Previous%20lecture:%20Fundamentals%20of%20radio%20astronomy

1
Previous lectureFundamentals of radio astronomy

• Flux, brightness temperature...
• Antennae, surface accuracy, antenna
  temperature...
• Signal noise.
• Detecting a weak signal.
• Some general considerations.

2
Blazar observing techniques
Word of warning It is possible that your
friendly support staff has no experience
whatsoever from blazar observations!
Special considerations continuum sources of
unknown fluxes (variable!) faint
sources sources that may not be suitable for
pointing (faint) absolute calibration needs to
be accurate exercise special caution if looking
for IDV!!!!!!
3
Receivers, problem areas

• Noise from the recei ver, gain fluctuations etc.
  large factor compared to the astronomical
  signal.
• Signal-to-noise ratio ? radio astronomical
  measurements arealso measuring noise!
• Signal ltlt background.
• Power levels are low (P ca. 10-15 - 10-20 W).
• Noise also from ground, atmosphere, etc.

4
We need

• Good stability.
• Good sensitivity.
• Good measurement techniques (elimination of
  background etc.).

5
Background noise

• Man-made noise
• Receiver itself, cables other components,
  other radio signals (GSM!), etc.
• Noise from the nature
• Atmosphere, CMB, black body radiation from ground
  etc., Solar radiation, thunderstorms ...

6
Receivers

• Heterodyne receivers
• Defining technologies HEMT, Schottky, SIS.
• Bolometers.
• Bolometer arrays.
• Especially in the receiver technology
  considerable differences btw. microwave /
  millimetre / submm!

7
Heterodyne receiver Superheterodyne or super
receiver, superhet

• Uses a local oscillator at a freqeuency that is
  different from the incoming signal, to obtain an
  itermediate frequency that isprocessed using
  conventional microwave technology.
• Different technologies at different frequency
  ranges,HEMT amplifiers vs. SIS mixers,currently
  the limiting freqeuency ca. 100 GHz.

telescope
sideband filter
Dicke-switch
mixer
pre-amplifier
IF amplifier
detector
comparison load
localoscillator
8
Preamplifier

• At lower freqeuencies, f lt 120 GHz(note upper
  freq-limit changing with new technologies!)
• At higher frequencies not in use (too noisy, or
  nonexistent).
• Common abbreviations
• LNA Low Noise Amplifier.
• HEMT High Electron Mobility Transistor.

9
Sideband filtering

• Not desirable/useful in continuum observations!
• Normally two sidebands, upper sideband (USB) and
  lower sideband (LSB), get through the mixer.
• Used for sideband rejection, before the mixer,to
  filter one of the sidebands --gt single sideband,
  SSB.
• For continuum observations, wide bandwidth is
  desired and both sidebands carry information
  double sideband, DSB.
• If sideband rejection is normally in use, check
  if you canget rid of it for continuum work!

10
Mixer

• The signal gets converted (mixed) to a lower
  frequency.
• Amplifying high radio frequencies (mm/submm) is
  difficult, amplifiying low-freq signals is easy.
• (Especially earlier) no preamplifiers for high f
  exist(ed), only amplifying after mixing
  possible.
• Signal frequency fs, local oscillator frequency
  fLO ? lower intermediate frequency fIF fs
  fLO fIF much lower, much easier to handle (
  amplify).

11
Mixer technologies

• Schottky diode
• Older technology, easy to build.
• Typically for lower frequencies ambient
  temperatures.
• SIS junction mixer (Superconductor-Insulator
  Supercond.)
• Requires cooling, more challenging to build.
• For frequencies gt 100 GHz (at lower freqs no
  direct benefit requires more maintenance),widel
  y in use at mm/submm telescopes especially for
  spectral line work.
• Useful for detecting faint signals, ie. when
  high sensitivity is needed.
• In practise often problems with stability!
• Bolometers much preferred for continuum work!

12
Local oscillator

• Typically a semiconductor oscillator that can be
  phaselocked to an exact frequency.
• Most common Gunn oscillator.Difficult to make
  for n gt 120 GHz multipliers.Gunn oscillator
  itself not very stable phase locking to a more
  stable oscillator.Phase lock loop (PLL) keeps
  the Gunn LO stable (n phase).
• Output signal intermediate frequency, IF.
• Stability required especially for spectral line
  observations and VLBI. At high frequencies
  stability may be a problem.

13
Dicke-switching

• Gain fluctuations from mixers.
• Varying background noise.
• Observed signal e.g. 1/10000 of the overall
  system noise.
• Amplifications by a factor of ca. 1012 may be
  desired.? comparison measurement by
  Dicke-switching (R.H. Dicke 1946).

