Optimization and Lab Characterization of the TIME Instrument
Emma Linscomb
Nexus Scholars
Summer 2024
TIME Background
Tomographic Ionized-Carbon Intensity Mapping Experiment
2
3
Recombination → Dark Ages → Reionization
~ 370000 yrs after Big Bang
Cooled through expansion
Condense into neutral hydrogen
Uneven distribution of matter
Galaxies and stars begin to form
Ionizing radiation from new galaxies
Neutral hydrogen now ionized
What other sources of ionization?
ALMA Observatory: https://www.almaobservatory.org/en/160616-header-web-eng/
What We Are Building
4
Figure 2.43 from Jon Hunacek Thesis, Section 2.5
Figure 6.3 from Jon Hunacek Thesis, Chapter 6
Cryostat housing detectors and spectrometers.
Will be attached to the bottom of a telescope at Kitt Peak for observation!
What TIME Does
5
Examining the [CII] emission line allows us to better understand the phase shifts in the universe.
We don’t understand reionization very well! We could find other sources.
Using [CII] instead of ionized hydrogen allows us to observe a slightly different wavelength than other similar experiments.
Artist rendering of a quasar from NASA JPL https://images.nasa.gov/details/PIA16114
Hubble Image of a star cluster from NASA GSFC
https://images.nasa.gov/details/hubble-catches-stellar-exodus-in-action_17644845341_o
Relevance
Tracing matter as it forms all over the universe.
Could measure wavelengths across all space and time for different forms of matter.
This kind of cosmology allows us to learn more about our history on a cosmic scale!
6
TES Bolometers
7
8
Figure 2.1 rom Jon Hunacek Thesis, Section 2.1
spectrometer
detector array
Figure 2.11 form Jon Hunacek Thesis, Section 2.2
Detectors sit under the spectrometer to receive light split in spatial and spectral axes.
Figure 2.7 form Jon Hunacek Thesis, Section 2.2
9
Figure 2.12 rom Jon Hunacek Thesis, Section 2.2
The Cryostat
10
TES Bolometer Design
11
Film of conductive material that has transition temperature
Small change in temperature causes big change in resistance.
Figure 3.2 rom Jon Hunacek Thesis, Section 3.1
12
Load Curve Transitions
Load curves compare the current across the detector (TES current) to the input current (Bias current).
The transition on a load curve correlates with the decreasing section of the graph.
Transition
13
Detectors sit on top of a bath that allows for venting of power according to the thermal conductivity
TES Bolometer Methods
Detector
Bath
Photons
Voltage
Venting
TES Bolometer Design
14
Detectors connected in parallel with shunt resistor
Gives constant
15
Limitations
We can only bias on each column, NOT on an individual detector.
COLUMN
Figure 2.23 from Jon Hunacek Thesis, Section 2.3
optimal_bias.py
16
AKA the majority of my time spent!!!
What does it do?
17
Use the optimal bias script to find the best bias voltage value for each column.
Allows us to have the highest amount of usable detectors.
Likelihood
Bias Current
cutoff_finder()
18
Mitigate multiple transitions due aluminum or manufacturing defects.
2nd Transition
(BAD!!)
1st Transition (GOOD!!)
19
No more funky slope decrease!
col_transition()
20
Turn these…
col_transition()
21
Turn these…
Into this!!
But the transition didn’t look right?
22
yikes…
How to fix this?
23
KDE
Rug Plot
Peak Probability
1 SD
OR
KDE + Chi-Square
Interval Overlap Counter
Bias Current
Likelihood
To summarize:
24
1
cutoff_finder()
Remove secondary transitions
2
Find individual transitions
Each detector has an individual bias value!
3
col_transition()
Find the bias value with the most overlapping intervals
Next Project: Beam Mapper!
25
Use IR sources to produce fake data for testing between Kitt Peak visits!
Motor Controller
Function Generator
Beam Mapper Mount
Other Highlights
26
Polish Party!
Cryo Disassembly with 10-ton Crane
Tip Test!
Detector Installation (DO NOT TOUCH!)
Heat syncing
The Future
27
The Future for TIME
28
After it all works and lab tests are done, then we take it to Kitt Peak for ~3 years, 3 months per year for observation!!
Cosmology requires long-scale observation.
Future experiments could look at different wavelengths at different times to better understand those phase shifts.
Figure 2.42 from Jon Hunacek Thesis, Section 2.5
The Future for Me
29
What I learned:
Future goals:
Bonus: Fails
30
So. much. code.
When the distribution board stopped working for a month and then fixed itself T-T
Questions?
Thank you to Victoria Butler for providing the base code and for brainstorming solutions with me! (and for being responsive to my frantic Slack DMs)
Thank you to Abby Crites for being a wonderful PI and being so active in my learning process throughout my time in her lab.
Lastly, thank you to all the other amazing people I worked with this spring and summer!
31
Sources:
32
All images without captions are my own.
Appel, J. W., Bennett, C. L., Brewer, M. K., Bustos, R., Chan, M., Chuss, D. T., Cleary, J., Couto, J. D., Dahal, S., Datta, R., Denis, K., Eimer, J., Essinger-Hileman, T., Harrington, K., Iuliano, J., Li, Y., Marriage, T. A., Núñez, C., Osumi, K., … Xu 徐, Z. (2022). Calibration of transition-edge sensor (TES) bolometer arrays with application to class. The Astrophysical Journal Supplement Series, 262(2), 52. https://doi.org/10.3847/1538-4365/ac8cf2
Hunacek, J. R. (2020). TIME: A Millimeter-Wavelength Grating Spectrometer Array for [CII]/CO Intensity Mapping (thesis). https://thesis.library.caltech.edu/13683/1/thesis-jhunacek-20200427b.pdf
Kernel density estimation. Kernel density estimation - SciPy v1.15.0.dev Manual. (n.d.). https://scipy.github.io/devdocs/tutorial/stats/kernel_density_estimation.html
33