Despite some pretty poor weather this past week we are finished with the bright time engineering run and ready to start science over the 14 days around new moon. In this past engineering run we found some problems with the wave-front sensor software and implemented a new forward model of the hexapod motion. This new model allows for us to compensate for an unintended very small rotation that is induced by hexapod motion over the course of a trajectory. The impact of this rotation was to cause the field to rotate by several arcseconds over a trajectory. Our new forward model eliminates all of the hexapod induced rotation but allowed us to see a much smaller Azimuth dependent rotation term. This much smaller term will be addressed after this next science run, during the next full moon. Overall, a good improvement to the telescope’s operation.
This week has been an exciting week ramping up to science. The Board of Visitor Staff Excellence Award Winners were announced by Director Armandroff and are as follows: Henry Cantu, Angela Davis, Steve Odewahn, and Trent Peterson. Congratulations to the winners.
We are quickly moving in to science mode with the upcoming new moon. Currently this week the first half of each night is being used primarily for science with the second half used for engineering due to the moon. The weather looks to be clearing over the holiday weekend, so hopefully we will be able to collect some good data for the HET community!
This week we are firmly back in our 2 weeks of engineering time. The big news is that we were able to get the wave-front sensors to properly operate in closed loop mode. This means that we are able to optimize our image quality during a trajectory small tweaks to the overall tip and tilt of the corrector above the primary mirror.
The mirror teams has kept up their rapid pace of pulling mirrors out of the array and putting in recoated mirrors. The current swap rate is 1.67 segments per week with 69 segments having been swapped since 29 February.
This week we finally had a break in the stormy weather and started wave front sensor commissioning. A team from Austin came out and installed an imager and wave front sensor in the IHMP (Input Head Mounting Plate) to calibrate all other wave front sensors on the telescope. These devices once calibrated will help keep the tracker in perfect alignment with the mirror to optimize the observations we are taking for science. This will be the main focus for the next month. Hopefully the weather is more cooperative than it has been. If you would like to see a time-lapse of some of the recent storms follow this link: https://www.youtube.com/watch?v=shRD2WmPWd4
This week has been a busy one for the virus units at the HET. All of the virus units were removed from the enclosure, backfilled, pumped to vacuum and reinstalled. While doing this cooling fans were installed on all of the virus controllers to help keep the electronics cool. They hope that this will make the virus system a bit more stable.
The HET has been churning out lots of science during dark time this month as well. All groups that have submitted targets have received data!
Last week we reported on a incident with the HET tracker that put us in a hard skew condition and shut us down. We had recovered from that skew condition but the software issues that led to the skew had continued. This week I am happy to report that we have not only installed additional skew sensors but we have found the source of the software issues. The software teams in Austin and West Texas were able to perform a series of experiments which allowed them to isolate the problem as coming from the ethernet card inside the dSpace box. This is a very specialized ethernet card but thankfully we were able to swap out a different card for it and found that all of our TCS communications problems, hexapod faults, and loading trajectory errors have gone away. We will look into getting a spare card and send the card in to see what kind of fault occurred in the card.
With the improved tracker performance we were able to start our early science operations. We took our first science targets on the night of the 16th. Unfortunately, we are now headed towards full moon which is not the ideal time to do science targets with LRS2. However, it does give us a chance to work out any inefficiencies in our PI interface, night operations interfaces and queue tools so that we can be fully ready once we get some dark time a few days after full moon.
Two weeks ago we had a fairly serious tracker issue. While recovering from a fairly typical tracker failure, the telescope operator requested that the breaks be set but instead the tracker moved in X and generated a hard skew. This hard skew actually caused the bearing to come out of the bearing block. This occurred the day that the blogger and most of the chief engineers were going to their summer conferences. Upon the return the engineers were able to determine that no permanent damage was done to the bearings or the tracker in general. The engineering team was able to move the tracker back into the block using the tracker motors. Unfortunately, the software issues that led to the skew are still present and have not yet been sorted out. This has severely limited our ability to run trajectories for engineering or the start of early science operations.
