From: CANSAS::IN%"white@irg.usask.ca" "Megan White" 3-OCT-1996 20:48:21.89 To: IN%"jeg320@mail.usask.ca", IN%"gene@irg.usask.ca" "Gene McDougall", IN%"osterried@dansas.usask.ca", IN%"woodh@engr.usask.ca", IN%"duanes@eagle.wbm.ca", IN%"scitec.eng@eagle.wbm.ca", IN%"jeffj@eagle.wbm.ca", IN%"efurkal@plasma.usask.ca", IN%"dolovich@e CC: Subj: optics team minutes Return-path: Received: from skycat.usask.ca by SKISAS.USask.CA (PMDF V4.2-11 #3676) id <01IA7X4UW0748WWBBH@SKISAS.USask.CA>; Thu, 3 Oct 1996 20:47:51 GMT Received: from irg.usask.ca by sask.usask.ca (PMDF V5.0-6 #15020) id <01IA7KJSANZ48Y90OG@sask.usask.ca>; Thu, 03 Oct 1996 14:47:22 -0600 (CST) Received: by localhost from irg.usask.ca (router,WinSmtp -Win32- V1.07beta1.7.s); Thu, 03 Oct 1996 14:44:28 -0600 Received: from baldur.usask.ca by irg.usask.ca (128.233.17.24::mail daemon,WinSmtp -Win32- V1.07beta1.7.s); Thu, 03 Oct 1996 14:42:57 -0600 Date: Thu, 03 Oct 1996 14:45:12 +0000 From: Megan White Subject: optics team minutes To: jeg320@mail.usask.ca, Gene McDougall , osterried@dansas.usask.ca, woodh@engr.usask.ca, duanes@eagle.wbm.ca, scitec.eng@eagle.wbm.ca, jeffj@eagle.wbm.ca, efurkal@plasma.usask.ca, dolovich@engr.usask.ca, jim_bugg@engr.usask.ca, jas140@mail.usask.ca, ron_bolton@engr.usask.ca, soteros@snoopy.usask.ca, raj@snoopy.usask.ca, david.steele@sask.usask.ca, smolyakov@sask.usask.ca, matthias.huber@sask.usask.ca, kustov@SKISAS.USask.CA, m_stocki@engr.usask.ca, Dave.Harris@sask.usask.ca, kkc290@mail.usask.ca, painter@lighthouse.usask.ca, JJ Offiong , dsl@skatter.usask.ca Reply-to: Megan White Message-id: <19961003144428.38694ceb.in@irg.usask.ca> X-Envelope-to: STEELE X-Mailer: Pegasus Mail for Windows (v2.23) Content-transfer-encoding: 7BIT Priority: normal Comments: Authenticated sender is Hello All: Below are the minutes of the Optics team meeting on Wed. Oct. 2, 1996 (rtf format). To view, save message to a file, open in a text editor, and delete everything up to and including my name. Then open with a word processor. Good luck! Megan. {\rtf1\ansi\deff0\deftab720{\fonttbl{\f0\fnil MS Sans Serif;}{\f1\fnil\fcharset2 Symbol;}{\f2\fswiss\fprq2 System;}{\f3\fnil Times New Roman;}} {\colortbl\red0\green0\blue0;} \deflang1033\pard\plain\f3\fs20\b\ul Notch Filter\plain\f3\fs20 \par \par \pard\li360\fi-360{\*\pn\pnlvlblt\pnf1\pnindent360{\pntxtb\'b7 }\plain f3\fs20 {\pntext\f1\'b7\tab}Combine two filters to cut out the red and green lines in the spectrum and keep most of the data available. \par {\pntext\f1\'b7\tab}Throughput should be sufficient (Dave is talking to the manufacturers for the complete specs) \par {\pntext\f1\'b7\tab}Would create known wavelength markers. \par \pard\plain\f3\fs20 \par Looking at the absorbed spectrum (with the unabsorption spectrum) that Doug provided: \par - At levels where the transmission is high, we will have good S/N, but the numbers in the comparison will be close to one. \par - At low transmissions (lowest point in the absorption), bad S/N but a large difference for comparison \par - Which of these would give better data in the end? \par \par Calculation: \par \par assume 20k cnts at spectral peak above O3 \par \par high transmission \par 500nm: above: 20k counts +- sqrt(2*20k+30*30) = 20k +/- 200 \par \tab within: 10k cnts +/- sqrt(2*10k + 30*30) = 10k +/- 145 \par where factor of 2 is from paper (given by Dave) and 30 is the readout error \par \par ratio = (10/20) +/- 0.012 = 0.500 +/- 0.0012 = 2.5% accuracy \par where 0.012 = (d(10k)*(20k) + (10k)*d(20K))/(20k)^2 \par \par low transmission \par 600nm:\tab above:\tab 20k +/- 202 \par \tab within:\tab 2K +/- 70 \par \par ratio = 0.10 +/- 0.0045 = 4.5% accuracy \par \par therefore, worse at low signal as opposed to low absorption \par *this is 'best case' for absorption, but should still have good accuracy higher in the layer \par *ratio gets bigger higher in the layer, but S/N goes up too \par \par ***Aside: best case resolving power = 100:1 S/N \par \par \plain\f3\fs20\b\ul Miscellaneous \par \plain\f3\fs20 \par Big FOV: - good for tomography (duplicity of measurements) \par Small FOV: - better for optics (less problems with abberation, focus) \par \par \pard\li360\fi-360{\*\pn\pnlvlblt\pnf1\pnindent360{\pntxtb\'b7 }\plain f3\fs20 {\pntext\f1\'b7\tab}could go with no lens if we use a concave grating \par {\pntext\f1\'b7\tab}are pressure and temp needed for Rayleigh scattering (what about $$) \par {\pntext\f1\'b7\tab}An intesifier would