Blog

Today is still ongoing, as I write this blog post at 1 am. I’m currently holed up in the library at the ASB writing my butt off for this 51 Peg proposal. It’s basically like 12-16 pages of highly polished material that they’re requesting for the application, so it’s a lot of work. To keep my morale up, Warren made sure I could read all about our avian brethren that live in the Argentinian region of South America by finding this in one of the many bookcases.

Lunch today consisted of some interesting Chilean-Japanese fusion. I grabbed the “pastel de choclo” which is like a shepherd’s pie, but with a corn topping instead of mashed potatoes. They also had vegetable tempura. This tourno of chefs is killing it, man.

After lunch we returned to our regular duties. However, before dinner we stumbled on an annular eclipse, so we had to go out and check it out. Apparently totality is happening directly over Easter Island, which honestly sounds like a religious experience. Well we here at LCO we ended up getting something like a 30% eclipse. Better than nothing…!

In MagAO-X engineering news…

We got a chance to admire the fruits of our alignment labors last night. A PSF comparison between camsci1 and camllowfs. We want these to look similar, indicating good alignment. So it’s looking good!

We had trouble measuring some optics. Inches were too fractional we couldn’t find any metric calipers. We present a new unit of measurement for the lab: a micromaggie, which is 6.75e-7 inches or like 2 microns or something. Much more useful and reliable.

ICC2 has been sick. In fact, this morning we awoke to it looking almost terminally ill. It wouldn’t boot to the OS and just kept getting stuck at the BIOS, despite working relatively fine the day before. Although it was having trouble recognizing hard drives before it got super sick, usually a reset was enough to fix everything.

Luckily, we got our best doctors on the case (Jared and Joseph). Despite the confusing symptoms and near-death experience, ICC2 managed to make a full recovery once Jared and Joseph found the right treatment for the ailment. I’ve been watching old episodes on House M.D. lately, so I’ve almost been picturing the legendary Dr. House and Wilson themselves working on ICC2.

I think it would’ve made for a great episode! Maybe you’d all like to know- it wasn’t lupus that was making ICC2 sick. Although, it’s *never* lupus…

Song of the Day

I was going to pick Massive Attack’s “Teardrop”, initially released in 1998 because of the House M.D. analogy. But I’m pretty tired. And I think everyone else is too especially after a couple back-to-back 16 hour shifts. So maybe Brain Stew by Green Day better captures the mood here.

Walking back in the dark with Jay tonight under quiet stars, I remarked that “even though I arrived yesterday it already feels like a week, but in a good way.” Long days can do that – leaving the cleanroom at 12:00am was pushing well into our sixteenth hour of work. Time in general though has a more, ethereal quality on top of the mountain. It can be hard to remember that life continues outside of the immediate challenges of the day, and that every day before now hasn’t been just like this one.

The main goal for today was to align the LLOWFS – this is Lyot Low Order Wavefront Sensor or “YO-FUSS” using the Spanish pronunciation. When paired with a transmissive coronagraph, rejected light from the Lyot stop provides “free” photons for driving auxiliary wavefront control. The system had been roughly assembled in Tucson and we had poetic alignment instructions as courtesy of Sebastiaan Haffert and Chat-GPT:

In the lab where lasers gleam,

Align the Lyot, chase the beam.

Turn it on, ensure the light,

Hits the center, sharp and bright.

Flip the clamps, turn them around,

Old positions, stoppers found.

Pull the wheel, the system’s grace,

Install the mirror in its place.

Put it back, a careful feat,

Center Lyot, alignment neat.

Tilt the laser, shift with care,

Card in hand, another’s stare.

Lens one’s heart, its center true,

Shim stock, tilt—just right for you.

Centering near, within a mill,

Precision holds, steady still.

Lens one now done, we move ahead,

Z filter next, where paths are led.

Remove the tube, the chip exposed,

The cap now off, the system grows.

Pivot here, or pivot there,

Camera moves with patient care.

Optical feedback, guiding hand,

Or move again, the next command.

Install the tube, frosted within,

Pinholes small where light begins.

Lens to guide, align once more,

Precision sought, forever sure.

