Sander Weinreb, Interviewed by Kenneth I. Kellermann on 17 September 2023

Kellermann: 00:01

This is Ken Kellermann, and I'm with Sandy Weinreb.  It's September 17^th [2023]. And Sandy was starting to talk about fixing the receiver that you installed at Harvard, or the back end.

Weinreb: 00:27

Well, you want to start with Ewen? How did I get into radio astronomy? Well--

Kellermann: 00:36

This is before you built your correlator? -- No, this is after you finished at MIT.

Weinreb: 00:45

No. Oh, okay. No. I was a junior at MIT, and I decided I would-- or maybe I was a sophomore. I would get a summer job, a technical job, outside of MIT. I worked at MIT, too. I did some technician work. But I went through the Boston telephone directory, and I maybe called 100 places, but there were four under physicist. And one of them was a company called Ewen-Knight. And I got an interview. And there was a guy named Pete Strum who was the vice president who interviewed me. And he was a very good technical person and good interview for me. And at the end, we walked through the lab, and there was this guy sitting on top of a pedestal on a big workbench. And around him were several assemblers. And he was sitting up there like a Buddha. And he didn't get down. Pete Srum said, "Here's Sandy Weinreb from MIT undergraduate school. He was looking for a job." And Ewen said something like, "Okay. Give him a buck. Let's hire him." That was my introduction.

Kellermann: 02:30

Was this a summer job, or?

Weinreb: 02:31

This was to be a summer job. And this had been after Ewen had discovered the hydrogen line. And Harvard wanted to continue doing hydrogen line. I don't remember whether that was Tom Gold or Bok wanted to do it. And so they had bought a receiver from Ewen Knight for $50,000, and it didn't work.

Kellermann: 03:06

By receiver, you mean front end or back end, or both?

Weinreb: 03:09

I think it was both. So Ewen arranged it so that Harvard would pay me to make the receiver work. Well, here I was an undergraduate. I knew a lot about electronics, I built an oscilloscope when I was in high school, designed an oscilloscope. But radio astronomy, I knew nothing. And the receiver was a can of worms. But I took the job and eventually made it a one-channel receiver, which did work. And it was on the 60-foot telescope out at Harvard, Massachusetts.

Kellermann: 04:08

What year would this have been?

Weinreb: 04:10

Well, let's see. I was class of '58. So this must have been '55 or '56.

Kellermann: 04:19

Were you by any chance at the dedication of the 60-foot telescope?

Weinreb: 04:26

I probably came after that. I don't know.

Kellermann: 04:29

It would have been the spring of '56.

Weinreb: 04:32

Okay. I may have been there.

Kellermann: 04:34

Well, I was too. Because I was taking a freshman astronomy course at MIT. It was the only astronomy course I ever took.

Weinreb: 04:44

Oh, I never took one.

Kellermann: 04:45

And they took us out there for the dedication. It was on a Saturday.

Weinreb: 04:50

Who was he?

Kellermann: 04:52

Well, teacher. He wasn't an astronomer. He was a geophysicist, but.

Weinreb: 04:55

Well, one of the other people working for Ewen- Knight was Frank Drake. And I think he was a graduate student at the time. I was an undergraduate. And Kochu Menon was around-- I don't know what position he had. But I got another undergraduate student named Ron Weimer, who eventually worked for NRAO and was in Green Bank for years and years.

Kellermann: 05:29

I didn't know that.

Weinreb: 05:31

So he got started in radio astronomy around the same time I did. And I worked on the receiver, and then Ewen said, "Why don't you do a thesis, a bachelor's thesis, about deuterium?" He says, "I detected hydrogen, but deuterium, we want to know the deuterium to hydrogen ratio." And an important measurement, and the Soviets have claimed they've done it, and there's a few other people, Davies, maybe, in England, very controversial. It was very hard to do it. And so I took it on as a bachelor's thesis. I didn't do any experimental work. I just wrote up what you had to do. And the next summer, I didn't work in radio astronomy. I went down to Melbourne, Florida, and worked for radiation military receiver work. I had grown up in Florida, and I wanted my new wife to like living in Florida, and I wanted to eventually live in Florida. So we spent the summer there.

Weinreb: 07:01

But prior to that, let me say one other little story about Doc Ewen. I was working for him during the summer, and I was married on July 4th, and he had me over to his house about a week before. And he said, "Where are you going for honeymoon?" And I said, "Oh, I'm going down to the Cape on July 4th." He said, "Oh, where are you staying?" I said, "Oh, I don't know. I'll find someplace." And he said, "Oh, no. You've got to have a reservation." And, "No. It's too late for you to get a reservation. Here, I want you to take my reservation." And he had reserved a beautiful room, Blue Waters Motel, and he turned the reservation over to me. Thanks, Doc. That was great of you. Margie was at this dinner at his house. And one of the things I remember him saying, I mean, he was very much into cars, Lincolns, one after another. But he said-- I think I have it right-- that he was boxing champion of the US Navy. I've never checked that. But he did have a nose that looked like he'd been punched a few times. And he was a character. I kept track of Doc. I saw him often for the next 40 years. That is, he became a vice president of Militech when I was teaching out at UMass. And I used to go over to Militech all the time and see him. I went to his house many times. Great friend, Doc Ewan.

Kellermann: 09:06

Did you meet his son?

Weinreb: 09:08

No, I met his daughter who was-- as he got older, she took care of him.

Kellermann: 09:15

Well after he died, which is about four or five years ago, his son contacted Jim Moran and myself about arranging a-- let's see. Would it have been the 50th anniversary? No. Some anniversary of the hydrogen line discovery.

Weinreb: 09:42

1951 it was discovered.

Kellermann: 09:43

Maybe it was the 70th or something like that. And he took after his father I guess. He was a wheeler-dealer and organizer. And he was going to arrange a big conference and everything. Jim and I were helping. But then COVID came, and the thing sort of died. David I think was his name.