14
... Dicke switching

• Swithching rate e.g. 5 100 Hz.
• Reference source noise diode attenuator
  signal from the blank sky(e.g., two-feedhorn
  method in Metsähovi).
• Two beam-method for beam switching two
  feedhorns or one beam chopper wheel.

15
Millimetre-domainspecial considerations

• Components are small.
• Tolerances are small.
• Components are often expensive.
• Circuit losses larger than in the microwave
  region.
• Amplifier technology still in development.
• Less power.
• Atmospheric attenuation is large.
• Millimetre domain technology
• Quasioptics (Mirrors, lenses diffraction must
  be taken into account!)
• Cooling is necesssary.

16
Quasioptics

• In the (sub)mm domain the interface from the
  telescope to the detector.
• Mirrors, lenses, grids are small ? geometrical
  optics can not be used (diffraction) ?Gaussian
  optics quasioptics optimizes the feeds of the
  receiver to the antenna beam pattern.
• Can include a polarizer plate.

17
Cryostat

• SIS requires very low temperatures.
• Helium cooling.
• Cooled parts of the receiver in vacuum container,
  dewar.
• Dewar enclosed in a radiation shield.
• Radiation shield cooled by a closed-cycle
  refrigerator.

18
Bolometer

• A total power detector.
• High sensitivity from a cooled device.
• Wide bandwidth.
• Relatively easy to construct operate.

Classical semiconductor bolometers Superconducting
TES devices Silicon pop-up detectors (PUDs)
19
... Bolometer

• Classical Germanium bolometer.
• The temperature rise causes a change in the
  resistance of the bolometer and consequently in
  the voltage across it. V is amplified and
  measured.

20
Bolometer characteristics

• Noise Equivalent Power, NEP The power absorbed
  that prduces a S/N of unity at the bolometer
  output.NEP2 NEP2(detector) NEP2(background)
• Thermal time constant t a measure of the
  response time of the bolometer to incoming
  radiation C/G
• Compromise btw response time and NEP!
• In practise, with wide bandwith the bolometer
  performance can degrade due to power loading from
  the background ? background loading
  photon noise.

21
Bolometer arrays

• Until mid-1990s bolometers were mainly
  single-channel devices.
• Main advantage mapping of extended sources,
  i.e. not directly applicable for blazar studies.
• Can eliminate the need for separate pointing
  observations (depending on the obs.
  configuration).

22
Bolometer arrays currently in use

• SCUBA _at_JCMT
• 450 / 850 mm, 91/37 pixels, 300/65 mJy/sqrt(Hz).
• MAMBO-2 _at_IRAM
• 350 mm, 384 pixels, 500 mJy/sqrt(Hz).
• Many hundreds or even thousands of pixels.
• Silicon-micro machining, thin-film deposition and
  hybridization techniques.
• Integrated SQUID multiplexers in the same plane
  as the detector chip.
• E.g., SCUBA-2 10000 pixels, for JCMT.SPIRE for
  HERSCHEL 200-700 microns, 3 arrays of 43, 88 and
  139 pixels.

Future
23
Observing techniques beam switching

• Source size, line/continuum, etc --gt Observing
  method (Position switching, Frequency
  swithching, Load Switching)
• Continuum observations of point sources beam
  switching.
• Single beam switching The source is in the
  signal beam, the sky is observed in the reference
  beam.
• Dual beam switching Alternates the source/sky in
  the signal and reference beams.
• Dual beam switching produces good results when
  sky noise is the problem most of the time ?
• Note ON/ON, ON/OFF terminology not fully
  standard!
• Technology dual horn setup or chopper nodding
  secondary telescope movement.

24
... beam switching, things to remember

• Relatively large overheads do not use too short
  integration times per beam.
• Data reduction make sure to know whether your
  intensities are to be divided by 2 or not!

25
Dual beam switching

• 2 feedhorns / beams, A B
• Dicke-swithcing measures signal A signal B
  (integrated in e.g. 1 s chunks).

I A background (b) B background source
(b s) II A background source (b s) B
background (b)
(s b)B - bA
D 2S
bB (s b)A
Not the same background! Eliminates e.g. effects
of a radome very well!
26
Point source observations...

• 1-point observation, problemsOne must rely on
  the pointing pattern offsets required earlier.
• Drift scansThe moving source drifts over the
  beam.Gaussian fit ? Smax.
• 5-point observationFor relatively bright
  sources!Default center point 4 other
  positions.2-dimensional Gaussian fit amplitude
  offsets.

27
Pointing

• Radio telescope pointing can never be quite
  perfecttelescope size, gravitaional
  deformation, heating(more exotic problems
  pedestial tilts, earth quakes, ...)
• Pointing pattern.
• Pointing checks.
• 5-point (9-point) measurement offsetsamplitude
  correction (2-dimensional gaussian fit).

