HXDS DATA REDUCTION TASKS

This site under construction.

$Id: reduction.html,v 1.14 1998/03/20 19:56:27 emk Exp emk $

REQUESTS FOR DATA FROM PTB

• Errors associated with the ring current measurements that have already been sent to us. We have the ring current ascii file from PTB, but there are no error estimates associated with the measurements. [Mike, 2/25/98] . Mar 19, 1998:COMPLETED.
• Photodiode measurements taken on the SX-700. If this is already included in the ring current files we have, then we need to know the conversion factor for converting the photodiode current (or ring current) to photon flux at a given energy with respect to the SX-700 monochromator. Specifically we need the photon flux hitting our detector for each of the measurements made at the SX-700. We need the errors for these measurements too if they are separate. [Mike, 2/25/98] Mar 19, 1998:COMPLETED for SSD.
• Ring parameters from white light calibration. Magnetic field, electron energy, etc that are necessary to calculate the photon flux. [Mike, 2/25/98]
• Ring current data for all FPC and photodiode data taken during the December (SX700) and January (KMC and 12.20) runs.[Brad, Mar 2, 1998]
• A discussion of the uncertainties in ring current measurements. (Different monitors are used in different current regimes, with varying uncertainties.)[Brad, Mar 2, 1998]
• Conversion from ring current to x-ray rate at all the energies we used on the SX700 FPC measurements with the 600-um aperture, and the uncertainties. As Mike said, what we really want is to know what the incident x-ray rate was on the FPC, and the net uncertainty in that rate, taking into account the photodiode dark current, photodiode current measurement uncertainties, beamline filter transmission uncertainties (and remember that for a few energies on the SX700, we used different-from-usual filters), SX700 efficiency, ESR errors, etc.[Brad, Mar 2, 1998][data received this week didn't include discussions of energy scale offsets, photodiode current fluctuations, white beam/aperture alignment errors, unfilled ring buckets, Brad, Mar 19, 1998]
• Energy scale offsets on the SX700, with uncertainties.[Brad, Mar 2, 1998] [data received this week didn't include this item, Brad, Mar 19, 1998]
• Energy scale offsets on the KMC, for each crystal, with uncertainties.[Brad, Mar 2, 1998]
• Photodiode data for all transmission measurements: HSI uv/ion shields on SX700 FPC window on SX700 unstretched polyimide/Al window material on SX700 filter transmission on SX700 filter transmission on KMC (beware of higher orders with beryl) with estimated uncertainties; also photodiode dark current.[Brad, Mar 2, 1998] [data received this week didn't include this item, Brad, Mar 19, 1998]
• Information on SX700 beam size with the 50-um aperture at 330, 500, and 930 eV. (We made measurements of beam size only at 330 eV. The other energies are less important, but if size information is available it would be useful.)[Brad, Mar 2, 1998]
• An appraisal of the reliability of the 12.20-beam aperture size measurements. One set of apertures did not have vertical position adjustment (and thus might not have been centered on the beam) and some apertures appeared to be partially blocked by something.[Brad, Mar 2, 1998]
• Ring parameters from white light calibration. Magnetic field, electron energy, etc that are necessary to calculated the photon flux. Preferred ring data format (in ascii): column 1 = decimal hours (e.g., 12.345) in UT or Germany time Preferably, do not reset to 0 when the day changes column 2 = ring current (in specified units). [Brad, Mar 2, 1998]

FPC DATA REDUCTION [Brad Wargelin]

• Count rate linearity data from BESSY. -- XSPEC fits Mar 9-20, 1998
• Window transmission -- Mar 9-20, 1998
• Analyze Pipe data:
  • calculate net fpc_hn collecting area ; relate fpc_x2 flat field relative QE to focussed beam QE ; calculate corrections for anode wire absorption, [Mar 9-13] [worked on during week ending Mar 20; not done.]
  • Plan to finish FPC support mesh blocking fraction data, and begin analyzing FPC window bowing and QE uniformity data, week ending Mar 27.]
• Line Response Function,relative QE of FPCs, -- Mar 20-Apr 10, 1998
• Window uniformity -- start on Apr 10.
• Absolute QE, awaits synchrotron model for fitting. Not yet scheduled.
• First draft of FPC report, April 15, 1998.

FPC data analysis (taken from bradslist)

This is a shopping list of things that Brad has kept. We need to consider how to implement it.

March 19, 1998 [Ed Kellogg]

1. gain nonuniformity (fpc_hn)

2. aging (fpc_x2)

3. aperture sizes (Ap Plate and blocking plates)
   • detector = photodiode
   • detector = FPC
   • optical measurements

4. window (and filter) transmissions
   • fpc_?? window trans with photodiode on SX700
   • fpc_?? window trans with photodiode on SX700
   • fpc_?? window trans with photodiode on SX700
   • unstretched polyimide trans with photodiode on SX700
   • fpc_?? window trans with photodiode on SX700
   • fpc_?? window trans with photodiode on SX700
   • filter trans with photodiode on SX700
   • filter trans with photodiode on KMC

5. spectral response functions
   • main peak shape--fano/polyah, width, etc.
   • shelf
   • tail
   • rate dependence
   • energy linearity
   • escape peak fraction