Weinreb: 10:16

I didn’t know him. Don't remember him. All right. I would have liked to come probably to the conference if I was invited.

Kellermann: 10:26

It didn't happen.

Weinreb: 10:27

It didn't happen, okay?

Kellermann: 10:29

But there's a lot. He donated all of his papers and everything to NRAO. It's in the Archives. And so there's quite a bit about him in the book.

Weinreb: 10:43

All right. Well, no, I saw a picture of him.

Kellermann: 10:48

Yeah. So apparently, he became aware of the van der Hulst paper through his Navy connections because it was translated by the military. And van der Hulst made some errors in his calculation. So he was very skeptical about where the line could be detected. And Ewen seemed to be just interested getting his thesis done.

Weinreb: 11:25

Was it the frequency that was missing, or?

Kellermann: 11:28

No, no, the transition probably.

Weinreb: 11:32

Coefficients, yeah.

Kellermann: 11:34

And so he was trying to--

Weinreb: 11:36

Purcell believed it could be detected.

Kellermann: 11:39

He didn't know. He just knew about-- he just knew about the line and that it had been measured in the lab.

Weinreb: 11:49

Oh, Purcell measured the frequency in the lab. No.

Kellermann: 11:52

No. Somebody else did, but Purcell knew about it. That's all he knew. He knew nothing of astrophysics and the probable density. But it wasn't until Ewen found out.

Weinreb: 12:10

Ewen wrote a $500 proposal.

Kellermann: 12:12

Purcell did.

Weinreb: 12:14

Oh, Purcell did?

Kellermann: 12:15

Yeah.

Weinreb: 12:16

Okay, to fund it. And you have a copy of that? It must be in the files. What--?

Kellermann: 12:22

Yes.

Weinreb: 12:23

--details of the $500?

Kellermann: 12:24

Yes, It's just a letter. You know--

Weinreb: 12:30

Crystals.

Kellermann: 12:31

No. an electronics wizard a Harvard, physicist, just retired, wrote this fantastic book years ago-- The Art of Electronics.

Weinreb: 12:40

Oh. Horowitz?

Kellermann: 12:41

Yeah.

Weinreb: 12:41

Sure. Paul Horwitz.

Kellermann: 12:43

Yeah. And I can send it to you. He wrote a little thing about the line discovery in the Harvard newsletter-- some newsletter. And he discusses the one page proposal.

Weinreb: 13:01

Economics, maybe. [laughter]

Kellermann: 13:02

Yeah. I think I have a copy of that letter.

Weinreb: 13:06

Yeah. No, that was something. And Ewen didn't move on in astronomy. Business was his thing.

Kellermann: 13:23

So how did you go from hydrogen line receiver? And then, you came down to Green Bank to observe deuterium line.

Weinreb: 13:33

Okay. Now, pick up the story. Okay. The next summer, I didn't work for Ewen. I worked on a communication receiver for the military. But then, I became a graduate student and started looking for a thesis-- a doctoral thesis. And the only person with connections with radio astronomy at MIT was Wiesner-- Jerry Wiesner, who was a professor in electrical engineering. And he had had a student prior to me who did something in radio astronomy. And I told him about the deuterium line and gave him my bachelor's thesis as an introduction. And he said, "Great."

Kellermann: 14:31

So you were specifically looking for something in radio astronomy?

Weinreb: 14:36

Me?

Kellermann: 14:38

You.

Weinreb: 14:39

I was looking for the deuterium line.

Kellermann: 14:41

Yeah. No, but I mean--

Weinreb: 14:42

Oh. And I was looking for a thesis in the astronomy.

[crosstalk].

Weinreb: 14:46

And the only professor-- and if I want to stay at MIT, it was Wiesner. So he took it on. And I don't remember how much of it I wrote, how much of it he wrote-- a proposal to NSF. I think it was $60,000 to build the receiver, including a digital correlator.

Kellermann: 15:19

And whose initiative was that?

Weinreb: 15:21

Well, that was my initiative. By that time, I had taken digital circuits and I knew how to build it. And to build it-- for $60,000, I think that I only got about $30,000 [laughter] for equipment. So, boy, that was hard to do in those days. We were talking about 1960, 1959. People didn't do digital processing even at hundreds of kilohertz rates. That is, I was trying to analyze 150 kilohertz bandwidth needed for the deuterium line. Needed a clock rate of 300 kilohertz. And so there was another guy who worked at Ewen Knigh named Harry Adams who had a little company, Control Equipment Corporation. And they wouldn't build the correlator for $30,000, but they would sell me boards. And a big part of the correlator was counters. The correlation function was accumulated in counters, that is it had-- you have to multiply to input, but then you have to accumulate. You have the sum, and we wanted the sum for months even. So my job, the cards that I got from Control Equipment would do the multiplication for the correlation. But the accumulator, I did with a combination of flip-flops that I designed from transistors, two N398 transistors, driving a neon bulb. And the readout was the telescope operator reading this row of neon bulbs, and he'd better get it right.

Weinreb: 17:49

But in addition to the neon bulbs, there was then an electromechanical counter, click, click, click counter. I had trouble with those. Some of those failed to produce an error in the correlation function; very bad. But anyway, we built these three racks of equipment, two of the racks being counters. And at that point, the test of it-- the big thing was accuracy. This correlator had to be-- had to see very small changes in the correlation function. And so I did that in Building 20 of MIT, the Rad Lab building. I had a big lab. I had a technician who worked for me. It was great.

Kellermann: 18:59

That would have been the early '60s.

