Bicep2-list December 2012

bicep2-list@lists.fas.harvard.edu

20 participants
19 discussions

Re: [Bicep2-list] FTS prep, updated figures

by John Kovac

Glad to see the initial interferograms, Jon. But I'm frustrated you and Grant never answered the below email, and have not confirmed what your plan is for complete consistency checks--I spelled out many options last month. Grant's email below only indicates that you'll do as many pointings as needed to get complete coverage, and then try several angles to fill out A/B response. I'm not sure what he means by "Incidental to this would be redundancy of coverage" but that plan does not sounds nearly sufficient to ensure we have a handle on systematic uncertainties in this measurement. Positive vs. negative going scans is a start, but it only probes a very limited class of systematics. As Zak says, take more pointings than you think you need--at least 9 sounds wise--so that every pixel is hit with multiple alignments. Multiple mirror speeds is a good test. Please answer my question below about chop vs scan modulation. And please take multiple input grid configurations. I would also like to know how whether the spectra you obtain have any systematic dependence on your internal FTS alignment procedure. If you adjust the two arms within a range of reasonable alignment how do the spectra change? John On 12/13/12 2:58 PM, John Kovac wrote: > > How about different output grid configurations? In one of my previous > emails I talked about many options there. I think that is a very > useful probe of repeatability. > > Are you going to chop or scan modulate? Can you try both? > > And what mirror travel are you using? Symmetric about zero path or > one sided? In either case, can you test for FTS alignment related > systematics by comparing spectra from positive range of mirror travel > vs. negative travel? This may also be worth thinking about. > > J > > > > > > On 12/13/12 2:40 PM, gteply(a)caltech.edu wrote: >> What I had in mind was that we would tilt the output mirror to several >> positions to assure complete coverage of all rgls. Then we would rotate >> the entire system with respect to the cryostat by several different >> angles >> to get both A+B. Incidental to this would be redundancy of coverage, >> assuring repeatability. Also, at each position, we would take multiple >> interferograms. >> >> I can't think of any reason to restrict the travel range and lower the >> resolution except in real time to help with alignment. >> >>> Jon and Grant, >>> >>> Could you please tell me exactly what different configurations of the >>> FTS you will be trying on in order to test systematic repeatability? >>> >>> What combinations of changes of grid orientation, output port, input >>> port, modulation, and/or mirror travel range do you expect to be >>> trying? >>> >>> John >>> >>> >>> >>> >>> >>> On 12/13/12 1:48 PM, John Kovac wrote: >>>> Hi Jon, >>>> >>>> We just spent some time on the phone discussing improvements that can >>>> be made to your comparison plots so they are more useful for judging >>>> whether the pixel-to-pixel variation we see in FTS-derived quantities >>>> (bandcenters, bandwidths, relgain mismatches) is repeatable or not. >>>> To do this we need to be able to directly compare pixel-to-pixel >>>> variations and variations among measurements. >>>> >>>> - histograms of both of these things, i.e. >>>> 1. differences between repeated measurements >>>> 2. differences of each pixel's average value vs. FP mean >>>> both of these things should be signed (don't plot abs of the >>>> difference) >>>> >>>> A basic question is whether we've detected any significant variation >>>> among pixels, or whether the variation in 2 is consistent just with >>>> the "noise" indicated by 1. Your plots should attempt to answer this. >>>> >>>> Also, make scatter plots, please. >>>> >>>> Colin, Grant and others on the phone can add to this feedback. Despite >>>> the poor quality of the B2 measurements in your March posting, you can >>>> still try making the above plots. I expect they will illustrate that >>>> last year the repeatability was too poor to significantly measure >>>> actual pixel-pixel variation in these quantities...but make them >>>> anyway and explicitly draw that conclusion. >>>> >>>> John >>>> >>>> >>>> >>>> >>>> >>>> On 12/12/12 7:17 PM, Jon Kaufman wrote: >>>>> Hey John, >>>>> >>>>> I couldn't get the old data with the extremely choppy satellite >>>>> schedule yesterday but I took the k5 lab run data and broke it into >>>>> two data sets for comparison. I've added a few plots to my posting >>>>> <http://bicep.caltech.edu/%7Espuder/analysis_logbook/analysis/20121207_FTS_p…> >>>>> >>>>> showing >>>>> differenced band centers and bandwidths for the two data sets. I >>>>> plan to make very similar plots to figures 10 and 11 but for rgls and >>>>> plots similar to figures 3 and 4 here >>>>> <http://bicep.caltech.edu/%7Espuder/analysis_logbook/analysis/20120319_FTS_s…>. >>>>> >>>>> This won't take me longer than an hour once the data is in the can. >>>>> >>>>> Are there more plots that you think might be useful? >>>>> >>>>> -Jon >>>> >>>> -- >>>> ___________________________________________________________________ >>>> John Kovacjmkovac(a)cfa.harvard.edu >>>> >>>> Assistant Professor, Astronomy and Physics, Harvard University >>>> 160 Concord Ave rm 310, Cambridge MA 02138, 617-496-0611 >>> >>> -- >>> ___________________________________________________________________ >>> John Kovac jmkovac(a)cfa.harvard.edu >>> >>> Assistant Professor, Astronomy and Physics, Harvard University >>> 160 Concord Ave rm 310, Cambridge MA 02138, 617-496-0611 >>> >>> >> > > -- ___________________________________________________________________ John Kovac jmkovac(a)cfa.harvard.edu Assistant Professor, Astronomy and Physics, Harvard University 160 Concord Ave rm 310, Cambridge MA 02138, 617-496-0611