6. deadtime and pileup corrections
   • SSD with Fe55
   • pileup time from SRFs
   • counting rate linearity
   • root out bad pulser data from flat field tests
   • compare MCB vs. pulser deadtime for FPCs

7. SSD vs. FPC rate ratios with 2 and 4 mfp filters---XRCF flat field tests

8. QE from SX700

9. Ar-K edge scan on KMC
   • discontinuity in work function
   • QE jump across edge

10. QE reduction from anode wire absorption
    • by quadrant--no effect on T and B

11. QE from white beam
    • with no,Be,Ti,Cr,Fe filters
    • correct for wire obscuration and bowing

12. QE mapping (at 330, 500, 930 eV)
    • within windowlets
    • samples across entire detector

13. Gas Opacity
    • 1700 eV on SX700

14. Window support wire obscuration
    • relate fpc_x2 and fpc_hn small-beam QE results to XRCF flat fields
    • include energy dependence from reflections off wires

15. Write up GSS absolute pressure calibration at XRCF

16. Temp and pressure deltas
    • Define a baseline QE w.r.t. SSDs
    • Devise corrections to QE as f(T,P)

17. Ring focus FPC slit scan

18. Incorporate results into XSPEC/JMKMOD, Ping's mesh/aperture/window-bowing raytraces, etc.

19. Write SSD continuum analysis program, based on brute-force application of pileup rejection and escape event probabilities.

SSD DATA REDUCTION [Mike McDermott, Shaun Serej]

• Derive a spectral response matrix for the SSDs for use in XSPEC [Peter Barnes, COOMPLETE as of Feb 28, 1998, see SPIE paper]
• Reduce SX-700 data to produce QE(E) -- awaits PTB photodiode --> -- measurements. Planned for Mar 9-13, 1998.

  We have been working on the ssd_5 QE(E) measurements on the sx-700. A preliminary graph is /home/mcdermot/ssd5qe.ps. We processed the data by doing ROI sums over the peak, ROI sums over the interpeak events, and ROI sums over the pulser peak. We multiplied the interpeak events by 2 and added them back into the main peak. We then corrected for deadtime using the number of pulses recorded. Then we divided the number of counts by the true time. This result was divided by the photon flux supplied by Jutta Auerhammer of PTB. The statistics were about 0.1%. The photodiode uncertainty is about 0.6%.

  Next week, we plan to fit a couple of the spectra and see if we get a peak normalization that is comparable to ROI sum method. If this proves to be OK, then we will do the same ROI reduction of the SSD_X data to get its QE curve.

  We are also preparing to fit the data to a model of window transmission vs. energy, including carbon, oxygen and aluminum, and perhaps nitrogen. We are considering how to deal with the EXAFS near the Al K edge.

  Mar 12, 1998.

• Estqablish an energy scale for ssd_x

  Since we have no built in calibration source on ssd_x, and the BESSY crystal monochromator proved not to be reliable for energy reference, we are investigating alternate ways to find an energy scale for ssd_x. We are trying to add together a number of BESSY spectra to see if we can get good enough statistics on the fluorescence peak from Al-L. The first pass at it was less than thrilling. Mar 17, 1998. [M. McDermott]

• Reduce BESSY whitelight data on QE uniformity -- started, half done as of Feb 28, awaits PTB photodiode data; expect completion Mar 6.
• Reduce BESSY data on absolute QE-awaits synchrotron model. Not yet scheduled.
• First draft of SSD report, April 15, 1998.

HSI DATA REDUCTION [Ian Evans]

• CAT053 MST Due 980318 Changes in HSI quantum efficiency During the AXAF calibration activities, charge depletion decreased the gain in the central region of the HSI microchannel plate, resulting in an apparent decrease in quantum efficiency. OPEN Develop a strategy for approximately correcting the HSI data for changes in quantum efficiency. Estimated completion mid May 1998. [Ian Evans]
• Develop improved degapping algorithm to correct for variable gap geometry along the length of each gap. Started as of Feb 28, 1998.
• Produce spatial distortion map. Not yet scheduled.
• Correct flat field data for large scale beam non-uniformities. Not yet scheduled.
• Develop algorithm to correct flat field and experimental data for spatially periodic detector sensitivity variations. Not yet scheduled.
• Develop algorithm to smooth/fit flat fields over subarcsecond spatial scales and correct flat field and experimental data for spatial gain non-uniformity. Not yet scheduled.
• Develop algorithm to correct data for gain degradation, if feasible. Not yet scheduled.
• Produce electronics count rate linearity calibration. Not yet scheduled.
• Correct angular dependence of QE calibration data for vignetting, and determine MCP pore-angle incidence vectors for calibration data. Not yet scheduled.
• Develop algorithm to correct data for angular dependence of QE, including extrapolation to energies not calibrated directly. Not yet scheduled.
• Determine absolute on-axis QE by reference to FPC absolute QE at calibration energies. Not yet scheduled.
• Incorporate ion shield transmission measurements to produce improved QE model across relevant edges. Not yet scheduled.
• First draft of HSI report, April 15, 1998.