhere. But the problem then is to extrapolate those solutions to the whole flight. So you have to work yourself along from one section of the chart to the next and slowly identify stars that don't have too great a separation.
We don't have enough room here to really show another roll period, but we can go over here to this diagram to see what the trigonometric problem is. If this is the center of the yaw cone, the nose is passing along this path. Then our detector is looking out 90 degrees from this point, and we'll scan out a great circle along this path. All of these triangles are moving with time and will pass through zero and continue; they'll go through more than 360 degrees of angle. So you have to know how to assign algebraic sines to these angles. There's a rule of thumb that if you assign only 0 to 180 for all angles, no angle can be greater than 180, but it can be plus or minus, then you can make a rule that says all the angles within a triangle will have the same algebraic sine. Now, to make sure that they're 0 to 180, all you have to do is say that the algebraic sine of an angle is equal to the algebraic sine of the sine function of the angle. From that, you can work your way around. You have to solve this triangle right in here as a function of time.
DeVorkin: That's between the North Pole, the celestial pole…
Byram: The North Pole, the nose, and the star. With this angle varying through at least 360 degrees. So as this sweeps around here, this angle sweeps around. There's one difficulty that you run into in spherical triangles, and that is that it isn't like with plain triangles. The angles can add up to much more than 180 degrees. But if you use this little rule of thumb that the angles are only between 0 and 180, they can be plus or minus depending on the algebraic sine of the primary angle that you're interested in.
We flew magnetometers, which helped. The magnetometers then are referenced not to the North Pole, but to the magnetic pole. Now, the magnetic pole is much closer to the zenith. If the zenith were, say, here, the magnetic pole is somewhat to the east of that, and it's only 15 degrees from the zenith. The magnetic field is almost vertical, is what I'm trying to say.
DeVorkin: From White Sands.
Byram: Yes, White Sands or here, either one. But that helps you, in a way, because as the path of the rocket passes near the direction of the magnetic field, the period of the roll appears to change. When you're very close to the magnetic field, the period is very short. When the nose of the rocket is far from the magnetic field, the period appears to be much longer.
DeVorkin: That's the pole of the magnetic field?
Byram: Yes. So by finding out where these minimum and maximum times are, you'll find out roughly where you are with respect to the magnetic field, and it will also tell you what the period of the yaw period is. There is also an algebraic relationship that will tell you what — if you know the ratio of the shortest period to the
maximum period, you can determine what the actual size of this yaw cone is. That pretty much solves the whole problem for you. It's just a matter of picking your way along this path, identifying stars as you scan across them, and interpolating slowly. Then you plug all these parameters that you have determined into a computer program and let it predict what you will be doing at a considerably later time, like one quarter of the yaw cone away, one quarter of the yaw period away. You find out how much you're missing stars that are likely to be in the field, and from that you can adjust your input parameters and go back and go through the whole thing again. By cut and try, eventually you come up with the right solution.
DeVorkin: How accurate is it? Because I know that one of the biggest problems in the early sixties was really nailing down the positions of some of these X-ray sources. Was that limited by your ability to determine aspect or the resolution of the detectors?
Byram: Mostly it was the aspect solution, because we did not know where the X-ray sources were. We had nothing to guide us. In the early days, there was no association between visible sources and X-ray sources.
DeVorkin: Right. So you'd say your accuracy was limited by aspect.
DeVorkin: I see. Okay. Now, you developed these — should we possibly go back to the map? I'd like to see the magnetometer record.
Byram: Well, there is no — this record doesn't have that data on it.
DeVorkin: I see. Okay. This itself, could you explain what these records are of Sco X-1 itself, if we go back here?
Byram: Well, there were two detectors.
Byram: There are two detectors. This is a single detector, and this is the sum of the two detectors.
DeVorkin: You're referring, then, to this —
Yes, the lower peak is a single detector. This is a two-detector sum. This channel here is a scaled version of the sum of two — this ragged-looking signal here is really a series of pulses, and here you see these pulses reduced by a factor of 32. That is, their counting rate is 1/32th of what this counting rate is. So you can see the individual counts as distinct peaks. Down here they just look all jumbled together.
These signals are the output of something called a rate meter. That is, the amplitude of the signal is proportional to the counting rate in the detector.
