Attending: Julian, Toki, Tom, Kris, Brian, Shaul, Qi, Karl, Jeff, Amy, + (…some who joined late)

Notes by : Karl

• Focal Plane layout + Sensitivities (Karl)
• Initiating simulations (Julian)
• Readout mux factor, wire count (Roger?)
• Temperature stability from Planck?
• Sun Shield configurations: (Qi)
  • Deployable trade-offs and "Add-on" Sunshields

E-mail from Matt regarding FDM Power

Hi Shaul,

We're at about 3.2W per module assuming FPGA technology and existing DAC/ADCs that are flight qualified. Changing to ASICs could bring this down slightly, but not a lot (the FPGA is about 25% of the power, if I remember correctly).

The power does not go up much if the mux factor per module increases - so getting to 128x or 256x is very helpful in power consumption. The warm readout can support this already. The cold readout is limited by strays to increase the mux factor, and its hard to predict where we'll end up there. If I were choosing a target, I would specify a goal of 200x, and baseline of 128x (375W for 15,000 detectors). Real investment in the cold system would be needed to get there. Kam Arnold at UCSD and Toki Suzuki at Berkeley are the main ones working on this. Moving the SQUIDs to the cryogenic stage can help.

Matt

Minutes

Focal plane and noise:

• Karl to make memo of all assumptions going into these calculations
• Using curved focal plane with larger DLFOV.
  • More flexibility in number pixels per band, can trade synchrotron pixels for CMB pixels
  • with larger mirrors can do 28k bolos.
• CMB sensitivities for cases discussed last week.
  • Numbers in presentation.
  • Biggest effects:
    • Emissivity to 0.06% from 1%, 20% gain in sensitivity
    • stop to 6.5 K from 4 K, 10 % loss in sensitivity

Aside: observing efficiency

• Continuous 100% ? vs single shot 80-85% ADR
• Toki: may need downtime for data transfer. RF interference with SQUIDS
  • Jeff: SPIDER saw problems. Tech dependent, some systems are more RFI sensitive than others. our specific case will need investigation.
• Downlink time? Julian: 15 TB / day uncompressed. With fewer bits, compression ~ 1TB/day
  • Amy: Possible tech (implementation in ~2 yrs) Ka band, 30 GHz, 150 Mb/sec.
    • 1 Tb is about 10 hrs. Likely DSN time is limited to few hours.
  • Shaul: We may need to do some amount of onboard preprocessing.

Readout:

• need wire counts and mux factor. Roger done TDM. Toki will do FDM.
  • 100 mK –> 4 K
  • Toki: what is cooling power? Can give 2 cases.
    • Just wire count for now. lengths and materials are helpful as well.
    • Assuming: Achieved, x68. Can also do a possible value, x105 from EBEX or similar.
    • Berkeley development working to x300. Requires significant work still.
    • Do in next 1-2 wks. Karl relay number px and band layout, etc.
• Power use, Roger calculating

Simulations starting (Julian)

• software running, was developed and used for Planck.
• running noise map of few detectors.
• Andreas Zonker running this
• systematics group deciding what to include
  • sidelobes (brad doing grasp calculations)
  • calibration and gain variations
    • metrics such as S/N on dipole
  • Polarization calibration a work in progress

Sun shields

• options to get alpha = 30.
  • Julian: alpha of 35 optimum was for LB. should recalculate optimum for PICO
• deployable shields
  • for alpha 30, need 50 cm shields
  • really shields need to be 'flatter' (same angle as inner V-grooves) so slightly longer.
• 'add-on' shields. 2nd set above first. 2nd set covers primary.
  • Does this make sense?
  • Definitely would need more thermal modeling.
  • Amy: Make everything a tulip? What about just a very steep tulip, single set of V-grooves.
  • Amy: Useful for Team-X. Concept is currently volume limited telescope. When Team-X says we're in budget, this lets us have examples to show them.
    • In same vein, if telescope is cost limited we should have a design plan for this. Could have some smaller systems prepped. Karl will prep a 1.2m Open Dragone system.