PFS Upgrade Series, Day 5: So Good It Can't Be Right...Can It?

This is part of a series of posts about upgrading an instrument at Las Campanas Observatory. If you want to start at the beginning, it's here.

Yesterday we aligned the lenses in PFS as best we could, and measured small optical aberrations, small enough that we didn't want to change any of the lens orientations. Today we wanted to isolate and measure any potential optical aberrations in the prism in the instrument, which is rather thick so could have a gradient in its index of refraction that could contribute to the distortions we see in the data (or at least did, prior to these upgrades). 

First Jeff and I installed the "can" that houses the grating and prism for the instrument. We made a time lapse that isn't terribly exciting, but gives a good idea of what instrument work can be like, I think. I stopped recording before we removed the grating, although you can see Jeff take off the grating cover in the video (when we are wearing the face masks to avoid spitting on it). It was hard to find a good position for the camera to record the grating removal, plus I didn't want the added pressure of looking pretty on camera. :) The grating was heavy and in an awkward position, but Jeff took it off without a scratch, like a pro (which he is)!

Above: View of the grating (big rectangle thing) and the prism (mirrored thing) after we took the lid off the can. You can also see the *real* clean room in the background. We're being pretty careful but haven't gone full bunny-suit. Below: View of grating from the other side. Those silver poles help hold the lid of the can, which we took off. The black background is one of the long sides of the instrument. 

After we removed the grating, we put the same reference mirror from yesterday in the place of the grating; you can see it in the first photo below, looking at the mirror from the back towards the wavefront sensor. Jeff build a special holder (metal plate in the second photo below) to support the mirror at the right angle. You can also get a sense in that second photo for how efficiently the design of the instrument is in terms of space -- it's a tight fit but everything is perfectly organized. Steve told me a few days ago that this instrument is a work of art, and Jeff is "an artist" (pronounced "ar-teest").

As you might expect, we had to futz around with the wavefront sensor for a few hours to get a measurement of the prism that we believed. But the result was surprising -- the small aberrations that we found yesterday from the lenses alone changed *very little*. The only really significant change was in one of the astigmatism values, which changed by about 0.1 microns; everything else changed by 0.01-0.02 microns at most! We think (and spent some time convincing ourselves) that this means the prism introduces no optical aberrations, and that the source of the distortions we see in the actual data from PFS is the exact placement of the detector/dewar. Yet again, it seems that Steve and Jeff did an amazing job when first putting together this instrument! Oh, and that the prism is good; Steve suggested we send a holiday card to the company/people that prepared the prism for the instrument. 

Our set-up for the day. Note the mirror on the right side. It's sort of starting to look like a spectrograph again!

After lunch Steve was determined to see if we could get an even better measurement on the whole lay-out. We had been using a reference image that was provided with the wavefront sensor, to get a relative measure of the aberrations that took out whatever "initial" aberrations were present in the wavefront sensor measurement (without going through any lenses or mirrors). It's kind of like if I stood on a step and you measured my height -- first you'd want to know how tall the step was, then you could back out my actual height from (step+Johanna) and (step alone). 

After lots of iterating on this intrinsic measurement reference image, and applying it instead of the provided reference image, we *were* able to decrease the aberration values! Not by a huge amount, but at this point even 0.1 microns was significant. I don't have a picture of that pupil image or those zernike values, but the largest one was about 0.25, and the final wavefront RMS was 0.14 microns! (Okay, that's not counting the spherical aberrations, but we can't do anything about those.) We are happy. 

Before dinner, Jeff started pumping down our shiny new CCD to get it ready for testing by our colleague Christoph, who arrived today from California. Below you can see the difference between the old CCD/dewar (right, standing upright) and the new one (left) -- the new one is larger because it holds a 10k x 10k CCD with 9 micron pixels, versus our old 4kx4k CCD with 15 micron pixels. The new CCD was necessary to increase the sampling to match our new fiber feed (that will replace the traditional slit where light enters the instrument). The fiber feed isn't ready yet (whomp whomp), but hopefully that will get installed in the next trip. 

The theme of today seemed to be birds that hunt/eat meat. On my walk this morning I saw a beautiful owl -- the photo below doesn't do it justice -- who followed me with its eyes as I walked by. It's a little disturbing how their heads turn all the way around. Then later during the day, we all had encounters with a hawk (or maybe hawks, there are two) that flew close to us and in Jeff's case was curious enough to approach him while walking on the pavement. Are birds of prey our good luck charm? We'll see. I'm a little worried they are driving away the viscachas!