Weinreb: 19:00

Yeah, I was probably finishing the correlator around 1960. I have to think about when did I go to Green Bank. Now, before that, an unfortunate, for me, event occurred, Kennedy appointed, or Wiesner got appointed by John F. Kennedy to be his advisor. Oh, darn. I had him first. So I went to see Wiesner. "Congratulations on your appointment. What do we do now?" He said, "Well, I'm going down to Washington. You're going to Green Bank. Why don't you stop by on the way back and tell me what happened?" "Sure." I did that, actually. I went to the Executive Office building, all kinds of security. I got to see Wiesner for 10 minutes. He said, "I've got to go see Kennedy now." He rushed off. But I did the deuterium line search. Six months of observing did not detect it, had a negative result, which, the Soviets had a positive result. They were wrong. But I came back to MIT, had a thesis exam, which I failed. Here I was looking for this line that the Russians had found, and I was looking for it with a new piece of instrumentation, and I didn't find it. "Oh, gee, you better go back and check your electronics." But I had built an artificial deuterium line and put it at about the level that I should have gotten deuterium. And it worked. And I was sure the correlator worked, but Wiesner wasn't there. And I realized you've got to have a thesis supervisor to help you in your thesis exam, particularly at MIT. I see other thesis exams since then. This was a tough group. There was a guy named Bill Siebert, who was in--

Kellermann: 21:49

This is all electrical engineering?

Weinreb: 21:50

All electrical engineering. Siebert was more interested in the theory, and the one-bit correlation, oh, he wasn't sold on that, and asked a lot of questions, and so I failed. I forgot what I was supposed to do, go test it some more, which I did, but around that time, MIT hired Alan Barrett. Wow. He rescued me. He knew something about radio astronomy and about receivers, and he was appointed then on the committee. And I don't remember whether Wiesner was supposed to come or he didn't come the second time either, but I passed, I think, thanks to Barrett. And then they said, "Well, use the same equipment to look for the Zeeman effect in hydrogen."  I went back to Green Bank and didn't detect the Zeeman splitting of the hydrogen line either. But I then came back to MIT, got my degree in Lincoln Lab, a guy named Lit Meeks [M. Littleton Meeks]--

Kellermann: 23:13

Yeah, I knew him.  Lit Meeks.

Weinreb: 23:15

--and Alan Barrett and John Henry said, "Come out to Millstone, 85-foot dish," and Barrett knew the frequency of the OH line. He had been a student of Towne's and had key information, where to look for it. And I was so negative, discouraged that I had two negative experiments, and I was just shocked that in the first 10 minutes of observing, bang, there was a spectral line at exactly the frequency that Barrett had expected it at, and it was with zero velocity. We looked at the galactic center, which the velocity should have been close to zero. And sure, we saw a little wiggle there right away. It was pretty big. Several Kelvin, I think, of signal. And bang, there was a newspaper article. There was a letter to Nature, a big discovery. And we later found out that little wiggle was on the side of an enormous line that they saw at Berkeley or in the Netherlands. There were other people looking for the line. And at that point--

Kellermann: 24:52

It was a deep absorption line.

Weinreb: 24:53

It was a very deep absorption line with a wiggle on the side of that absorption line. And it's in your book, Ken. So at that point, NRAO said, "Gee, we should build a correlator." And you only had a 50-channel. “We want a 100-channel, one megahertz correlator,” and they hired Art Shalloway to--

Kellermann: 25:24

From where?

Weinreb: 25:26

Oh, Stromberg-Carlson. He worked for some commercial company, new digital hardware. And I remember him coming to me. I lived in Watertown. I guess I was still working for Lincoln Lab. And we sat at my dining room table and went over how to build the digital correlator. And he went back and, sure enough, built the 100-channel, which went on for many years, used on the 140-foot telescope, very successful. And around that time, I got to know Hein Hvatum, and he offered me a job to come down to NRAO, be head of the Electronics Division. And here I was 26 years old, 30 people in the Electronics Division. Hein was moving up to the Assistant Director, or Associate Director. And it was an administrative position, some technical, I know, but I accepted. Convinced Margie, my wife, to move to West Virginia. Wasn't easy. We had a two- or three-year-old son.

Kellermann: 27:02

Were you told at that time that the observatory was going to move to Charlottesville?

Weinreb: 27:06

Absolutely. That was the only way I could accept and get Margie to accept that we'd move to Charlottesville in a few months. That turned out to be three years that we were in Green Bank. We lived in the Rabbit Patch of houses, the ten houses. It was a great time. I met Ken Kellermann there. I met so many people. Rabi was on the Visiting Committee.

Kellermann: 27:36

No, the President of AUI.

Weinreb: 27:38

Okay. He was President of AUI before Jerry Tape, who became later President. And I met Reber. I think I met Jansky. I'm not sure about that.

Kellermann: 27:56

No. No.

Weinreb: 27:57

No. Reber?

Kellermann: 27:58

Yes. Reber

Weinreb: 27:59

Okay.

Kellermann: 28:00

Jansky died in '50 or '51.

Weinreb: 28:02

Okay. I didn't meet Jansky.

Kellermann: 28:04

He died before the hydrogen line was discovered.

Weinreb: 28:06

All right. But I had a good time talking to Grote Reber. And I think Bolton was at Caltech. I met him. I met Stanley, who was a receiver guy at Caltech.

Kellermann: 28:25

So why did they hire you? I would have thought-- why not Shalloway?

Weinreb: 28:31

Well, he was acting head. But Shalloway wasn't the microwave guy. I was the whole receiver.

Kellermann: 28:41

That was going to be another question. Backing off a minute. Had you not appreciated that the biggest contribution of your thesis was not the negative result on deuterium, but the one-bit correlator, which as far as I know, was an entirely new concept.

Weinreb: 29:01

I didn't appreciate it. MIT didn't appreciate it. Barrett, I think, appreciated it, and Hein Hvatum did. Shalloway did. People started appreciating it, but I didn't realize that it was--

Kellermann: 29:20

And why did you go through one bit? Was it just the limitation of the data rate?

Weinreb: 29:24

Sure.  Digital processing to build a multiplier with more than one bit was very expensive.

Kellermann: 29:35

And complicated.