give bigger noise than dark current \par {\pntext\f1\'b7\tab}We are making continuous measurements of background signal from pixels without spectrums \par {\pntext\f1\'b7\tab}How are we going to de-spin? Can we use the parachute? \par {\pntext\f1\'b7\tab}We can afford to throw away stars which are too high or low, but not the ones on the sides of the CCD picture \par {\pntext\f1\'b7\tab}Assuming 40 deg. FOV, could we squash 20 deg. vertically to smaller area on CCD to save in transmission? \par {\pntext\f1\'b7\tab}Relative calibrations of different pixels will be required (we will never see the same star on the same pixel) \par \pard\plain\f3\fs20 \par \plain\f3\fs20\b\ul Which part of Spectrum do we Use? \par \par \pard\li360\fi-360{\*\pn\pnlvlblt\pnf1\pnindent360{\pntxtb\'b7 }\plain f3\fs20 {\pntext\f1\'b7\tab}Will NO2 absorption give us problems if we don't go with just 70 nm range (from 560-630nm)? We have the NO2 cross-section, now need density to determine how much problem it could create. \par {\pntext\f1\'b7\tab}***Rayleigh scattering being worked out more accurately. Will we avoid scattering probs using just the 70nm range? \par {\pntext\f1\'b7\tab}70 nm bandpass = smaller optics (good point) \par {\pntext\f1\'b7\tab}If we reject wavelengths beyond the red and green lines, what will the actual passband look like (transmission, fall off at edges, etc.)? \par {\pntext\f1\'b7\tab}Stellar features aren't wide enough to cover a pixel (over which we average). \par {\pntext\f1\'b7\tab}Do we want lots of stellar structure or not? \par {\pntext\f1\'b7\tab}In the range between red and green lines, there is less star structure. \par {\pntext\f1\'b7\tab}We will have to deal with star structures for each star (will that create problems?). \par \pard\plain\f3\fs20 \par \plain\f3\fs20\b\ul The CCD \par \plain\f3\fs20 \par \pard\li360\fi-360{\*\pn\pnlvlblt\pnf1\pnindent360{\pntxtb\'b7 }\plain f3\fs20 {\pntext\f1\'b7\tab}Consider the spin of the instrument. Do we want to orient the CCD so that the smear of the spectrum is along the 'wavelength axis' (so as to stretch it out) or along the 'amplitude axis' (so as to increase the height of the spectrum)? \par \pard\plain\f3\fs20 \par With a grating of 600 lines/mm: \par \par off-axis angle (1st order):\tab \tab 17.5 deg @ 500nm \par \tab \tab \tab \tab 21.5 deg @ 600nm \par \tab \tab \tab \tab 25.6 deg @ 700nm \par \par therefore, we have 2.8 deg. for 70 nm (probably too high an amount) \par 20 deg. vertical extent \par \par 200nm = 8deg. \par 5 nm = 0.2 deg. = 1 pixel \par therefore 20 deg. = 500nm = 100 pixels \par \par 500 pixels available \par using (1/5) optics without smearing effects \par \par focus = ? for reasonable size CCD \par \par biggest CCD/intensifier pixel: 10-20 microns \par \par 100 pixels = 2mm \par f# = diameter/f = 1mm/L=tan(10deg.)=0.17 \par * looks like we will get 1st order for sure \par \par ** which equals 6mm focal length \par ** we want higher resolution to see absorption line more clearly (which will increase the S/N) \par \par \pard\li360\fi-360{\*\pn\pnlvlblt\pnf1\pnindent360{\pntxtb\'b7 }\plain f3\fs20 {\pntext\f1\'b7\tab}If we have the zeroth order, we can get rid of video camera \par {\pntext\f1\'b7\tab}What are the data rates for the video cam (will this limit use of vid cam?) \par {\pntext\f1\'b7\tab}Can we set the data rate if carrier freq. is 1.6 GHz? Is there something stopping us? \par {\pntext\f1\'b7\tab}Do these factors depend on the modulating frequency? \par {\pntext\f1\'b7\tab}If we have 0.1 bits/Hz === 150 Mbits/sec (no data transmission problem) \par {\pntext\f1\'b7\tab}minimizing angular resolution of pixels will help with aurora problems (less background noise per pixel) \par \pard\plain\f3\fs20 \par \plain\f3\fs20\b\ul Rockets\plain\f3\fs20 \par \par \pard\li360\fi-360{\*\pn\pnlvlblt\pnf1\pnindent360{\pntxtb\'b7 }\plain f3\fs20 {\pntext\f1\'b7\tab}Met rockets have cheaper launch costs \par {\pntext\f1\'b7\tab}one type: burns for 2 sec and coasts to 120 km \par {\pntext\f1\'b7\tab}looks like most feasible choice is the 2-3kg payload, range to 85km \par {\pntext\f1\'b7\tab}Launch site: Churchill has lots of aurora, but maybe deal on launch? Cold Lake has a bit less aurora, but dealing with Military might be a problem \par \pard\plain\f3\fs20 \par } ______________________________________________________ Megan White Institute of Space and Atmospheric Studies University of Saskatchewan < white@irg.usask.ca >