11 years, 5 months

Fwd: Re: FTS prep, updated figures

by John Kovac

-------- Original Message -------- Subject: Re: FTS prep, updated figures Date: Thu, 13 Dec 2012 13:48:02 -0500 From: John Kovac <jmkovac(a)cfa.harvard.edu> To: Jon Kaufman <jkaufman(a)physics.ucsd.edu> CC: gteply(a)caltech.edu, Colin Bischoff <cbischoff(a)cfa.harvard.edu>, ibuder(a)cfa.harvard.edu, Kirit Karkare <kkarkare(a)cfa.harvard.edu> Hi Jon, We just spent some time on the phone discussing improvements that can be made to your comparison plots so they are more useful for judging whether the pixel-to-pixel variation we see in FTS-derived quantities (bandcenters, bandwidths, relgain mismatches) is repeatable or not. To do this we need to be able to directly compare pixel-to-pixel variations and variations among measurements. - histograms of both of these things, i.e. 1. differences between repeated measurements 2. differences of each pixel's average value vs. FP mean both of these things should be signed (don't plot abs of the difference) A basic question is whether we've detected any significant variation among pixels, or whether the variation in 2 is consistent just with the "noise" indicated by 1. Your plots should attempt to answer this. Also, make scatter plots, please. Colin, Grant and others on the phone can add to this feedback. Despite the poor quality of the B2 measurements in your March posting, you can still try making the above plots. I expect they will illustrate that last year the repeatability was too poor to significantly measure actual pixel-pixel variation in these quantities...but make them anyway and explicitly draw that conclusion. John On 12/12/12 7:17 PM, Jon Kaufman wrote: > Hey John, > > I couldn't get the old data with the extremely choppy satellite > schedule yesterday but I took the k5 lab run data and broke it into > two data sets for comparison. I've added a few plots to my posting > <http://bicep.caltech.edu/%7Espuder/analysis_logbook/analysis/20121207_FTS_p…> showing > differenced band centers and bandwidths for the two data sets. I plan > to make very similar plots to figures 10 and 11 but for rgls and plots > similar to figures 3 and 4 here > <http://bicep.caltech.edu/%7Espuder/analysis_logbook/analysis/20120319_FTS_s…>. > This won't take me longer than an hour once the data is in the can. > > Are there more plots that you think might be useful? > > -Jon -- ___________________________________________________________________ John Kovacjmkovac(a)cfa.harvard.edu Assistant Professor, Astronomy and Physics, Harvard University 160 Concord Ave rm 310, Cambridge MA 02138, 617-496-0611