Weinreb: 29:36

And complicated. I don't remember the 100 channel that Shalloway built, whether it had two bits--

Kellermann: 29:44

No, I think it was still one.

Weinreb: 29:45

One bit. People had to learn, how much do you gain? Now, what was in my thesis was what do you lose by going to one bit instead of an infinite number of bits or an analog correlator. And it was something like 1.39, a factor of around 1.39.

Kellermann: 30:16

The two bits only get back about half of that, or?

Weinreb: 30:18

Yeah.

Kellermann: 30:19

And it introduces a lot of other problems.

Weinreb: 30:21

Yes. But as digital processing became easier, it was easy to do two or three bits even.

Kellermann: 30:30

But not at the time you did your thesis.

Weinreb: 30:33

Right. Yeah. So I came to NRAO, and there were administrative problems.

Kellermann: 30:46

Let's stop for a moment. Your room is ready.

[brief interruption]

Weinreb: 30:49

Okay. All right. In Green Bank. And the most immediate problem was the Teamsters Union. This is West Virginia. NRAO employed 100 people. And the Teamsters wanted to unionize it. And the telescope operators were 10 or 15 of 100 people. And the electronics technicians in the electronics division were another 10 or so. And there had to be an election to determine whether NRAO would be - the technicians and the telescope operators, telescope mechanics, 30 or 40 people would become unionized. And there's very strict rules about how this has to be done. I couldn't say anything. Very restricted what I could say to the technicians. Of course, I told them technically what I wanted them to do. And I had had a lot of technician experience myself. There's a picture I have of when I was 12 years old working in a radio-TV shop in Miami, Florida, and there's a sign that says, "Technician on duty, Sandy." The guy who owned the shop put that sign up there. He didn't have that sign up when a customer came in because he didn't want a kid fixing their radio. But anyway, I knew how to talk to these techs. I'd worked in a trucking company with a lot of veterans. Most of these were veterans. And I could get along well with the technicians. Engineers were a little harder. I won‘t name any names but there was one guy who didn't know what he was doing. He was fooling everybody. And just the project go on and on time-wise. And I had to fire him. And that was hard.

Weinreb: 33:38

I figured I now had the experience of having to fire other people. Give them plenty of time. Tell them, "Go find a job." And give them the reasons why they weren't doing well in their present job. So the next thing that came along is, well, the Green Bank Interferometer, the VLA proposal, which if you look back at it, you'll see it was for continuum only, 3 centimeters and 11 centimeter wavelength. Not at all what was built. And I say that I take a lot of the credit for advancing it. Using a digital correlator that was a natural for me and Shalloway. And building receivers with cryogenics even. Wow. That was a big step.

Kellermann: 34:59

We couldn't even keep one working on the 140-foot.

Weinreb: 35:02

Well, I kept it working.  Only when you observed did it not work. But it was a risk. And there was another enormous risk and that was how to connect 21 kilometers of antennas with hundreds of megahertz of bandwidth. How to transmit that signal. We thought we would do it with coaxial cables, one and five-eighths inch diameter, coaxial cables, which not only were expensive, we needed 21 kilometers, three times that of it. But it had a lot of attenuation. It had to have amplifiers along the way. And around that time, good old Bell Labs was developing loss-less waveguide or TE 01 mode waveguide, a hollow pipe 60 millimeters in diameter, which the electromagnetic wave flows down the middle of the pipe and doesn't have losses due to the walls. The walls had to be a helical wire, which any mode that got converted into a mode that generated currents in the walls was attenuated. And so you had to launch this TE 01 mode, which was difficult. But the telephone company had worked several years on that. And there were four companies in Japan who worked on it for the Japanese telephone system. And so I visited Bell Labs, got a lot of help from Bell Labs, and convinced Hein Hvatum and Heeschen that this would work. And in fact, they took so many risks with me. I suggested the risky things. But they had the confidence in me and knew enough themselves to know that these were good bets.

Kellermann: 37:34

So it was your decision to go to the waveguide?

Weinreb: 37:36

It was ultimately, I'd say, Heeschen’s decision, but it was my recommendation.

Kellermann: 37:43

Your recommendation?

Weinreb: 37:44

Yes. So the next thing was, "Where do we get the waveguide?" Bell Labs had made one or two kilometers of it. Well, Japan was further along. The four companies, maybe it was three companies, Sumitomo, Fujikura, and Furukawa, were competing for the telephone system of Japan to put in waveguide. And all three bid on doing it for the VLA. And I think it was bids around $10 million dollars. It was a lot of money. And there wasn't any by Japan question -- there wasn't anybody in the US to do it. Bell Labs, at that point, was getting more interested in fiber. They still couldn't do the fiber for the VLA. But they didn't want to put more money into waveguide, so.

Kellermann: 39:00

I think they were also looking at satellites then.

Weinreb: 39:02

That could be. Yes. Yeah. So a lawyer who worked for NRAO named Jay Marymor - he was in charge of procurements, $10 million dollars - he had to be heavily involved. We went to Japan. Great trip. All three companies, we visited them. We were impressed by all three. And eventually, the Japanese government decided who would win this bid. And they said, Furukawa would do the waveguide and Fujikura would do the connections of the antenna. And that's what we contracted with. They sent some engineers to Socorro, I think, at that time to help with the waveguide system. With the couplers, John Archer came over from Australia and helped with the design. The directional couplers launched that lossless mode. And electrically went very well, met the specifications. We needed no repeaters, no amplifiers, that could do the whole 21 kilometers. But a problem did come up. It was on the corrosion of the pipe, steel pipe. And everybody told us, "You must have protection of the steel pipe." And sure enough, there's a company in the US called Pipe Protection Company that water pipes all over the United States, gas pipes are protected by Pipe Protection Company somewhere in California. And the advice from the experts was, "Don't let the Japanese do the protection. They don't have a good protection system.”