DeVorkin: So to get a measure of the intensity, then, of Sco-X-1, what would be the measurement you'd make — just the distance from the continuum up to the peak?
Byram: Yes. Now, this one is saturated, so if we wanted to know what this sum is, we would go to this channel here.
DeVorkin: But it seems to me very hard to measure that. Do you measure it from this data, or is there another form of the data that you use? Or is this the primary record?
Byram: This is the primary record. We would have to accept that that's saturated and not — well, there is one thing you can do. From this record you can determine what the half-width of the response is, and then compare that to here. It looks like it's saturated for just the half-period, and so the true amplitude here is probably just twice what this is.
DeVorkin: What is the significance of the shape of the curve here?
Byram: That's just the angular response of our honeycomb.
DeVorkin: Okay. Now, in order to determine the exact position of that source, what would you do? Draw lines through the center up to where you have the stars identified?
Byram: Well, we would use our computer output to tell us where we were looking at the time at which this peak has reached.
DeVorkin: And you're saying that the resolution of that, the spatial resolution, was poorer than the spatial resolution of the detector itself. So you could not get to within plus or minus 2 or 3 degrees using aspect. Is that correct?
Byram: Oh, no, you can do better than that. Our original location of Sco XR-1, for instance, held for ten or
fifteen years before somebody got a better location on it. That was certainly within a quarter of a degree.
DeVorkin: How did you get that kind of accuracy out of this chart?
Byram: By cut and try, going back and forth from sources here to stars here. It's all related by time. There was one other little effect that we had to include, and that is the sloshing fuel remaining in the tanks gradually slowed the spin rate. It would decelerate it, so we added a little factor in our computer program to allow for that. The computer we used in the early days was a Bendix G-15, which, compared to a modern IBM-AT, the Bendix machine was about a tenth as good, maybe worse than that.
DeVorkin: What was it like to use?
Byram: It was slow and clumsy and not too easy to program and it used a language nobody ever uses now called INTERCOM. INTERCOM was a pseudo-machine language. That is, like "multiply" would be "44," followed by an address. If you wanted to multiply one number by another, you'd go to the accumulator with the first number and then say "multiply by the number at the following location." And there would be a number for "multiply," another for "divide," and formatting was clumsy, too.
DeVorkin: This tracing must have been made around 1963 or some of your earlier tracings, '63 or '64.
Byram: It was '64 or '65.
DeVorkin: So is that when you were using the G-15?
No, we didn't use the G-15 for too long. In 1960, I think we got a new machine. Yes, in 1960 we got a Control Data machine. This really wasn't — it was really those early ultraviolet experiments that used the G-15. Actually, what happened was that when we bought the CDC machine, we insisted that they provide us with INTERCOM, because we already knew that programming language.
DeVorkin: How did you learn INTERCOM then?
Byram: Oh, it was easy. It was an easy language to learn. There were only a few operations. It wasn't as flexible as FORTRAN by any means, which made it harder to get the answers but easier to program.
DeVorkin: Did you go through this whole derivation of the aspect of the rocket course manually before you had computers?
Byram: We got our first computer about the same time that we had a rocket to analyze. Our very first analysis was done before NASA was formed, so the first rocket was sometime in 1957 — Kupperian's rocket, Ultraviolet 25, I think.
DeVorkin: Wasn't your first ultraviolet back in '55? Your first ultraviolet stellar work was '55, '56? '54?
Byram: I don't think so. I think it was '57.
DeVorkin: Okay. And when you first flew the ultraviolet rocket experiment, at least, the stellar experiment, that's when you first used the computer.
DeVorkin: Would you have wanted to do it without the computer?
Byram: I didn't do the first one. Kupperian and [Richard] Milligan and [Albert] Boggess [III] did those of that first one. I learned to do it by taking that same experiment and redoing it, so in the year after, in 1959, the year that Goddard was formed, I spent about the first half of that year just learning how to do the aspect.
DeVorkin: That's when you knew the others were leaving.
Byram: Yes. They had already left, and I wasn't sure I trusted their solution. Besides that, I wanted to learn to do it. My solution agreed with theirs very closely.
DeVorkin: Marvelous. Is there anything we haven't covered, then, on aspect, or that you would like to cover?
Byram: No, I think we've covered it pretty well.
DeVorkin: Okay. I thank you very, very much. This has been a very fine session.