Karl Franz Flasch Hendrickson

Truman State University
Kirksville, Missouri 63501
REU 2003
University of Wisconsin-Madison
Madison, Wisconsin 53706

Final PowerPoint Presentation
I ponder the setting sun...

Development of Bolometric Detectors for Cosmic Microwave Background Study

What is the Cosmic Microwave Background?
What is a Bolometer?
The Research
Future Goals
Lab pictures!
Links

What is the Cosmic Microwave Background?


Simply put, the CMB is a collection of photons at a (roughly) constant temperature of 2.7K received from all directions in the universe. To understand from whence these photons come, we must look at the initial moments of time, as we know it now.

Shortly after the Big Bang, the Universe was a hot dense plasma ball. It was so hot that simple hydrogen could not form. During this time, photons were constantly being scattered (Thomson scattering) by the charged particles (imagine yourself trying to run through an extremely dense forest... you won't get far before you hit a tree, and then another tree, and so on...). Eventually the Universe expanded and cooled to the point where hydrogen could form (by cool, we mean something like 3000-5000 Kelvin). Effectively, all photon scattering stopped, and the photons continued in straight lines through the universe until they reach us now, ten to fifteen billion years later. Thus, the 'image' we see from these photons is called the Surface of Last Scattering. The photons are now 2.7K because their wavelengths have lengthened (redshifted) as the universe has expanded.

Figure 1: The Universe underwent a period of extreme inflation, until an age of roughly three hundred thousand years, which is the time of the surface of last scattering.
Figure 1


For more on the CMB, see Wayne Hu's Homepage.
Back to the top

What is a Bolometer?

A bolometer is a form of detector used to study the CMB, as well as other forms of electromagnetic radiation.

Figure 2: A sketch of a bolometer.
Figure 2


For simple understanding, a bolometer is a piece of detecting material held at a very cold temperature by a weak thermal link to a cold reservoir (See Figure 2). When incident photons strike the detector material, the temperature will increase. We can learn about these incident photons (some of which are the CMB - so you need to filter a lot out!) by studying the change in temperature. In order to study the photons, however, you need a way to measure changes in the detector.

TES Bolometers


Our Bolometers are called Transition Edge Sensor bolometers. The detector is constructed of a piece of superconducting material biased in the middle of its superconducting transition.

Figure 3: When a material transitions to its superconducting state, the resistance falls off sharply to zero.
Figure 3

By holding the detector near the middle of the transition curve (See Figure 3), very small changes in temperature will create very large changes in resistance. Thus, if we can measure the resistance, we have a very sensitive detector to study the incoming photons. Learning how to measure this resistance was one of my projects for the summer.

For more on Bolometers, see Berkeley's Introduction to Bolometers.

Back to the top

Research: Summer 2003


A)Magnetic Sheilding

The first complication with using TES Bolometers comes from external magnetic fields. In the presence of a magnetic field, a superconductor will begin to circulate current within its interior. If the magentic field is too strong, too much current will be produced, and the superconductor will reach its critical current - and thus cease superconducting. This ruins the advantage of using the superconductor as the bolometer detector. Thus, the bolometers must be shielded from stray magnetic fields.

I studied the abilities of superconductors to act as magnetic shields. An important property of superconductors is known as the Meissner Effect. It says that "a metal in the superconducting state never allows a magnetic flux density to exist in its interior" (Rose-Innes. Introduction to Superconductivity p.19). This is actually the property that creates the currents inside a superconductor in the presence of a magnetic field.

To use this property as a shield, I put a layer of lead tape (Tc = 7K) on the top and bottom of a metal box that would be used to hold an array of bolometers. When the lead goes superconducting, current will be created inside the lead to cancel any magnetic fields within itself. If any external magnetic fields are then introduced, the lead will produce current to cancel these as well, creating a magnetically stable area between the to layers of lead.

To test this, I placed a magnetic field sensor inside the box and placed the box inside a dewar to cool to 4 Kelvin.

A metal box lined with lead tape mounted on the 4K plate of a dewar.


I then placed the cooled dewar between a pair of Helmholz Coils, which create a near-uniform magnetic field along their shared axis.

The Dewar is sandwiched between two Helmholz Coils, underneath a bucket used to support the coils.


By varying the voltage across the coils, varying strengths of magnetic fields were introduced to the box inside the dewar. The response of the magnetic field sensor (Hall Probe) was then recorded. The data are as follows:

Figure 4: Three sets of data are shown. For the data with T=4K, the change in voltage across the Hall Probe (proportional to the magnetic field) was an order of magnitude smaller than for the others.
Figure 4


The voltage readout on the Hall Probe (y-axis Figure 4) is inversely proportional to the magnetic field strength (decreasing voltage => stronger field). The data show that at superconducting temperatures, the magnetic field increase detected was an order of magnitude smaller than at temperatures where the Lead tape was not superconducting. It should be noted that the varying levels for the magnetic field when the power supply voltage was zero is most likely due to the Hall Probe itself, which has a tendency to fluctuate its reading, even when left undisturbed. The change over the increase in external field is what is most important. When the power supply votage was 20V, the corresponding field produced by the coils is approximately 2.3 mT, which is significantly greater than Earth's roughly 0.04 mT field (location specific, or course).