Weinreb: 41:36

And so we went to Pipe Protection. The waveguide was taken by ship to Pittsburg California, where Pipe Protection had a plant. And they put tubing, essentially, over the pipe. But the problem was the pipe only came in something like 20 meter lengths, and they had to have connections between the 20 meter lengths. And there was a bulge at that connection point, and it leaked water. And the pipes started corroding at the connections. And well, we had Jay Marymor, and Pipe Protection made it good. We had to dig it up, I don't know, a few kilometers, it had already been put in and send them to Pipe Protection. And they put a sealant in between the outer sleeving, and I think that worked for the next 20 years or so. Of course, then fiber came along and was put in the waveguide, inside. That was the end of the waveguide story. But fiber is now the key thing in large arrays for radio astronomy. And we sure need it in the DSA 2000 that Caltech is now assigning, 2 thousand antennas in a 20 kilometer diameter. I've forgotten how many hundreds of kilometers of fiber are needed. And we're designing that now. Should I go on here? Is this--

Kellermann: 43:47

I was going to add, even the VLBA is getting connected up now. That's through commercial fiber.

Weinreb: 43:55

Yes. In other words, you're using Internet. I thought they would charge you too much.

Kellermann: 44:05

The bandwidth is limited.

Weinreb: 44:08

Oh, but you send it over a period of time.

Kellermann: 44:11

No, you can't because we're observing all the time.

Weinreb: 44:13

Well, oh, I see. You want to observe all the time also.

Kellermann: 44:18

Yeah. So it's a hybrid system for--  Going back to Green Bank, I mean, like you say, after three years, you moved to Charlottesville and you at some point became Assistant Director. Were there issues with the dual management of the Green Bank engineers? They reported to you. They reported to the head of Green Bank Electronics who reported to the Green Bank Site Director.

Weinreb: 44:58

It wasn't an issue with me. I had my hands full with the VLA.

Kellermann: 45:05

So you let Green Bank run itself, essentially?

Weinreb: 45:09

Well, I hired Mike Balister to be the Associate Division Head and run Green Bank. And he was very competent. And I used to go to Green Bank once a month from Charlottesville, a long trip. We first tried to do it by air. We had a once-a-week flight. Too many times it was canceled in foggy weather. Hein was improving the telephone service.  He was really an instigator for Zoom and telephone meetings. He believed in that early on, greatly helped the Observatory. So, yeah, any other questions here, Ken?

Kellermann: 46:10

Well, on a personal side, you worked at research organizations like NRAO and JPL. You've worked at universities, Caltech, UMass, and you've worked in industry, Bendix and--

Weinreb: 46:30

No, no. Martin Marietta.

Kellermann: 46:31

Martin Marietta, Applied Physics Lab. How would you compare the pros and cons of--?

Weinreb: 46:40

Yeah. NRAO was the best because of leadership at the time. I can't speak for future leadership. They've had some good ones, but.

Kellermann: 46:57

I'll ask you why did you leave?

Weinreb: 47:00

Why did I leave?

Kellermann: 47:01

You were at the height of both your career personally and NRAO with the VLA. I remember you went to UVA, but that was just for a year or two. It's understandable. You wanted to teach and everything. And that was on leave, I believe, from NRAO. But then you left.

Weinreb: 47:24

Yeah. I got an early middle age crisis thinking, "Gee, is this my whole life? Am I going to end up staying at NRAO and retiring from NRAO, or is there more to life?" And let's see, did I-- I went to sabbatical. Yeah, the sabbatical leave to Berkeley was the start. And I also thought my expertise is not as an administrator. And I was becoming more and more an administrator at NRAO. The Electronics Division was getting bigger. And then as Assistant Director, I was in charge of the computer and mechanical engineering division also. And I did okay, but I thought, "Well, I'm really a researcher and an engineer of new things." So when I went to Berkeley, it was to build a cryogenic low-noise amplifier, which consisted of a parametric up-converter and a 22 gigahertz maser. It was a 5 gigahertz up-converter, 22 gigahertz. It would have very low noise. And I worked on that for two years. It was successful, but there were better ways to do it that didn't require the maser. The maser required 4 Kelvin cooling. So I then had to make the decision of going back to NRAO or teaching.

Weinreb: 49:27

And I did some teaching at UVA and then got the bug to work in an industry, defense industry. Martin Marietta, Baltimore. It was the research division, very well-funded, millimeter-wave radars. Why millimeter wave? Well, Martin Marietta, the big business they wanted was small missiles that could be fired out of a 100-millimeter cannon. Or it could be a missile out of a submarine, but small. And they were to stop the tank invasion from the Soviet Union. This was about 1990, 1989. Around in there. It was the Cold War, but they were a little behind the times in thinking it was going to be a tank invasion. And so what I worked on was a radar seeker that would fit within 100 millimeters and would come into the top of the tank. You could find a tank out in the field. And I worked on that for a few years. It was a radar-phased array.

Kellermann: 51:15

Oh!

Weinreb: 51:16

Something like 40 elements, 32 elements, within that 100-millimeter diameter. There were little dipoles and little MMIC LNAs. You see, it was the start of microwave integrated circuit low noise amplifiers. It wasn't anything cryogenic, of course, but it had to be tiny, and it had to have a transmitter in it too. I worked on that. And had the nice experience of being sent out to San Diego to test it from the 15^th floor of a hotel, a penthouse. I had a penthouse suite up there where-- out on the balcony. I could set up my radar. And there was a vacant lot next door with radar targets-- calibrated targets. I'd test this system, and it worked. But around that time-- this was all Martin Marietta. Around that time, Lockheed and Martin Marietta combined and became Lockheed Martin, and they decided they had seven research labs, and that was too many, and the one in Baltimore was to be shut down. That was terrible. That was a very productive research lab making the integrated circuits. They did solid-state processing, very well-run.