11 years, 5 months

Fwd: Plan for taking/analyzing this summer's Pole FTS measurements, both B2 and Keck

by John Kovac

-------- Original Message -------- Subject: Plan for taking/analyzing this summer's Pole FTS measurements, both B2 and Keck Date: Wed, 21 Nov 2012 12:38:55 -0500 From: John Kovac <jmkovac(a)cfa.harvard.edu> To: gteply(a)caltech.edu, Jon Kaufman <jkaufman(a)physics.ucsd.edu>, Kirit Karkare <kkarkare(a)cfa.harvard.edu>, ibuder(a)cfa.harvard.edu CC: Zak Staniszewski <zks(a)caltech.edu>, Abigail Vieregg <avieregg(a)cfa.harvard.edu>, Colin Bischoff <cbischoff(a)cfa.harvard.edu>, Clem Pryke <pryke(a)physics.umn.edu>, Brian Keating <bkeating(a)ucsd.edu> Hi Jon, Grant, Kirit, and Immanuel, I've talked with each of you about the FTS measurements at Pole, which will start with B2 in early December. I just want to facilitate communication on this thread: Jon and Grant have volunteered to take the lead on B2 and Keck measurements, respectively. Immanuel will arrive at Pole ~Nov 5, and Kirit a week later--both will make completing FTS measurements and analysis their top priorities. Since this FTS will return to Harvard they should become expert in using it, and with Grant can hand it off to Abby and me in Jan for final Keck use and retro. We need to turn these data around as quickly as possible, verifying completeness of coverage on our ~250 light B2 detector pairs and data quality as judged by repeatability of conclusions. Our goal should be to take multiple sets of spectra on all B2 rgl's, with different FTS mounting orientations and/or grid polarizations, and verify repeatability of not just the spectra but the actual parameters we derive from them which we will use in analysis. These parameters should include *band centers* for an assumed source spectrum (e.g. CMB), and more importantly predicted *elnod relgain mismatch* as Evan calculated from B1 spectra in this posting. <http://bicep.caltech.edu/%7Ebicep/analysis_logbook_north/20101021_specleak3/> You could start working right now on the analysis that will be needed to derive these parameters from your spectra. A posting could demonstrate repeatability using scatter plots between two measurement sets, and plot significance assessed from scatter among measurements as in this recent posting <http://bmode.caltech.edu/%7Espuder/analysis_logbook/analysis/20121109_BeamP…> on beam parameters. Our goal is to see such postings before warming up B2. We'd like to aim for Dec 10. As you know, Dec 25 is the drop-dead deadline. Please work together, but don't be afraid to produce independent analyses of the same data--we need multiple team members on this to ensure timely convergence. John -- ___________________________________________________________________ John Kovacjmkovac(a)cfa.harvard.edu Assistant Professor, Astronomy and Physics, Harvard University 160 Concord Ave rm 310, Cambridge MA 02138, 617-496-0611

