Attending: Al, Peter, Tom, Roger, Jacques, Shaul, Qi
Notes by : Qi
Noise/band Figure
Proposed Caption:
“PICO polarization sensitivity (left ordinate and vertical bars), and expected RMS polarization signal intensity (right ordinate and lines) on angular scale of ~1 degree. For emission from Galactic synchrotron radiation and dust grain we use levels averaged over 60% of sky (as defined in 'Planck Intermediate Results. XXX.'). The BB amplitude is obtained with primordial B-modes assuming r=0.001 and lensing B-modes.”
Al Kogut's e-mail
Hi Josquin,
Are those units correct? uK-arcmin is a noise density (map noise
with pixels 1 arcmin in diam, hence pixel area 1 arcmin^2). Dividing
this by the beam FWHM in arcmin gives the noise density per beam
(uK/beam), which can then be transformed to units Jy/beam using the
usual conversion from thermodynamic temperature units to Jy.
That would suggest changing the Y-axis labels from Jy/sr to Jy/beam.
But that makes it harder to compare between channels, or between PICO
and other experiments, since everything on the plot would then be referred
to the specific PICO beam widths. I'd prefer leaving the units Jy/sr
and performing the conversion from arcmin pixelization to steradians.
This allows simpler interpretation for any choice of beam.
Al
Carlo's e-mail:
And another small check, does the synchrotron curve include curvature in the spectral index?
If yes ok, if not we could adopt the one which is used in sims, or add another shaded area with that, since it's rather baseline synchrotron modeling at the moment,
Carlo
Temperature Stability and effects on responsivity variations (Peter)
This was from 2007, not much progress was made due to lack of opportunities
things like better control system could help
+/-1 ohm sensitivity, which is +/-7 uK; so the data was at noise level, which prevented them doing better at the time
blue curves is the current in the magnet, exerted to stability, period about 20 mins
shorter time scale on temperature (red), when currents change direction
improvement
key question: given responsivity model, what would 20 uK change do?
Peter to send raw data of the plot (AI)
Roger volunteers to take charge of this, calculate equivelent signal from this fluctuation (AI)
Shaul: can affect responsivity, not just noise. If we were fluctuating a few mK, large signal far beyond we call noise.
Jacques: how 3K will be modulated; on ground, noise from all sorts of things are dominant, but in space, this source could be dominant in fluctuation
T_focalplane and T_ADR
time constant C/G: focalplane ~30 kg, C is not hard to estimate, but G may be. Ask Chris Payne to estimate (AI).
Peter: two different time constants, internal of arrays and between ADR and system. Internal time constant of array is also important.
Planck active T control
Jacque: Very long time constant, the fluctuation from dilution does not transfer to focal plane; most fluctuations were from cosmic rays
Shaul: why was active control still useful if time constant was large? Check with Brandon (AI)
The plot is from 4-stage, 4k to low T; 5-stage should outperform with 1K
Shaul: 1K stage would have its own fluctuation
Peter: ASTRO-H lab tests with full space instruments, 3 ADR: 2 upper stage + 1.5 K constant. Fluctuation was within 1mK. Two modes, ADR with liquid helium and without helium. Turned out without was more stable, gave better detector resolution, because of more stability under ADR; with helium, there were drifts.
Noise/band Figure
Jacques: this one is mostly wrong, no need to discuss. A lot things have been fixed.
Jacques sends a new figure for Shaul to present. Units are consistent between bands and signals.
Synchrotron not completely clear
BB dominated by lensing
Shaul: six dark blue bands are wrong, should be three
CO lines
Current units are for broader community, who do not understand uK.
Shaul will be gone for next 3 weeks until June 18th. After Shaul comes back, report writing mode.