Telecon 20170911

Attending: Al, Shaul, Amy, Brad, Julian, Brian, Roger, Bill, Jacques, Jeff

Agenda:

• New Permanent Telecon Time: Tuesday 2 pm
• White noise and focal plane model Noise model status (.pdf) (Karl)
  • Focal plane size comparison fov_comparison_09112017.pdf (.pdf) (Qi)
• White noise comparison; 1/f noise (Roger, Jeff)
  • TDM noise :TDM noise Study (.pdf)
  • 1/f noise summary : 1/f noise Study (.pdf)

Notes:

New telecon time. Tuesdays 2 pm CDT starting 9/19.

Noise model (Karl)

• Added multichroic pixels
  • stop illumination changes with frequency
  • Comments:
    • A/I should set lowest band at 25 dB to be conservative.
    • illuminations, diffraction, spillover may need study with GRASP.
      • Amy can look for person or software access at JPL. A/I: pending discussion with Shaul.
• 100 mK bath temperature added.
  • effects low frequencies where phonon noise was dominate at 100 mK bath.
• Single detector NET, Open performs better, particularly at high frequency.
• DLFOV calculated.
  • in F*lambda units, crossed diameter is 2x open diameter.
  • pixels numbers added which satisfy DLFOV requirements. Hex packing assumed
• Final polarization weights are similar
  • crossed is 0-20% better at most bands.
  • open 5% better at only 3 bands.
• Discussion:
  • Since roughly equal sensitivity open is probably preferred as focal plane is smaller, fewer detectors, less massive, easier to cool, overall cheaper. Also lower sidelobes.
  • may need to consider diffraction pattern on primary of open. Does primary need to be oversize? How much? A GRASP calculation could be done.

Focal plane sizes (Qi)

• Past report said DLFOV for Open/Crossed ~ 75%. This was stop at primary.
• DLFOV reduced for stop between primary and secondary on open dragone.
• Key points:
  • stop location matters and reduces DLFOV.
  • results consistent with and explain focal plane sizes in noise presentation.

1/f noise and TDM (Roger)

• presentation tied up in JPL release procedure. Should be shared with 1 week.
• brief summary of work done:
  • TDM version of white noise calc in rough agreement with UMN numbers.
  • Likely to need 100 mK bath temperature just so bolometers are build-able. Need a very low G and low T_bath helps achieve this.
  • little to no 1/f input from readout developers yet.
  • total instrument 1/f compiled from various deployed sources
  • how low has been achieved? Unclear, most experiments use modulators or focus on higher l science.
  • Comments:
    • SPIDER has a non-scanning data set which can be examined for 1/f levels. A/I Roger to follow-up.
    • Niemack may have data on TDM showing 10 mHz 1/f knee.
    • will the community doubt 1/f knee as low as needed is possible? This is a critical component.
    • what 1/f knee is needed? if 1 RPM need 16 mHz.
    • 1 RPM is fast. Past experiments are 0.5-1 RPM.
    • Large broadband HWP is not possible. Need to address/understand 1/f.
    • (Jacques) 1/f not terrible problem if uncorrelated. only reduces low l sensitivity.
      • Gain changes and temperature stability are large problems.
    • (Julian) making single detector maps is useful in dealing with 1/f, i.e. scan stratagey matters.
      • Has pipeline to do large scale simulations. “Space is easy.”