B) Bolometer Readout with SQUIDs

In order to record the changes in the bolometers, we need to measure the changes in resistance across them. To do this, we use SQUIDs. Superconducting QUantum Interfered Devices, or SQUIDs, are devices that can be used in many ways, butfor our specific purposes, they are the most sensitive, least noise ammeters possible.

At its most basic level, a SQUID is simply a ring of superconducting material (See Figure 5). While it is possible for a loop of superconducting material to have magnetic flux through the loop, the amount of flux will stay constant. Any attempt to introduce more flux through the loop (by, say, running a current through the inductor in Figure 5) will create a current in the loop to counteract the introduction of flux. This changes the voltage across the SQUID.

Figure 5: A simple schematic diagram of a SQUID.
Figure 5


In order to use the circuit to read a bolometer, we construct a circuit like that seen in Figure 6.

Figure 6: Circuit diagram for reading bolometers. A DC Voltage is connected to three parallel circuits: one with a capacitor, one with a shunt resistor, and one with a variable resistor (the bolometer) in series via twisted superconduting wires to an inductor in a SQUID circuit.
Figure 6


The shunt resistor Rs makes the circuit essentially a voltage divider, and the capacitor acts as a low pass filter. The variable resistor Rb is the bolometer. With Rs << Rb, the voltage through the SQUID loop of the circuit is mostly constant. For our SQUID, the voltage V across the SQUID is proportional to the current I through the inductor. Knowing the current through bolometer loop of the circuit allows for a calculation of the variable resistance value.

For my testing, instead of using a bolometer as a variable resistor in the circuit, I simply put in a fixed resistor with a known resistance to check to see whether I would be able to calculate the resistance based on the SQUID voltage. I used a shunt resistor that was 6.2±0.6[10%] mOhm and a fixed resistor of 24±2[10%] mOhm. After significant amounts of struggle with the SQUID (they can be very finnicky!), I obtained three good sets of test data. They produced the following results:

TrialResult
A21±2[10%]mOhm
B24±2[10%]mOhm
C*24±2[10%]mOhm

These data agree with the theoretical value (24±2[10%]mOhm), which is a major step towards successful use of the SQUID to read bolometers. The third result is noted with an asterisk because this value was obtained using a second SQUID, for which a proportionality constant used in the calculations was unknown. In place of the specific constant, the value from the first SQUID was used in its place. The sQUIDs should be similar enough that the results should be comparable.

These data were obtained by putting the circuit on a dunk tube and immersing it directly into liquid helium. Efforts have also been made to use the SQUID inside the dewar, but, at this point have not produced positive results. The SQUID circuit has been cooled in the dewar and has worked with the SQUID controller; however, the data obtained from the circuit in the dewar did not produce satisfactory results. This could be due to a number of reasons, including bad circuit connections and the possible need to assure a common ground among all circuits.

For more on superconductivity and SQUIDs see:
Rose-Innes, A.C., Rhoderick, E.H. Introduction to Superconductivity 2nd Ed.
New York: Pergamon Press, 1978.

Future Goals


It is hoped that SQUIDs will be the key to bolometric detectors for studying the CMB. If the bolometer circuit is biased using an AC current, then it is possible to read out the specific bolometer voltage from the SQUID using a lock-in amplifier. Hopefully, then, multiple bolometers - perhaps as many as a few hundred - will be able to be read using a single SQUID by biasing each bolometer at a different frequency and reading the SQUID voltage using, again, a lock-in amplifier.

Anisotropies in the CMB have already been discovered and studied by such instruments as COBE and WMAP. The goal of future CMB studies is to analyze the polarization of the CMB with the hopes of learning more about the first few moments (relatively speaking) of the universe. Any anisotropies in the polarization due to tensor mode oscillations in the early hot plasma will be one to three orders (perhaps more) orders of magnitude smaller than the anisotropies already discovered. As a result, detectors need to be refined and developed to be extremely sensitive to these anisotropies. This is the motivation for the development of bolometers and the use of SQUIDs (high precision) for the study of the CMB.

Back to the top

Acknowledgements


Special thanks to:
  • Dr. Peter Timbie, Advisor
  • Dr. Shafinaz Ali
  • The rest of UW ObCos
  • The University of Wisconsin-Madison
  • The National Science Foundation
  • My fellow REU Students
    Back to the top

    Pictures!
    My dewar smokes as it fills with Liquid Nitrogen.

    A liquid helium fill.

    The inside of the dewar - showing the liquid nitrogen and liquid helium tanks.

    Back to the top

    Links

    University of Wisconsin Observational Cosmology
    Truman State University Physics
    Truman State University Society of Physics Students
    The American Association of Physics Teachers
    Physics 2000: Textbook on CD

    The Vatican
    Diocese of Jefferson City
    Kirksville Catholic Newman Center

    Back to the top

    E-mail Karl
    Last Updated: 8-Aug-03 Feast of St. Dominic