28
Focusing

• Antenna deformation may be caused by gravitaional
  or wind forces, or by differential thermal
  expansion.
• ? illumination pattern is controlled by
  changingthe position/shape of the secondary
  mirror
• Very important in the submm region!

29
Calibration

• From the observed DV or DT into Sobserved.
• Load calibration (noise diode, chopper wheel).
• Opacity (tau) measurements (skydips).
• Flux calibration.
• (Pointing check sometimes called pointing
  calibration).

30
... calibration

• cm-frequencies noise diode injection for
  receiver calibration Gain fluctuations
  (temperature, electronics, mechanical stress).
• mm-frequencies absorbing blakcbody load (blocks
  the sky emission, corrects for atmospheric
  attenuation), can be done frequently? corrected
  antenna temperature TA
• Skydips
• Take up observing time.
• Assume homogeneous, plane-parallel atmosphere.
• Corrections usually done at data reduction stage.
• Primary flux calibrator a bright source with a
  known, constant flux.
• Secondary calibratora bright relatively
  well-known, relatively constant flux.

31
... calibration

• Microwave domain primary DR21, Jupiter,
  Marssecondary e.g., 3C274, 3C84.(Baars
  scale).
• Submillimetre domain primary Mars,
  Uranussecondary planetary nebulae, giant
  stars, ultracompact HII...
• Consult the support staff for advise on the
  observatory calibration procedures suitable
  calibrators their flux information!
• Always make sure to make frequent flux
  calibrator observations!!!

32
... calibration

• Tau measurements / skydips
• Measure the noise temperature of the sky at
  various elevations.
• Effect on the flux calibration exp(tau/sin(el)).

33
Telescope performance, sky noise

• Performance is represented by the Noise
  Equivalent Flux Density (NEFD).
• Depends very much on the weather and varies with
  sky transmission.
• Often the fundamental sensitivity limit is set by
  the sky noise.
• Sky noise ? spatial and temporal variations in
  the emissivity of the atmosphere.
• Sky noise can degrade the NEFD by more than an
  order of magnitude.
• Chopped beams travel through slightly different
  atmospheric paths ? Use narrow beam switching!
  Still some residual.
• With a bolometer array one can try to remove
  large-scale effects to high order.

34
From an observation into a data point

• Observing strategy
• Is focusing needed?
• Is load calibration needed?
• Is skydip needed?
• Choose target source. Bright enough for a
  pointing source?No check pointing or use
  reliable offsets in the same direction.
• Observe source, keep an eye on S/N if
  possibleif not, make a quick-look analysis to
  check if integration time was optimal.
• Observe calibrator source.Avoid unnesessary
  slewing, avoid blind pointing, make sure to keep
  an eye on the weather changes, observe primary
  and secondary calibrators often enough!

35
... From an observation into a data point

• Data reduction
• Enough flux calibration data?
• Enough tau information?
• Was the weather OK all through your run?
• Bolometer array data reduce the image following
  local instructions.
• Single channel bolometer or heterodyne data
  obtain the average result of the various ON/ON
  pairs with their cumulative error.
• Make tau corrections any other general
  corrections.
• Obtain flux conversion factors from the
  calibrators.Are they consistent???
• Use the FCFs for your own sources.
• Error estimates from rms values deviations in
  FCFs.

36
Bolometer data reduction

• E.g. SCUBA photometryFlatfielding Extinction
  correction Despiking Examine individual
  bolometer quality change if neededSky
  removal Average or parabola fit over the
  9-point mapCatenation of individual measurements

37
How to write a good proposal
or OK you think youve got a good science
case...but THEY dont know it!

• Why this telescope?
• Why this/these observing band(s)?
• Read the manuals observing manuals, data
  reduction,show them that you understand it all.
• Any specific constraints? Sun, gain/elevation,wea
  ther, pointing, ...
• Work out the realistic integration times
  including realistic overheads, explain them all.
• If using heterodyne receiver, pay attention
  totuningsidebandsbackenddata reduction

38
Observational radio astronomy

• N. Bartel et al. 1987, ApJ 323. 507No data
  were taken at station D during the period 0830 to
  1630 GST due to the presence of a red racer snake
  (Coluber constrictor) draped across the
  high-tension wires (33,000 V) serving the
  station. However, even though this snake, or
  rather a three-foot section of its remains, was
  caught in the act of causing an arc between the
  transmission lines, we do not consider it
  responsible for the loss of data. Rather we blame
  the incompetence of a red tailed hawk (Buteo
  borealis) who had apparently built a defective
  nest that fell off the top of the nearby
  transmission tower, casting her nestlings to the
  ground, along with their entire food reserve
  consisting of a pack rat, a kangaroo rat, and
  several snakes, with the exception of the
  above-mentioned snake who had a somewhat higher
  destiny. No comparable loss of data occurred at
  the other antenna sites.