NAS Fitcheck Day 12: In Transit

Extraction of the NAS Fitcheck team is underway. We all came down off the mountain today, feeling good about a successful run.

The first view of the ocean. It's at this point on the drive that your skin starts to feel normal again as the air becomes humid.

Here’s our quote from the ride down: “Pirate mining? . . . no no, you probably said iron, but I heard pirate.” (Tyson Hare)

Jason is spending the night in La Serena, doing some exploring. Laird, Alan, Tyson, and Jared have made it to Santiago, despite a bunch of airline computer problems. Here we commemorate Jared’s first Pisco Sour at the SCL Holiday Inn:

We have a 6 hour layover in Santiago, so we're spending it telling war stories from the run.

Days without a motherboard failure: 11

NAS Fitcheck Day 11: Back To The Aux

A light day today, if you use the usual MagAO tempo as your standard. Alan and Tyson did a little work on the guider this morning.

Tyson at work on the guider with the NAS still mounted.
Here's the new and improved x travel limit of the guider probe.

We had some excitement today. Power to the telescope was accidentally shut off, and there was a brief attempt to blame Jared. It wasn’t him (based on what I was doing at the time, it wasn’t a dumb idea). We learned that the rotator breaks release if this happens, so we will be extra careful to make sure the rotator is pinned when we are woking inside the NAS.

Once the power was back on we craned the NAS off the telescope, and lowered it back down to ground level.

The NAS weighs 1800 lbs now, after adding counter weights during the balancing procedure. We apply this much preload with the crane so we can take the bolts off without dropping the whole thing.
The NAS making the journey from the Clay to the Aux building.
Laird says goodbye to his baby.

There was also a long earthquake this afternoon. Anecdotes suggest you needed to be sitting down to feel it. It was a 3.9 according to Universidad de Chile Depto. de Geofisica

After lunch we cleaned up after ourselves, inventoried our gear, made a shopping list (we go through zip ties like nobody’s business), and generally took it easy. We did install our brand new ASM cover:

The new ASM cover, and Jason Lewis.

After that, there was some relaxation, a run down to the 100″ and back, and maybe some snoozing. Once you get kicked off the telescope that’s it.

Another solitary horse today.
The LCO sky put on one last show for us. This was late afternoon. (click)
Tyson grabbed a nice sunset picture at dinner. (click)

Days without a motherboard failure: 10

NAS Fitcheck Day 10: MagAO In Control

Today started with balancing the NAS on the rotator. We added two big pieces of steel at the top to keep the rotator motor currents even as the instrument rotates.

A shot of the NAS, with a meter hooked up to the rotator for reading currents. You can see the two weights we added by the guider box. The NAS started out very bottom heavy (it's upside down here).

Once that was done, we attached the Anaconda – our affectionate name for the cable that we’ve been spending so much time worrying about wrapping and passing through the hole. We’re in South America, and it has almost strangled the entire project several times in the last week so it seems appropriate.

Here Jason and Tyson feed the Anaconda through the hole:

We feed the cable up through the hole. Here we see that we have 1 meter of slack when the rotator is all the way over.

Here’s Jared connecting the data fibers:

Jared connecting the network fibers.
After being attached to the NAS, the cable runs through the floor. That hole was cut just for this purpose. You can see why our design has a bent PVC pipe in it, since as-is our electronics boxes will snag it. For now, we always have a person there to guide it around.

Jason was our snag watch:

Jason signals that the rotator is at 0 degrees (that isn't a gang sign). We tested how the cable moves and drapes at various orientations of the rotator and telescope.

And he also connected the cooling circuit:

Jason working on the CCD cooling system. We ran it, and the electronics box loop, all without incident today. No leaks.

Once the Anaconda was under control, and the pumps were running, we powered up. Everything came right up, and we aligned the system on our test source.

Jared, Alan, Tyson, and Laird celebrate a working WFS and VisAO camera attached to the Clay Telescope.

And we also made sure it works sideways:

The PWFS pupils with the NAS rotated 90 degrees. I had to think about what the X-Y-Z stages were doing, since they weren't the same X-Y-Z anymore (they are the same optically, but gravity did some funny things).

The next big item to check off our list was testing for possible collisions between our moving X-Y-Z stages and the telescope rotator bearing. The problem is that our stages move at an angle relative to the plane of the rotator ring that we bolt to, and it has a step in it so it changes height. The problem is hard to visualize, model, measure, and not worry about. To test this, we ran everything from one end to the other, with all scary combinations of full travel in the stages. Despite some doubt on the part of some, we are safe.

Maybe this helps you appreciate why this has been driving us nuts for a long time. This shows that we clear with X and Y at their 0 positions, the Z stage all the way back, and the CCD 47 focus stage all the way back. For the record, we have about the thickness of Jared's thumb to spare. Everybody should remember that when we can't focus with the Wollaston and need to change some limits.
This just shows that our stage travel limits, though safe, aren't even astronomically useful. L1 is too high to even work.

All of this initial testing was done from the platform in the dome. The next big step was to move into the control room. Here we’ve just set up the PWFS pupils of MagAO in the Clay control room for the first time:

The MagAO system being controlled from the Clay control room.

Once we moved to the control room, it was time to take control. An AO system periodically needs some help from the telescope to correct some low-order wavefront errors, such as being pointed in the wrong direction. We call this ‘offloading’. To do this, our software has to tell the telescope to move. We also send position commands to the secondary support system (the ‘vane ends’) and – take a deep breath – actually send commands that change the shape of the 6.5 meter primary mirror. We tested this process on the real telescope tonight after dinner. In this video you can see Glenn and Jared nerd-out when it all works as we planned:

We then tested the software interface for Clio2, successfully performing a ‘nod’ in RA and Dec. It’s also important that we are getting data from the telescope so that we can record it for analysis later. In this clip you can see that the VisAO system is getting information about the parallactic angle of the target, as evidenced by the rotating green arrow.

At the end of the day, we had some dark time in the dome which we used to do some scattered light tests. Our system looks really dark. We also did some read noise measurements with the CCD39 and nothing changed from the lab – a big relief. The NAS is now uncabled, and will be pulled off the telescope tomorrow morning. We still have to pack it up and put it away safely, but all of our major testing is done.

Quotes of the day:

Alan: “I’m starting to have Laird’s nightmare over here.”
Jared: “Don’t go all wobbly on me now.”

Glenn: “We’re perfect. We’re on target.”

We spotted our first horse at lunch today. Only one though.

An LCO horse. They got a lot closer last time.

And Jason found a whole new population of Viscachas on the other side of the Telescopes.

A Viscacha suns itself north of the telescopes. Click.
A Viscacha on the move. Click.
They have striking black stripes on their backs and tails. Click.

And finally some more ornithology from LCO. These little guys make a lot of noise, and until I really pay attention, my farm kid ears hear redwing blackbird. That doesn’t make sense (there are no cattails here), so I tracked the singer down.

Days without a motherboard failure: 9