Rubin trip Nov 2022, Stubbs

Action items and observations:

Interference filter degradation!
can't ssh or scp to SLAC data from the summit, and they won't give us accounts on the summit. Not OK
Slack messages saying "dome might move" are now unclear as to which telescope is being discussed. Not OK.
EFD data aren't accessible in real time at SLAC data repo, and they won't give me a summit data access account.
No units in Choronagrf entries?
No real sense of US astronomers coming down- need an updated FAQ page, post-pandemic.
Takes about 1 minute to move 14 MB of data from SLAC to laptop on summit.
need to deal gracefully with dual telescopes in operation. Use a different office? Multi-zooms are going to be tough in commissioning.
We need to stress test the butler vs. number of accesses open to it.
prepare for using astrowidgets stuff for cursor feedback.
support notion of DM visitor at all times.
No way to log things like having moved anemometer.
What night-time support can be expected for Aux Tel operations? Who is on call for summit IT issues, in particular?
We aren't running project WFS code on Aux tel? Why?

For next time-

arrange for summit and up-down driving permission - Giovanni Corvetto is summit safety person.
tools and tool box- find or replace
Chromatic scattering test system with remote corner cube inside PVC tube?
combination of semrock quadnotch plus Edmund disperser, in disperser holder frame.
do thermal control experiments with Aux Tel dome
perhaps with big fan next to slit?
Take thermal imaging camera for iPhone.

Aux tel stuff

four changes to operations:

a. store telescope at horizon not zenith
b. open dome sooner, pointed East, with forced airflow
c. initiate celestial pole throughput monitoring
d. start sooner and end later, with sensible filter sequences.

we are ending sooner than needed, and starting too late. Last image on night of Nov 23-24 was taken at 0830 TAI, in r band. Sky background was (Rubin TV value) with 21 ADU median flux level in 30 seconds. First image was 64,000 ADU in 30 sec for CWFS so that seems better.
scheduler should start with zirgu and end with ugriz.
While just tracking same target, no slew, image 432, 433 had 5 arcsec pk-pk azimuth drive errors. Why? Image 412 (on same object but earlier in night) was middle of 3, same issue.
Map out az motor current vs. az at fixed rate?
point to the W, crack the slit, and open lower vents much earlier in the day, like around 5 or 6 pm, to get air exchange going sooner.
Think about clamp-on heat sink for cryotiger hose.
12 degree twilight is arbitrary- need to use sky brightness.
Add EFD check to afternoon checkout list.
Consider parking Aux Tel away from zenith. Need to verify this but it looks like there is 10 C of vertical temperature gradient in the dome.
Confer with AO team to predict FWHM after WF correction- are mirrors suffering from figure error? Need to take donuts AFTER corrections are done as well.
Add nightly imaging of celestial pole, with and without filters in the beam. This will allow us to monitor filter degradation.
Do seeing test vs. thermal turbulence, also at pole

Elana:

Elana should bring down three prong Chile to US power adapters. Ordered.
Elana needs to do inventory of what we have in boxes, in container, with video.
also see equipment (including Mac Mini) in fan box at base of telescope pier.
SSH connections into temp sensor system.
1. do we need to take down keyboard and monitor?
think about where to put our anemometer. Really useful diagnostic.
Think about monitoring systems for main telescope!
Early dome opening with fan installed? Need to know about all temp sensors in the enclosure.
stuff in fan box at base of pier in Aux tel, including mac mini from before for DIMSUM?

Nov 22, 2022

Aux Tel banding:

Tried a bunch of things, none of them made a difference. from auxtel-detector-effects Slack channel:

Image quality instrumentation:

Moved anemometer onto Nasmyth fork.

Use chronograf at https://summit-lsp.lsst.codes/chronograf/sources/1/status

Explore

efd.autogen

Anemometer updates at 10 Hz. Reported quantities are speed0-speed2 in m/s and acoustic temperature in Deg C. Other items in Choronagraf are computed from these, but Craig says beware of computed magnitude.

Anemometer axis conventinos

So...

positive x direction points towards primary along optical axis
positive y points away from telescope, horizontal always
positive z makes Right Handed coordinate frame with the other two, such that z=cross(x,y)

It seems rational to assume, based on Campbell scientific documentation, that

speed 0 = x
speed1 = y
speed2 = z.

So if we point to horizon, wind through slit shows up in x
If we point to zenith, wind through slit is along z

If HSW is horizontal slit wind, then x component (up to a sign) is HSW*sin(zenith angle), and z component is HSW*cos(zenith angle).

We likely want to compute wind speed as slit-wind and cross-wind. The windspeed in the plane of the telescope is sqrt(ux^2 + uz^2). Call that slit-wind. Out-of-plane it's uy.
That avoids having to know elevation angle.