11 years, 5 months

Re: [Bicep2-list] [keckarray] South Pole Notes 12/18/2012 -- FTS plan

by John Kovac

Hi Roger, Jon, and Grant: > I wanted to give a more detailed update on things down here especially in light of the BICEP-2 mailing list email discussion about the FTS measurements- Immanuel’s notes and Jon Kauffman’s post did not fully communicate those activities Roger, thank you for the report. I am responding to your BICEP2 FTS plan and questions below, and cc'ing the B2 list. It is great to see a plan finally laid out. Grant and Jon, you really should have responded to my suggestions and request for a plan in my emails to you on Nov 21 and Dec 13. My current request for tests should be no surprise to you--please re-read those emails to understand where they are coming from. I will forward them to Roger. Mark Halpern and Zak: thanks for the suggestions on specific FTS systematic effects to watch for. Please keep feedback coming. Time is short, and we need this to be a thorough job. Here is a (partial) list of systematics we know can affect FTS spectra, shifting band edges, band shape, and/or out-of-band features: 1. nonlinear detector response 2. detector (/FPU) transfer function effects 3. multipath interference within the FTS 4. multipath interference between FTS and BICEP2 5. FTS-arm alignment effects 6. mirror travel dependent alignment effects 7. mirror velocity error 8. mirror velocity non-uniformity 9. microphonics, mechanical resonances 10. frequency-dependent beam coupling efficiency Others with FTS experience, please feel free to add to that list. The different tests we are proposing attempt to eliminate these things (and many tests probe more than one) as significant sources of systematic uncertainty. We've discussed how to quantify "significant" for many spectrum-derived quantities--the B2 science analysis is ultimately what sets the spec. It would make sense to focus first on addressing all the systematic/repeatability concerns, using a subset of detectors, and THEN to take complete data. Until now, Jon and Grant have been saying that taking all the data we want will go extremely fast...and perhaps it has. I certainly hope that the complete coverage data you've taken is archival quality, but we won't know until you've passed consistency tests. > -Roger, Grant, and Kauffman have developed a plan for consistency checks on the new FTS being used with BICEP-2, **which I elaborate on below.** We have taken much of the data and Kauffman is reducing it into *.mat files with spectra and interferograms so we can parallelize the analysis between multiple team members. Kauffman has already started some analysis, posted to the BICEP-2 analysis logbook. The initial posting looks promising. > -We think we need 1 week to warm up BICEP-2, investigate wiring issues (mux col 14 & row select shorts), and close D2 into RX3. We’re scheduled to cool down RX3 on 12/26, so we need to be mindful of this schedule when deciding which FTS measurements to focus on. Our drop-dead date for starting to warm BICEP2 is Dec 24--and if we do push it that late then obviously it will have an impact on Keck's rx3 closeup. We would like to see complete FTS results with convincing analysis of systematic uncertainties finished as soon as possible. > PLAN FOR FTS: > We are taking data now to address the following systematics concerns (hopefully the answers to each of these questions will prove to be “not significantly”): In your postings you should work to quantify "significantly". We don't want this to be open-ended. Eyeballing spectra is a first step, but we need science-guided spec's on each of the quantities derived from these spectra. For example, I've stated that we'd like to see predicted relgain mismatch measured to < 0.2%, based on scaling from Takahashi et al. > 1) Are FTS spectra pointing dependant? We are currently taking the 9 goniometer/rotation stage settings suggested by Zak. This is being done at a dec angle 45deg between polA & polB. > 2) Are FTS spectra scan direction dependant? We have been taking data with the FTS running both directions and Kauffman’s post begins to address this. > 3) Is there any hysteresis in the stage that might cause differences between spectra that accumulate over several runs? We have taken FTS scans with eight passes, and can compare the different passes to look for such problems. good tests--they probe some combinations of systematics 4,5,6,8,10. > 4) Are the FTS spectra speed dependent? Tonight, we are taking spectra at faster and slower speeds (a factor of two different) with the goniometer and rotation stages fixed for both speeds. This will be done centered over every tile. probes systematics 2,7,8,9 > 5) Are the FTS spectra dependent on orientation of the FTS wrt BICEP-2? We have taken spectra oriented in alignement with pol A, pol B , 45 deg between polA and pol B, and finally aligned with pol A but rotated 180 deg form above. We have taken these centered over each of the four tiles. also a good test, probing 4 and 5 > 6) Are the interferograms asymmetric and influencing the spectra? We will FFT interferograms stitched together from half sweeps of the FTS and their mirror images (mirrored with respect to time). We will then compare the pairs of spectra. > 7) Do the spectra depend on alignment procedure? While the alignment procedure is a multi-dimensional space, most components are heavily constrained except for the lateral position and orientation of the beam splitter. Grant and Roger spent four hours using an alignment procedure that sought to produce visually symmetric spectra with zero-path positions at consistent locations on a sample of pixels in the FPU. We will compare spectra between ‘aligned’ and ‘misaligned’ settings of the beam-splitter. Thank you for describing the alignment procedure. Given this, did the procedure allow you to record interferograms from the latter stages of alignment that can be analyzed to determine sensitivity of the spectra and derived quantities to the final adjustments? Of course you need only explore the "reasonable" region of alignment parameter space to judge what the systematic uncertainties might be. A separate test would be to change the path length to the fixed roof mirror by a few inches. This is a strong probe of systematic 3 (as well as 5). It would be very interesting to see whether the fringe structure in the passband responds to such a change. Is this easy? > 8) Do the spectra depend on an overall 90 deg rotation of wire-grids in the FTS? This is a good one to check and we are interested in doing it. That said, the grid frames are not designed to be rotated 90 degrees and we may need to drill holes in the frame to do this. We do not feel confident doing this, although we can discuss this with our machinist (Steal). Do people on the team feel that such modifications in the field are prudent? my email from Nov 21 also suggested removing the output grid. This forces the BICEP2 detector polarization to define the coupling through the FTS--A and B detectors will measure opposite input ports at zero path. This seems like a very stringent test--it probes a combination of at least 1,3,4,5. If the negative-going interferograms for a given A detector at dk=90 yield consistency with the positive-going ones for A taken at dk=0, and if differential A vs B quantities are consistent under this inversion, then I will be very reassured. Another test of 1 (also suggested by Mark Halpern) is to alter the source temperature. > 9) Do the spectra depend on modulation technique? (i.e. chop or scan modulated). This has been suggested by John Kovac, but we are unclear what symmetry this tests and hope he can provide clarification. If his concern is simply signal-to-noise after experiences with the UCSD FTS, this may no longer be a concern, as evidenced by the figures in Kauffman’s posts. Regardless, this may require a significant modification to the FTS. We hope Zak and comment on how this might be accomplished. No, S/N was not the motivation--I've been taking high S/N for granted. Changing the source in this way would probe 1, 2, and 7-9. I asked about this some time ago, when there was plenty of time to prepare, but if the above tests all pass, I'm OK dropping this. John -- ___________________________________________________________________ John Kovac jmkovac(a)cfa.harvard.edu Assistant Professor, Astronomy and Physics, Harvard University 160 Concord Ave rm 310, Cambridge MA 02138, 617-496-0611