As most alignment projects go, lived experience quickly derailed from our flowery aspirations. While Jared fought down a flurry of unexpected problems that I won’t pretend to summarize, Maggie, Jay and I slogged through the full range of alignment processes and human emotion necessary to correctly focus the reflected beam onto the LLOWFS sensor.

The main difficulty in alignment came from the sheer number of things (I’m going to stick with “things” here because my gut reaction at the end of the day is to use a couple swear words but a lot of time went into each part) in the neighborhood around the Lyot wheel. Opening the panels this morning made me think of the frog that hops into lukewarm water and quickly finds itself boiled alive. When I started looking at PIAA designs in early 2022 this area was fairly crowded but manageable. Years later the optomechanics have grown layers deep and maneuvering a stubby Allen key through the optical jungle can take thirty seconds and Zen breathing techniques.

We added a mirror to the Lyot wheel to use a reflected beam for the alignment. Picomotors on the stage means that translating the wheel in x and y are quite straightforward, but using the wheel as a mirror instead of a simple aperture means that the tip and tilt in both directions became critical. The immediate descent into shim madness gave a portent of the long day in store for us.

Aside: it’s rather silly to spend much more time writing the details of our work today when it’s already late and breakfast will be served in six hours. I’m going to scatter a few photos through the post and call it sufficient. The biggest takeaway from the day though is that external challenges are always manageable with the right team around you. Despite the setbacks and technical frustrations that were part of the alignment process, problems became fun, interesting challenges when shared with the group. I appreciate doing this work on an engineering run, with no rush at night and enough time to listen to each other’s perspectives. Amid this collaboration and working in such a beautiful place, one long day makes us only more prepared for the next.

Song of the day

I had about twenty songs stuck in my head through the course of the day. When doing delicate or technical work I like to nurture their memory on repeat because I find that tends to relax the high stakes. Of the songs that I played during breaks afterwards, sitting outside the cleanroom and appreciating the warm sun, this was the only one that met the blog rule.

Warren and I made it up the mountain without any necessary running. However, due to the leftover storms from Hurricane Helene, the ride was bumpier than usual thus less sleep than usual.

We didn’t want to be left out of Empanada Sunday, even though we technically didn’t arrive till Monday morning, so we managed to find a tasty new, or at least new to us, spot in the Santiago airport.

Up on the mountain, spooky clouds were a-brewing.

Okay they were kinda pretty. After dinner we got to work. Jay and Jared worked on telescope proposals while Warren and I staged the instrument for upgrades tomorrow. We moved PIAA hardware out of the system so Warren can install alignment motors on them and I’ll have room to install a reflective Lyot mask for LLOWFS. For unfamiliar readers, PIAA stands for Phase-induced Amplitude Apodization and is a type of coronagraph that utilizes four specially shaped lenses, two on either side of a light’s focus with a diffraction mask placed directly in the focus. This device will allow us to block out light from stars with Earth-like planets orbiting very close-by. Lyot-plane low order wavefront sensing (LLOWFS) will allow us to track lower order modes of turbulence in the atmosphere and enhance the PIAA’s function.

Only bad news of the day: The focal-plane LOWFS (FLOWFS) shutter gluement failed. Luckily, we have three jam packed days here in Chile to work on a bunch of engineering tasks in prep for our November run.

Song of the Day:

We aren’t at a hotel in California, but an observatory motel in Chile is close enough. Hotel California, The Eagles, 1977.

One of the big goals for this visit to LCO was to troubleshoot a coupled pair of actuators on our precious 2040 actuator “tweeter” deformable mirror (DM). Being coupled meant that when we poked actuator 498, actuator 650 moved (not 498). Actuator 650 also moved when we poked 650. We first noticed this problem after shipping MagAO-X in March, when we ran a post-cabling function test. However, digging through our archives we found that it showed up back in November, 2023, but we just didn’t notice.

Now our tweeter is expensive. So we don’t like it when things go wrong, and it makes messing with it really stressful. But having those two actuators acting up means MagAO-X didn’t work perfectly, so we had to do what we can to fix it. We deferred it until after our March and April runs to give us time to plan and do it right.