Nov 24, 2022

Modified differential temperature installation to place dT_x along the anemometer support pole, so it's measuring difference along the optical axis.

Attempts to log into data collection computer failed, so wasn't able to verify things are working. DVM tests looked promising.
Also the interior of the data collection box is not long term reliable. boards are rattling around inside the white box and will fall apart if moved. We need a second generation of this.

For Nov 23:

driving test
measure rails and take images
move temperature system.

Bad Sats

Nov 23.

try subtracting pairs of Aux Tel imaging survey images to look for Dan Weatherill's 'bar codes'
Images to use from night of Nov 22 are

714, 715, 716

724, 725, 726

Elana's December trip to Chile

December 2

Differential temperature sensor

In morning, restarted differential temperature sensor probe computer a few times before it would finally show me the login screen. Once it showed me the login screen I could ssh in.

Started a 24 hr run on the differential temperature sensor where it's saving 10% of the detailed data (per Sasha's notes). At 2:39pm local time, cracked open the dome slit by ~1 foot (don't want to open it further because don't want to have the sun fall on the telescope). Door to AuxTel is also open. Slit closed at 3:39pm due to concerns about the sun hitting the telescope.

Effect of opening dome on acoustic temperature:

standard deviation of sonic temperature, different x axis:

and here is corresponding plot for our X-axis differential temperature variance:

Conclusion:

We can easily tell when dome slit is open, with x-axis differential temperature sensors. Note our dT variance is computed over a wider frequency range than the sonic temperature. Goes from much higher frequency to much lower frequency.

This shell script takes the fragmented TempVar data files for a single day and makes a merged file with delta-t and temperature variance as two columns:

#!/bin/bash
# wipe out any old files
rm timestamps.dat
rm timestamps.tmp
rm timestamps.tmp2
rm VarPoints.dat
rm VarPoints.tmp
rm TimeVar.tmp
rm TimeVar.tmp2

# Loop over files we want
for tempfile in TempVar*.csv;
do

# extract only the timestamps
echo $tempfile
cat $tempfile | grep Delta | awk '{print $3}' >> timestamps.tmp

# get rid of commas
cat timestamps.tmp | sed s/','/''/ > timestamps.tmp2

#extract second variance numbmer over the interval in each TempVar file
cat $tempfile | head -3 | tail -1 >> VarPoints.tmp

done # end of loop through files

# construct seconds elapsed
awk 'BEGIN { FS = ":" } ; { print $1*3600+$2*60+$3 }' timestamps.tmp2 >> timestamps.dat

# extract second data point for each minute
awk 'BEGIN { FS = "," } ; { print $3 }' VarPoints.tmp >> VarPoints.dat

# now merge the two columns into one file
paste timestamps.dat VarPoints.dat > TimeVar.tmp

# sort by timestamp
sort -k 1 -n TimeVar.tmp > TimeVar.tmp2

# fix up the horizontal spacing
awk '{print $1" "$2 }' TimeVar.tmp2 > TimeVar.dat

# check on lengths
wc timestamps.dat
wc VarPoints.dat
wc TimeVar.dat

DIMM

Had lunch with the people from Astelco who are here installing a new mount for the DIMM. They said this DIMM dome is unusual - usually they open all the way, so dome seeing is expected here. They also said that they think the fans are unlikely to do much and while spending time on the DIMM tower there was often wind from below, through the chimney (with the dome both open and closed). This is bad.

Main dome visit

Mario gave a tour of the main dome. I took lots of pictures and think there are quite a few places we can put dome seeing instrumentation. I couldn't figure out where on the dome the CBP is going, but Brian suggested that we attach the solar cell to the yellow railing on level 8. I've taken several pictures of the yellow railing. Note also in the images attached that the DIMM is currently placed on the side of the TMA. We could mount other dome-seeing instrumentation in similar locations on the side of the TMA. There are also quite a few places on the dome near the dome slit where we could place instruments. There is also a catwalk around parts of the TMA that I think will remain. I can take more pictures and possibly measurements next week. Brian also suggested I look at the Tokovinin system he's planning to put in the main dome. It's in his office and I will look next week.

SHWFS

Mac mini is still working and we can ssh into it. Found cables to connect the Sony camera to the Mac mini. Attempting to charge battery now (as long as it can hold some charge it should be okay, because it can also charge through one of the cables).

Stubbs group Anemometer

Found replacement power supply. Need to decide where and how we want to install it.

Storage from last time

All items we left here have been found. We have many extension cords, U.S. to Chile adapters, tools, cables, etc. Will do a full inventory next week.

Note: the rest of Elana's trip is logged at Dec 2022 Chile Trip, Urbach