11 years, 5 months

South Pole notes, 2012-12-11

by Bischoff, Colin (Guest)

hi all, We had some good progress over the weekend. [Thursday / Friday] * Finished the fiducial set of four dk angles for BICEP2 RPS observations. * Ran a schedule to repeat the RPS observation for one row of detectors at a single dk angle three times in a row. * Deliberately misaligned the RPS by 15 degrees in azimuth and repeated the same row of detectors three more times. * Took down the RPS, put up the uberchopper, and ran unpolarized beam maps at dk=+-45. [Saturday] * Immanuel arrived. * Bench tests of the rotating polarized source, to look for collimation offset and beam ellipticity. * Put the RPS back up, but with the wire grid set to horizontal alignment to run pixel pol raster schedules (RPS orientation fixed, rotate BICEP2 in dk to map out the detector angles). [Sunday] * Fixed an azimuth offset in our pixel pol rasters, so now we are actually hitting the source. Continued running this schedule through Sunday night. We should have data for 30 rgl pairs spread across the four tiles. [Monday] * Took off the BICEP2 far field flat! * Ran yukical on BICEP2. * More bench tests of the RPS -- looking at stability of the power output. * Set up source for BICEP2 sidelobe mapping. First sidelobe map is at very low source power -- chop amplitude is only 1000 ADU for a co-polar detector. (not dubstepping the detectors) Also, during this span of time, cooled down K5 in MAPO with just the camera installed (but no detectors, obviously), to observe whether the bellows contraction is significantly different than when the cryostat is simply under vacuum. I have started transferring some arc files from Keck RPS observations. These are going through the usual route, so I think they should make it to spud / odyssey automatically. Can someone check to see if there are any Keck arc files on odyssey from late on November 15 to early November 16? [Plan for today] * Finish RPS bench tests -- last test we want to run is to observe the source power output while we let it cool down outside. * Repackage BBNS from RPS to use for Keck sidelobe mapping. Set up source for Keck sidelobe maps. * Keep moving along with BICEP2 sidelobe map itinerary. [Rest of the week] * BICEP2 sidelobe mapping will take us through to Thursday morning at least, but there are additional sidelobe schedules that we would run if more time is available. * Chin Lin and Colin are leaving on Saturday. I'm hoping to get BICEP2 out of the mount and into the lab stand before I leave, which means Friday at the latest. Seems to imply that we could continue sidelobe mapping through Thursday (either one more source configuration, or maybe just contingency in case something gets screwed up). * Plenty of Keck sidelobe mapping to do, since we don't really have any useful sidelobe data from previous summers. Is there a particular date when we should cut this off to take down rx1?