Weinreb: 53:05

And just everybody was—they said, "You've got six months. We're closing down. If you want any of the equipment, if you'll pay" - I think it was 20% - "of what it cost us, you can have it." I bought a few items. There was one very expensive thing that I had bought for $100,000, and I didn't want to pay $20,000. It was the test set to measure the noise properties of a transistor-- the noise parameters it was called. Very difficult to do and they had this test set and I didn't pay for it-- I didn't buy it. But I was one of the last people in the building before we moved out, and it was sitting on the shelf. And I was told, "Just leave it there." The junk people were coming in, and I left it there. And for years, I was looking for it to appear on some used equipment. Never did. I don't know what happened to it, but I didn't get it. And let's see. Where did I go from Martin Marietta?

Kellermann: 54:44

UMass?

Weinreb: 54:45

Yeah. Back to Massachusetts. Three years at UMass. Not too happy years. Bad weather, and one of the things that happened-- I got shingles when I was there. Of course, that wasn't all that bad, but I fractured ribs trudging through the ice and snow to get into the astronomy building, and there was a shorter way to walk from the parking lot, and they wouldn't let me go through that building, and I was mad about that. If I was in the physics department, I could go through the building, but I had a broken rib at that time, and if I fell again, I could die. And that was one of the reasons I left. But yeah, I didn't have that great students. We were working on the millimeter array in Mexico.

Kellermann: 56:05

Millimeter, antenna, not array.

Weinreb: 56:06

Antenna. Yeah. I was working on receivers for it. And UMass had a 13-meter, 12-meter telescope. I did some equipment for that, along with Neil Erickson, who’s still there. But then I got the bug to move to California. The reason is that Margie's parents died. They were in Massachusetts and she wanted to be near them. And so that went away. And I think I initially worked for Hughes or I had a job offer for Hughes in Torrance, California. And we bought a home in Rancho Palos Verdes. And then Margie got cold feet. We had to sell the house. We didn't move. We didn't have to move. But then, I don't know, a year later or so, I got the bug again. And I had a graduate student at Berkeley named Dave Rutledge who had become a professor at Caltech in the EE department. And he helped arrange for me to get an offer. I don't think it was full professor, but associate professor was pretty high-level professor. And we thought it would go through. I had some letters, and it required approval of a faculty committee that eventually disapproved - that one vote against it. It had to be unanimous.

Kellermann: 58:15

This is in electrical engineering?

Weinreb: 58:16

Yes.

Kellermann: 58:17

I mean, I wrote one of those letters. Well, you knew that. You asked me.

Weinreb: 58:20

All right. Yeah. No. Well, that was a big disappointment. They didn't hire me. And I had another house. That time, I had put a deposit on the house of $2,000 I think I lost. That house, by the way, was $250,000. It was a lot of money in those days. It's now worth about $2 million. If I would have bought it and kept a hold onto it, it would have been great. But anyway, I eventually decided to work for JPL. That was about 1999. And I had an adjunct position at Caltech, worked with Rutledge, had graduate students, had five PhDs, great students at Caltech, all very successful. And well, I was still interested in large arrays. I'd been interested in large arrays after the VLA. What comes next? Well, Square Kilometre Array. Boy, I started working on that 20 years ago. So I went to JPL. I convinced the director, Charles Elachi, the DSN needed a large array, that they could do much better than 70-meter telescopes. They broke it apart into 12-meter telescopes. And he awarded $3 million dollars for me to do prototyping of a DSN array. And I worked on that, had a lot of help, of course, at JPL. We made a 12-meter antenna. We made 6-meter antennas. We made the feeds, the low noise amplifiers, all very successful. But then it came to NASA. Do you want to build a big array? And it was $2 billion dollars, had to be built at three longitudes and had to be much better than the 70-meter antennas. I forgot how many antennas were going to be in the array. And they just didn't have it on their priority list. They would have to give up things that were space. Okay. Good. All right.

[brief interruption]

Kellermann: 01:01:51

Do you ever have the urge to come back to NRAO? No? The reorder won't understand shaking your head.

Weinreb: 01:02:01

No. Why not?

Kellermann: 01:02:11

Well, I'll rephrase. Of all the places you've been, which do you think were the best and you're most productive and contributed the most?

Weinreb: 01:02:25

NRAO is a general answer. I think my first 23 years of my career at NRAO was my most productive, and enjoyed it. But I wanted to leave because I had this urge to do something different, something else, and then longer-term, or go back? No, never. Well, first of all, we really liked California. And things were pretty good at Caltech except I missed that professorship. That would have made it a lot easier to be at Caltech. But what came later, is I got kicked out of the electrical engineering department.

Kellermann: 01:03:27

That was after you retired, or?

Weinreb: 01:03:29

No. No. I was a principal scientist at JPL, but also had an adjunct position in electrical engineering, and I had the five graduate students at different times. And it was this way. I funded the graduate students by selling low-noise amplifiers and feeds that they helped to develop. And it paid, I paid for the four years of salary and so forth for the graduate students, and good equipment for them. Quite successful. First question that came up is, "Well, you're selling amplifiers. Caltech doesn't sell. We are an educational and research institution." So I said, "Well, how can we do this?" Very good business department said, "Well, we need a letter, a document from each person who buys it that they're buying it for research, and they are an educational institution." So I could only sell the amplifiers on that basis. And that worked for about six years, and I brought in $6 million dollars. That was a dedicated fund called Amp, Amplifiers. And that was going great. Well, the new provost came in and he said, "Oh, no, we're going to get in trouble with the IRS." The business department said, "Oh, but we have all these letters saying, 'This is supporting research.' And the money that comes in isn't going to Weinreb. It's paying for research." Said, "Nope." I went back to the business department, the vice president of business. He says, "Well, whatever the provost says goes. It's his money, not your money."

