Guiding out the waves: engineering planetary radar

I spent last week observing of asteroids from the telescope: our first night was just another 8-4 workday night, where we looked at space rocks in our neighborhood and out beyond Mars for eight hours.

Arriving at the control room, I sat down with my binder full of… notes on how to observe at Arecibo, which hadn’t made much sense.  At any optical observatory you enter your coördinates into a computer and take an image.  After verifying that you pointed the telescope correctly, you then tell the telescope system to take data for the rest of the night, occasionally adjusting pointing or focus.

Hector está llamado por telefono

At Arecibo, you’re pointed in the right direction.  Alignment with the William E. Gordon Telescope is not an issue.  Even with the 305-meter dish, you’re good to a few millimeters.

Position of the platform, receiver, transmitter

Getting the signal out of the receiver and properly into the computer is the hard part of observing.  Instead of a few shiny silvered mirrors and a charge coupled device digitizing and sending your photons to a screen, here a maze of waveguides, cables, and wires brings signals from the matte metal dish, after being ushered into the receivers, along a path 1,600′ long to the control room.  Where computer monitors would display starfields at an optical observatory, wavy lines danced across oscilloscopes at Arecibo.  It felt like junior year electronics laboratory again in the physics department, so different from most of the things we were doing in astronomy, and not just because of cgs/MKS units arguments.

Ellen had considered walking me through cabling the week before our six-night-long marathon observing run, but ultimately decided that it wouldn’t make sense out of context.  Wait for the actual observing run to understand the cabling.


In the overly-air conditioned room the panels of cables with cryptical labels looked daunting.  Mike stepped in, ripped all of them out by the BNC connectors, and handed me a cable.  I idly fingered the BNC connector.

“Everything begins with the Planetary Downcoverter, labeled 7-19 in rack #7.  Split/attenuate the signal from there with 7-3 (opposite sense signal) and 8-3 (same sense signal).  For continuous wave, set to ~40 dB (OC ~41 dB; SC ~47 db as of 2012).  Filter with the Butterworth 125 kHz bandwidth filters, 7-6 and 8-6…”  Somewhat in a state of shock and disbelief I obligingly followed his instructions as I stuck BNC connector after BNC connector into the panels.

BNC cable octopus

Later in the night, Mike proceeded to recable for another asteroid with a different setup.  It was like watching an anime character’s arms in a blur, and a moment or two later, the blur subsided, and lo, things were cabled.

“How long did it take you to learn to do that?”

“About a month.”

After attenuating, splitting, filtering, amplifying, and connecting, the signal winds up on an ancient Sun Sparc Station, of similar vintage to the author, where you type a few commands to begin data taking.  Abstracted away are details of where the asteroid is, how long light takes to travel there, and what sort of data you’re taking (continuous wave, or encoded?).  Type five letters, confirm with the transmitter operator, and bam, you’re sending a million watts of S-band signal off to the asteroid.  It returns a few seconds (or minutes) later and you measure the few parts in a billion of signal that you get in response.

Sparc Station

The transmitter is controlled a delightfully archaic display, looking something like an old-school video game housing, with 8-bit graphics and audible alarms when the filaments or thermionic cathodes in the klystrons stop behaving as they ought.

Radar transmitter; old control console

Recable, process, and magically, you get something like this:

We had six nights on the telescope, the first one going until 4 am; subsequent ones going until around 2 am.  After the first few, Mike left early and I finished the runs.  The second-to-last night I tried cabling on my own, but was somewhat foiled by the remnants of the aeronomers’ cabling scheme from earlier in the day.  The last night it became a lot easier.  The transmitter had other ideas, but we have data and can process it.

Observing here is much more about engineering than physics  How can you reconfigure cables and machinery to do your bidding?  Arecibo is much more hands-on than any other place I’ve been: everyone rolls up their sleeves and gets covered in grease several times a week.  While some details are hidden away, the actual process of taking data is much more transparent, and you can visualize the photons through the waveguide and down to the control room.

Some irony: the control room is air conditioned to about 55 ˚F.  So cold that after weeks of wearing nylon hiking pants or shorts and a t-shirt, I was freezing in jeans, closed-toed shoes, socks, a scarf, sweatshirt, and fleece jacket.  Going outside, into the cable car, and up and down 250 stairs to the receiver room to evaluate the instruments was quite warm in comparison.

After the first night, the telescope operator on duty, Norberto (pictured above), asked if I wanted to go up to the platform and uncover the receiver for another experiment.  Sure!  I ditched the fleece and scarf and hopped on board the cable car in the dark.  We waved goodbye to the control room and car housing.  A firefly drifted by below us.

Cable car housing

Up on the platform we could see the first quarter moon descending to the west beyond the tower.

Moon and tower

Norberto led us into the rotary joint.

Rotary joint


Climbed through the joint

Below the joint, looking north.  To the left is the control room; to the right the visitor center.

Tower lights

Down the numerous flights of stairs.

Down the stairs


We uncovered ALFA and went back up 200-something stairs to the cable car, back down to the control room, and then I hiked another 125 steps back up to Visiting Scientist Quarters.  Bed at 4:30 am, zzz…  I think I prefer the telescope after dark.

8 thoughts on “Guiding out the waves: engineering planetary radar

  1. Wow, simply fascinating post, Sondy! Your photos are beautiful, too. I’m surprised their tech is so outdated there. It seems they might benefit a little bit from upgrading… any reason why they don’t?

  2. As for the cabling system, you can use it in ways it was never intended: it’s hard to program a box to be utilized in ways you never anticipated when building it. And it works, which is great when you’re on a budget.

    • The cabling system actually looks pretty fun to work with. Why route data through a software pipeline when you can route it through a hardware pipeline instead? 🙂

      • Hi Sondy!

        I was there many moons ago, building telescope control and data acquisition systems.

        Did a lot of work with the pulsar astronomers, and was privileged to watch Joe Taylor do the work that won him the Nobel Prize. Also worked with Aeronomers (including Michael Sulzer who it seems is still there! Actually helped build his house) and built the first system (named “Simple”) that allowed full control of an aeronomy campaign. On the astronomy side, spent a lot of time with Riccardo Giovanelli and Martha Haynes. Spent quite a bit of time with Jill Tarter as well when she would come down and do her SETI work.

        Ray Velez and Angel Vasquez were the lead telescope operators — Angel is still on the phone list — is he still there??

        I eventually bought an old wooden boat, fixed her up, quit my job and sailed the Northern Caribbean for a few years, before sailing back to the States and getting a job building ground support systems for the Hubble.

        I have been very fortunate and feel very privileged to have had these opportunities and to work with some of the world’s pre-eminent scientists.

        Please say hi to any that might remember me. None of them would recognize me — back then I had a ponytail and beard. (I will try to find an old picture to help jog their memories.)


Leave a Reply