11 years, 5 months

BICEP2 is off the mount!!!

by Jonathan Kaufman

And currently observing the ceiling... :) Thanks to the fantastic, hard work of all those deployed, BICEP2 came off the mount at approximately 8:00 PM Pole time. There were a few head-scrathing moments but no catastrophes. I again want to personally thank all those who are down here who worked so hard for this historic event and those up north who I'm sure wished they could be here with us. I intend to run a fridge cycle tonight and start FTS testing in the morning. -Jon

11 years, 5 months

BICEP2 team meeting + Pipeline call today (16 Oct 2012), 9am Pacific

by John Kovac

Hi all, BICEP2 team meeting today (Tues) at 9-10am Pacific. Thanks to each of you who sent reports AHEAD OF TIME this week, even late last night. Sorry these came too late for me to do the agenda yesterday, but please do look over the below before the meeting. Phone: 1-866-890-3820 (toll: 1-334-323-7229) Passcode:59702175 Walt will take notes <https://docs.google.com/a/cfa.harvard.edu/document/d/1iVrfX19cXQr8fUgRYOUos…> for the team meeting. The bispud pipeline telecon will follow immediately, run by Clem. Except as called out below, all NEW analysis postings will be reviewed in the pipeline telecon, not the main telecon. 1) Weekly report from Pole [ Steff ] - yet another great observing week; good weather - Jon's thermal tests: 3 additional schedules run, report below. 2) Data quality / reduc czar report [ Walt ] AIFL: star pointing cut--fixed AIFL: El nod mismatch 20120929F10 <http://bicep.rc.fas.harvard.edu/bicep2/reducplots/201209/browser/20120929F1…> follow up next week's czar: ________________ 3) Testing and Deployment Planning: Thermal testing report [ Jon ] 5-10 min Oct 9: xfer function low freq <http://bmode.caltech.edu/%7Espuder/analysis_logbook/analysis/20121009_therm…> Oct 14: xfer function high freq (Jon, fix name!) <http://bmode.caltech.edu/%7Espuder/analysis_logbook/analysis/20121014_therm…> data great 0.0001-0.01 Hz, but 1e-3 response at 1 Hz? analysis/modeling needs work (offline!) - are we done taking xfer fun data? - PID characterization status November/DecemberMeasurement Plan <https://docs.google.com/a/cfa.harvard.edu/spreadsheet/ccc?key=0Ao4UxezLygx5…>(view only!) [ Colin (/Randol) ] 5 min - plan for pol rot? (need analysis of last year?) - plan for sidelobe maps? (who writes these schedules?) - plan for analysis? FTS status [ Zak ] 4) Analysis Status Reports: Baseline 2yr data reduction status report [ Walt ] 10-20 min Walt to give overview of this thread [ see his summary report appended below ] Characterizing non-beam systematics [ Walt / Angiola ] brief! briefly list threads / status / people: - relgain mismatch from per-channel abscal still in progress [ Walt ] - crosstalk, new posting to discuss next hour [ JET ] - other threads here: thermal, magnetic, ? Characterizing beam systematics [ Abby (/Randol) ] brief! briefly list threads / status / people: - main beam parameters [ follow up to Randol's post <http://bmode.caltech.edu/%7Espuder/analysis_logbook/analysis/20121004_beams…> last week ] - un-deprojected residuals (main beam and sidelobes) - other threads here: pol calibration errors, ? Deprojection / beamsim algorithm coding status report [ Clem ] brief! briefly list threads / status / people: - deproj sim test of linearity: add timestream noise [ Chris ] - deproj sim test of beam width/ellipticity [ Stefan ] - other: simulation of more complicated beam shapes, ? Data consistency check status report [ Chris (/Randol) ] big news: B2 2yr maps DK_jack cleans up dramatically <http://bicep0.caltech.edu/%7Espuder/analysis_logbook/analysis/20121008_sim1…> with deproj further discussion of new postings in next hour briefly list threads / status / people: - comparison of recovered deproj coefs w/ measured mismatches - full jack chi2's, sum_bp_devs, etc. Foregrounds status report: [ Jamie T ] brief! Statistics for reporting bandpowers, r constraints [ Colin / John ] brief! End. ---- Appendix, from Walt: Baseline two-year data reduction --------------------------------- What we've been doing: * Spent substantial time in 2012 addressing discrete artifacts in the data, including glitches and step-like crosstalk. * Cut definition is nearly complete. * The recent reducplots show the success of these efforts: there are only a very few examples of non-standard behavior remaining in the scan plots. Efforts now in progress and to be completed soon: * El-nod goodness of fit cut: - Now have strong evidence this cut is important for BICEP2: + 2012 Oct 12: Threshold-crossing gain change in GCP 165 (RWO) + 2012 Oct 12: Threshold-crossing gain excursion in more channels (RWO) - Martin has defined several new cut parameters - Angiola will regenerate cutparams files, cutplots, study thresholds - Regenerate cutparams, default cuts when ready. * Other cuts: - Thresholds for stationarity_ab and elnod_mean are now set. - Will take Jamie Tolan's adjustments to scanset_std and tfpu_std. - Angiola to make sure default cut thresholds get updated, include in new cut plots. * Follow-up on the few remaining oddities in reducplots. - Walt working on this now. - May require additional cuts or tweaks to existing cuts, thresholds, or deglitching. - In case of deglitching changes, could require new run of low-level reduction. * NTD calibration - Walt to propose scheme for applying offline cal for periods when needed. - Will redo low-level reduction for this period only. * Once this is done, can define stable TODs / data selection / cuts / pairmaps / reducplots. Next steps: * Write up final observing efficiency, etc. Easy -- follows SPIE papers with a few updates. * Human eyeball test for final version of all reducplots. - Will need volunteers... Decisions to make: * The high-weight catch has been removed from reduc_makepairmaps. Restore in reduc_coaddpairmaps or in reducplots? * How to make maps of dark TESs, dark SQUIDs, NTD? - How to apply cuts? - "clean" version would require substantial changes to code - put off until after BICEP2 first results? -- ___________________________________________________________________ John Kovac jmkovac(a)cfa.harvard.edu Assistant Professor, Astronomy and Physics, Harvard University 160 Concord Ave rm 310, Cambridge MA 02138, 617-496-0611