Weinreb: 01:06:03

And so Provost didn't like that, that I even questioned. And then he said, "I want some of that money, that 6 million, back as retroactive overhead." Well, I said, "If I knew there was overhead, I would have charged more for the amplifiers. The books are closed on that. How can it be retroactive?" The business office said, “Whatever the provost says goes.” But that also created a bad feeling about me, I'd say, in the Electrical Engineering Department that-- Rutledge had left by that time. Had retired. And the head of all engineering, who didn't like this either. He wanted to side with the provost. And so at that time, Gregg Hallinan came along and Astronomy rescued me. He said, "Oh, if Electrical Engineering doesn't want you, just come over to Astronomy." The provost got his retroactive overhead. I think it was about $2 million dollars he took out of the 6. I had 6 left about that time. And I have spent a lot of it down, but I still have about 200,000 in Astronomy. It's in an account called Amp. I can use it for development work of amplifiers, or whatever I want to do with. Actually, this trip I'm charging to Amp. If I give a talk about radio astronomy, I charge them for that.

Weinreb: 01:08:13

So that's the story of my life at Caltech. I have a student now who is a graduate student in electrical engineering. And he's going to get a degree in electrical engineering. He can have a supervisor who is a professor in another department and Gregg Hallinan is his supervisor. And next month, we're having his thesis. It's called the candidacy exam. It's not the thesis exam.

Kellermann: 01:08:46

Yeah. I remember that.

Weinreb: 01:08:48

Yeah. Yeah. He's very good. He'll do fine with that. And actually, I have another student now who's a senior who wants to go on for a PhD. I may be his, not official supervisor, that'll still have to be a professor. But he wants to do filters and RFI, the effect of RFI in arrays. And I think that would make a good thesis. We've had some RFI surveys of Nevada sites, but not a deep survey. I want to do a survey using my very low noise amplifier and integrating the RFI way beyond the commercial equipment RFI survey that we've done so far. So Sarin is his name and work on that.

Kellermann: 01:09:56

So you have an appointment in the astronomy department now of some kind?

Weinreb: 01:09:59

Now I'm retired. I'm retired from JPL, and my adjunct position is retired too. I'm all emeritus or retired.

Kellermann: 01:10:14

Yeah, but you have some sort of appointment. You must have some sort of appointment to be able to spend Caltech money. [laughter]

Weinreb: 01:10:20

Don't ask. That's going on. I use Gregg's secretary. But the business office knows it's me. I approve the purchases of that money. It's working so far, but--

Kellermann: 01:10:45

To change the subject, what do you know about-- I just read some news release. This power transmission has satellites, solar panels, and beams down to Earth.

Weinreb: 01:11:04

Yes. Yes.

Kellermann: 01:11:05

I mean, as you know, the Japanese were pushing this decades ago, and we were all terribly against it. And I was horrified to see Caltech supporting it.  It is very clever. It's an array of transmitters, but--

Weinreb: 01:11:23

I don't know if I would describe it as very clever. I would say the implementation, there's some good research being done on lightweight solar cells. And there's no doubt that the solar cells will be more efficient out of the atmosphere.

Kellermann: 01:11:44

I was thinking of the phased array transmission.

Weinreb: 01:11:47

Right. But the transmission of the data back to Earth, you need two things. You need the collecting area on Earth, and you need the transmitter in space. And I think those are just way too expensive. I think this is impractical except in some special cases, which the military has been convinced of if-- well, I don't know to what extent they're convinced. But if you have a battalion off in Afghanistan and you've got to send fuel, or you need power for that battalion, it's very expensive to truck the oil in. You might do that with a satellite in orbit with solar cells, and then transmitting it to an array that you got to build for that battalion to receive that power. So that gets funding. That helps it get funding. But I don't think it's practical. In the whole question of climate change and CO₂, I think the answer is nuclear. You can do so much with solar cells. You need a big solar farm, nuclear-- safe nuclear, and fail-safe nuclear fuel that doesn't melt down. And I think that can be done. So other questions?

Kellermann: 01:13:55

Oh, where do you see radio astronomy going?

Weinreb: 01:13:58

I'm reading your book [Star Noise]. [laughter]

Kellermann: 01:14:00

[That's fast, that's fast?].

Weinreb: 01:14:02

It's the last chapter on that. I think the idea of an array could get order of magnitude bigger. There's 2,000 might be 20,000 next. I think the low-cost receiver, which results when you don't have any cryogenics-- I think that will be proven with DSA 2000. We would like to move it to higher frequency. See, the noise temperature is still proportional to frequency. And we're doing 11 Kelvin up to 2 gigahertz. But if you want to do it at 8 gigahertz, it could be 44 Kelvin. That still might be--

Kellermann: 01:15:04

That's getting pretty high.

Weinreb: 01:15:06

Yeah. There's another approach and that's to cool. One thing that I did find out about the transistors - and this is what Marian [Pospieszalski] doesn't understand or agree with - is the noise temperature is-- goes down rapidly up at the 300 Kelvin end. If we can go from plus 20 to minus 40, we gained about 4 Kelvin out of the 11. And that could be greater at higher frequencies. If we had to build an 8 Gigahertz array, we would probably cool it with Peltier cooling, solid-state cooling to minus 40 or maybe minus 90 even. And that would gain back a lot of noise temperature. So that's something that could be worked on.

Kellermann: 01:16:18

You've been following the SKA developments?

Weinreb: 01:16:23

Yes.

Kellermann: 01:16:25

What do you think?

Weinreb: 01:16:26

The International SKA, I've followed it for 10 years, and it's always, "Where's the funding?" Do the countries agree to fund it requires a treaty between the countries? And I think that has been approved.

Kellermann: 01:16:44

Yes.

Weinreb: 01:16:44

Yes. So I think that SKA will go ahead. It's hundreds of antennas, and it's kind of a low-end in Australia and a high-end in South Africa. Yeah, I think that'll--

Kellermann: 01:17:04

Well, what's funded at a level that's somewhat vague, it's somewhere over a billion dollars or euros, is only a very small fraction of SKA.

