Attending: Al, Roger, Jacques, Jeff F., Shaul, Karl, Qi, Ranajoy, Brian

Agenda:

• Imager + Spectrometer (Karl) Spectrometer Slides
• Focal plane + noise model (Karl) Noise model slides
• optical system (Qi) crossdragone_d100cm_f4.5_fov10deg_10032017.pdf
• Scan parameters (Shaul (for Kris))

Notes:

Imager + Spectrometer (Karl)

• mini-PIXIE, 20 cm mirrors, 10x38x48 FTS
  • Same boresight angle
  • Share same cooling system
• mass of mini-PIXIE, ~20 kg in total, barrel contributes most of it.
  • detector assembly is < 1kg
• Imager focal plane mass: not known yet. A/I Brian to ping Roger.
• with Open Dragone
  • same boresight as Imager, although not necessary
  • quite small, fits easily
• with Cross Dragone
  • boresights don't match, but scan pattern does (beta is same in both cases)
  • Al: it’s okay. There will be some time delay between imager and spectrometer scans, but it does not matter much.
• Boresight and spin axis
  • It’s not required to have boresight along the spin axis. PIXIE did plan to align with spin axis,allows fitting of Q/U of each pixel before moving to next pixel.
    • It’s okay to not do this. The current main driver is the sky coverage.
• FTS box is not required to be with optical tubes, it can be anywhere as long as an optical path exists.
  • The detectors can even be on the Imager focal plane.
• Tradeoffs:
  • question of cost, might limit imager size
    • Current primary costs: cooling, telescope, readout, and detectors
    • marginal costs when increasing
• Aside:
  • Check Open Dragone fit, to see if there is conflict between mirrors and cone. A/I UMN to check.
  • structural issues with support for mirrors?
    • Brian: JPL is working on this question.

Noise Model (Karl)

• Biggest change: safety factor → 2
• Squid is the largest contribution
  • Using 3.5 pA/rt(Hz). A/I Karl checking with Hannes at NIST about proper number to quote.
• Roger: all these numbers are fine.
• Cross check between Karl and Roger. noise matches except for:
  • bunching noise, a factor of 2
  • bunching noise does not contribute too much, ~15%; all other are same
  • A/I Karl, track down 2 in bunching noise. Possibly polarization related.
• Richard’s correction on Coma
  • more detectors at high frequencies, for both Open and Cross Dragone
  • can weight CMB frequencies more heavily.
• Can optimize edge taper
  • including correlated photon bunching noise from CMB (correctly) and mirrors (incorrectly, but with factor of 2)
    • Next step: Karl to implement full correct calculation for mirrors.
  • best mapping speed is near 2 F lambda. Current Open Dragone pixels are at 2 F lambda.

Optics (Qi)

• New cross-dragone with +/- 10 deg FOV
  • f/4.5, 1m aperture, requires deployable shields.
  • deployable shields should be V-grooves as well. leave design to thermal engineers. Just put place-holder in current model
• both cross dragones' focal planes view sky after 1 reflection off tertiary
• total sensitivity is better than Open dragone (baseline)
  • assumes full telescope is 4 K
  • also number of detectors is larger. 3k for Open, 7k and 11k for new crossed designs.
• How does larger f-number give better sensitivity? Opposite what is normally said. From CORE, increasing f-number gave no benefit since pixels grew at same rate focal plane grew.
  • A/I Karl/Qi to sort out scaling of focal plane with f-number.

Scan delayed to next week.