11 years, 5 months

Thermal conductance of a Duband fridge?

by Runyan, Marcus C (389I)

Hey folks - I'm curious to know how good of an effective heat switch a 4He fridge is. Has anyone measured the thermal conductance between the 4He tip of a Duband fridge and the 4K cold plate when the 4He section of the fridge is filled with gas? Even a few Delta_T vs P data points would be helpful. Thanks, Marc Marc Runyan Marcus.C.Runyan(a)jpl.nasa.gov JPL Superconducting Devices Group (389I) MS 302-205 4800 Oak Grove Dr. Pasadena, CA 91011 Office: 302-205J cell: 818.298.5932, office: 818.354.8871

11 years, 5 months

Fwd: SPTR request -- optimal filesize for SPTR is ~1 GB.

by John Kovac

Hi all, Those who are using the SPTR to transfer summer data north: the below email recommends ~1 GB filesize for maximum efficiency. If we have a large number of much smaller files it may make sense to tar them up to chunks of that order. John -------- Original Message -------- Subject: SPTR request Date: Mon, 3 Dec 2012 08:37:38 +1300 From: McAfee, Bill (Contractor) <Bill.McAfee.Contractor(a)usap.gov> To: Kovac, John A-039-S;A-149-S <jmkovac(a)cfa.harvard.edu>, Pryke, Clement A-039-S; A-149-S; A379-S <pryke(a)physics.umn.edu> SPTR request Hi guys, Please pass this along to the appropriatemembers ofyourBICEP2 and SPUD teams. The File Catalyst software that we run on the SPTR servers(Pole and White Sands)is most efficient when transferring large files. When transferringa lotof small files, as some South Pole science projects do from time to time, the ratio of data to overhead is lousy, and performance is degraded. On the other hand, if those small files are bundled together into a few large files (~one gigabyteeach), the applicationruns more efficiently. Thanks, Bill

11 years, 5 months

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

Bicep2-list December 2012