Weinreb: 01:17:17

Yeah. It's SKA One. Yeah. Right. It's not a square kilometer.

Kellermann: 01:17:22

Right. But we were both very-- I'm not sure whether we're as ambitious or naive or-- you convinced me once that-- well, we put together between the two of us the proposal for the US and you costed it. It was somewhat under a billion dollars, but it was a full square kilometer or half of that because of the lower system temperature. But it met the spec and I believed it.

Weinreb: 01:17:53

Yeah.

Kellermann: 01:17:55

Where did we fall down? I guess partly the antennas being able to turn out these monolithic hydroformed dishes--

Weinreb: 01:18:08

Hydroformed antennas. Well, the billion dollars, of course, was 10 years ago. It's maybe two billion now. But I don't think we could build a square kilometer for $2 billion. The non-cryogenic-- well, the square kilometer array has to work up at higher frequencies. So, yeah.

Kellermann: 01:18:42

I would say what convinced me that you could do this electronics and the signal processing is that device that you're holding in your hand.

Weinreb: 01:18:50

By the way, is this recording working?

Kellermann: 01:18:52

Yeah. Which is fantastic when you think about what's in there for a few hundred dollars.

Weinreb: 01:18:56

Yeah.

Kellermann: 01:18:57

14-channel GPS receiver and everything else?

Weinreb: 01:19:04

Well, I see the data processing part of even a square kilometer array as becoming more and more feasible. I am not optimistic about the US part, the ngVLA. I think that SKA International has a head start now in funding.

Kellermann: 01:19:34

Construction [crosstalk]--

Weinreb: 01:19:36

Construction, yes. And the latest thing I saw about ngVLA is completion by 2037. And the $2 billion is still big chunk of NSF's budget unless the science funding situation changes, which I don't see happening. I don't have much optimism.

Kellermann: 01:20:09

Well, I think it can get funded because it's well within the cost envelope of what the NSF plans to spend on new facilities. Question is, by 2037, is that enough of a leap forward?

Weinreb: 01:20:32

Okay. Yeah. By the time it gets built and International and SKA, even phase one, is going, "Okay, it doesn't go quite as high in frequency as the W band."

Kellermann: 01:20:47

ngVLA goes a lot higher. Yeah.

Weinreb: 01:20:49

Yes. Well, it goes twice as high I think. I'm not sure whether phase one of SKA goes to 50 or 20 or 10 even 13.

Kellermann: 01:21:02

No. No. 10. Yeah. So the ngVLA is intended for high-resolution thermal like ALMA.

Weinreb: 01:21:21

Yeah.

Kellermann: 01:21:22

Whereas the SKA is non-thermal. So they're different, but.

Weinreb: 01:21:30

Okay. And DSA is non-thermal. There's a gap between DSA going to 2 and SKA going to 20 maybe. So I think there's-- and SKA, the funding is there. The treaty is there. It's a 50-year treaty I believe. And it provides a mechanism for doing the full SKA. It's an international radio astronomy observatory. So I wanted to get more feeling here about ngVLA. You just told me some of the rationale. I don't know if [Tony] Beasley is coming to my talk or I'm going to talk to Beasley at all. But I don't think I can convince him that ngVLA is not going to get funded, or certainly has to continue trying to get the money. Who knows what happens?

Kellermann: 01:22:55

As you may have heard, they just got funding for part of the engineering design.

Weinreb: 01:23:02

What was it?

Kellermann: 01:23:04

I think 21 million.

Weinreb: 01:23:06

In three years, over three years. Not each year.

Kellermann: 01:23:09

That's correct. Tony thinks it will be spent over just two years. He seems to be optimistic about--

Weinreb: 01:23:27

Where're the engineers going to come from? The engineering staff of NRAO is getting older too. Are they drawing in new help?

Kellermann: 01:23:37

Yes. In New Mexico, I think they've hired a lot of people. So in fact, I imagine that this $7 million dollars a year is mostly to maintain the marching army because there have been special NSF funds for the ngVLA for the past years. Certainly in the case of the VLBA, which was entirely done, or the design work was done in-house, without any special funding, until the construction project was funded. And I think the VLA was the same, wasn't it? Until the construction project was funded, I mean, you and others were just--

Weinreb: 01:24:37

We got millions of dollars each year to develop it, to design it. Before.

Kellermann: 01:24:42

[inaudible].

Weinreb: 01:24:44

Yes. Sure. And it created this change of the design completely to correlator--

Kellermann: 01:24:50

Oh, that was after the project has been approved. After the was approved and the funding was there, they sat down and said, "Okay, now what are we going to build?"

Weinreb: 01:25:04

I don't know. When the project was approved, a budget was approved, which included all of these improvements.

Kellermann: 01:25:17

[inaudible] the idea.

Weinreb: 01:25:21

It wasn't the original budget. Dollar-wise, it wasn't more than the original budget.

Kellermann: 01:25:30

Well, it because got stretched out. Same thing with the VLBA. Because the NSF stretched it out. They agreed to put in the extra funding, which saved the project, I think, in a way, and same with the VLBA. I know in the VLBA, the proposal was for three years, which was completely unrealistic. We never could have done it in three years. But because they stretched it out and then added the extra costs incurred by the delay.

Weinreb: 01:26:05

Who's in charge of the NSF?

Kellermann: 01:26:09

Oh, I don't remember his name.

Weinreb: 01:26:13

What about NSF Astronomy?

Kellermann: 01:26:16

I don't remember her name.

Weinreb: 01:26:18

Martha Haynes?

Kellermann: 01:26:21

No, no, no, no. Someone you won't know. She came from some other part of the NSF, I think. That keeps changing.

Weinreb: 01:26:28

I sure wish Bill Howard was still there.

Kellermann: 01:26:30

Yeah, he fought for a--

Weinreb: 01:26:35

Well, turn the recording off.

Kellermann: 01:26:40

Yeah. Right. We're getting into delicate territory.