We can all breathe a sigh of relief: with only minimal shenanigans we got it fixed. After testing various potential cabling problems, we found that a circuit board just needed to be replaced. We had a spare, popped it in, and the couple had split up and gone back to behaving normally.

And we got that done in time for Empanada Sunday:

After lunch we got onto the next set of tasks, which included disassembling a rack shelf so we can get better airflow.

We’re also upgrading the Instrument Control Computer (ICC) to go faster and do more stuff. We like to liquid cool our CPUs and GPUs. Normally we rig that up in Tucson, but we couldn’t get all of the components from our usual supplier in time, so we shipped what we had (and found some substitutes on Amazon!) and are doing some clean room plumbing.

Jay has also been leading the assembly and testing of new ICC. Other than the liquid cooling parts, it was fully working in Tucson before we took it apart to ship it down here. We’re getting it all back together.

Even after our delayed departure, we are making good time thanks to the relatively easy fix to the tweeter. I haven’t had enough running on this trip yet, so I took some time before dinner to go circle around the 100″.

Let me tell you about our trip.

Jared and I were scheduled to leave on Thurs. 09/27, however our Tucson flight was delayed to the point where we would have missed our ATL-SCL connection, so we got the opportunity to rebook. Whether or not this was due to Hurricane Helene, we’re not sure. Because ATL-SCL on 09/27 left on time, more or less, despite being closer in time to peak hurricane’ing. Our original TUS-ATL flight arrived about 3 minutes before the plane to Santiago left the runway, so all in all, it was the right move to rebook.

Because we rebooked, the revised departure date was the next day, Friday 09/28. By about 9a on Friday, the flight was already delayed by 20 minutes (we thought: “here we go again”), but thankfully the delayed departure time didn’t change again…. all the way until the VP landed in TUS which (we painfully learned) shuts down the whole airport for half an hour. This new delay was coupled by the plane being overweight as well because of the unseasonably hot tarmac temperatures this late into September. So, the poor souls on standby who got onto the plane had to be escorted right back off to bring the curb weight of the plane down. After all that, we departed TUS *finally* at around 2:10p, a final delay of about 1hr 20mins. We were set to arrive at ATL by 8:15p EST at Terminal B. Our ATL–SCL flight was set to depart right at 8:55p, from Terminal… E….

We ended up making it with mere seconds to Gate E14, but Delta misjudged our dedication (and ability to sprint) and automatically rebooked us to fly to Santiago the next day while we were still en route to ATL. After a few tense minutes, however, the gate agents were able to get us back on our original flight and thankfully they didn’t close the door on us as it was definitely after 8:40p by this point. After taking our seat, we left the ground a few minutes after 8:55p.

We landed at SCL without incident but about ten minutes past 7a local time. Normally not such a big deal, but because we had gotten rebooked initially, our new flight out SCL–LSC was at 9a on the dot. The line to get through customs was, of course, super long. We finally got through at 7:55a. Boarding began at 8a, so, more running…!

SCL has also moved the security checkpoint to near the LATAM check-in desks. Again, normally not such a big deal, but Jared was never given a boarding pass when we deplaned out flight from ATL, so he had to get one reprinted before we could get through security. It was 8:34a.

Boarding doors close at 8:45! and the B gates were, no joke, like a third of a mile from the security check point.

We. Just. Made it…!

After taking the transport up from La Serena (and sleeping for most of it), we were just in time for lunch and were greeted by some familiar views.

Truth be told, after all the delays and setbacks and just plain bad luck, I never thought we’d make it this far.

Bonus Content

For today’s Bonus Content we were greeted by a ravenous goaty parade at sunset…! This made for many, many small “landmines” when walking down later that evening in the dark. Make sure you wipe your shoes extra good before entering your room during goat season.

Blog Rule

Short trip, simple rule. The blog writer must pick the Song of the Day from the assigned decade below:

Day 0: 1950s
Day 1: 1960s
Day 2: 1970s
Day 3: 1980s
Day 4: 1990s
Day 5: 2000s
Day 6: 2010s
Day 7(?): 2020s

Song of the Day

The song of the day is Frank Sinatra’s “Come Fly with Me”, originally released in 1958.