In the mean time work continues on VIRUS enclosure 2 and with re-coating of the primary mirror segments.
We are currently in bright time (near full moon) and the currently active instruments, LRS2 and VIRUS are really designed to operate in dark time (when the moon is below the horizon), thus we spent this week analyzing the data taken in the last dark run, testing and updating the telescope control software and training the night time staff to make sure all of the staff have the same skills and knowledge of the system for the upcoming dark run.
One of the problems we were having in the last dark run was a mysterious end of track due to a hexapod hitting the software limit. Further investigation has revealed that this limit is part of the natural extension of the warming tracker. As the steel structure expands from warmer temperatures we have to move the hexapods down to compensate. We also found that we had set the software limits fairly far from the limit switches so we could give ourselves plenty more room without any changes to the hardware of the telescope. This is an example of the type of tuning that we continue to have to do with the control algorithms.
Both the LRS and the VIRUS are IFU (Integral Field Unit) spectrographs, which means that one could take the spectra for each fiber and reconstruct an image of the field which was observed. Below is the very first reconstruction that we did for a VIRUS field. It was a fairly bright (8th mag) standard star. The image scale is set very wide so that we can see all of the interesting features that come from the unusual design of the HET. The six image spikes comes from the edges of the hexagonal HET segments, and the stronger spikes along the X direction come from the tracker Y superstructure.
Not long after generating this the software team was able to reconstruct a number of images for each IFU in a pointing of a globular cluster, NGC5272. Below I show a couple of examples of these.
Here is a schematic of what the current array of IFUs would look like compared to a Digitized Sky Survey Image of the field we pointed at. This is a by-product of our current setup software.
We have had a very exciting couple of weeks. In the last blog post I had mentioned that we made some repairs and improvements to the top end of the telescope and that the LRS2 team had left the telescope in the hands of the night operations team to gain experience with the system. That is exactly what we did over the bright lunation. We had nights set aside with all of the telescope operators working together at night to make sure that we all know how to operate all of the new software the same way and to work out the kinks in the operations procedures. This was followed by a few nights with the resident astronomers all working at night to do the same. The result was that we had developed a few pieces of code that would allow us to determine the exact position of the LRS2 field of view and could setup any target that the Acquisition camera could see.
In the mean time, the day staff was prepping the right VIRUS enclosure for the delivery of the first 9 units (18 spectrographs). The spectrographs were brought out in two vans along with the VIRUS commissioning team for the dark run (where the moon is in the new phase). The fibers were strung, the units were vacuum pumped and finally cooled. Everything went ahead of schedule and on Saturday night we were able to make use of some very clear skies to get first light with the VIRUS spectrographs. On Sunday night we were able to get the first LRS2 science spectrum (a supernova for an astronomer at PSU) and a few long exposures for characterization of the VIRUS units.
The telescope is still occasionally fussy but generally performed well and the commissioning work continues on into the next week until quarter moon. At the present time the data reduction software is the thing that is lagging most behind.
We have had a great couple of weeks at the HET. Last week we brought out down the focal plane assembly to work on the field calibration unit and replace the acquisition camera. Both went well although some further work on a few specific calibration lamps and changes to the acquisition power supply will still be required. This work took the telescope down for a few days. The telescope control software also went through a large upgrade during this period. With improvements to guiding and offsetting. One of the more complicated aspects that has been commissioned is the ability of offset the telescope and the guide probes so that we can move from one instrument to another while keeping the same guide star.
The LRS2 commissioning team came back out to West Texas from Austin and despite very poor weather they were able to train a few of the resident astronomers in the use of the LRS2 instrument and the calibration scripts. When the skies did finally clear they were able to determine the positions of the LRS2-B and LRS2-R on the acquisition camera. The LRS2 team has departed but the instrument has been left online for the resident astronomers to gain familiarity and perhaps start characterizing on-sky.