Next: 7.5 Quality flags for Up: 7 Data Processing Level: Previous: 7.3 Signal Processing

7.4 In-band power calibration

Detailed description: section 5, 4.5.4

The steps described in this section are applied to all astronomical measurements except those collected with PHT-S. The PHT-S SPD processing finishes without this last step.

The detector responsivity varies along a revolution, in particular the long wavelength detectors P3, C100, and C200 show significant variations. To guard against such drifts, all PHT AOTs except P40 contain at least one FCS measurement per detector that has been used during an AOT. The FCS measurement determines the responsivity of the detector at approximately the same time as the sky is observed.

The derivation of the target's in-band power using the FCS calibration requires important instrumental corrections to ensure the best photometric calibration. The related processing steps are described in the following sections.

7.4.1 Determination in-band power from FCS

Detailed description: section 5

The electrical heating power p applied to the FCS directly corresponds to an FCS power on the detector P_FCS. The conversion is obtained from dedicated in-orbit FCS calibration observations where the signal of a calibration source is directly compared with the signal of the FCS for a given heating power p. However, the calibration observations could only cover a limited range in p due to limited availability of suitable calibration targets. FCS measurements with values for p outside the calibratable range will get unreliable P_FCS based on extrapolations.

In Derive_SPD the value for p is checked. In case the value is outside the valid range an extrapolated value for P_FCS will be retrieved, and a warning will be issued in the SPD file header.

Ancillary data required:

The valid heating power ranges are included in the FCS power calibration Cal G files PPxFCSPOW (x=1,2,3) for the P detectors, and PCxFCSPOW (x=1,2) for the C100 and C200 detectors, see section 13.10.

7.4.2 Signal to in-band power conversion for P detectors

Detailed description:5.2.3

From the electrical power (in mW) applied to the FCS, an in-band power on the detector P_FCS normalized to the aperture area (in ${\rm W/mm^2}$ )is obtained using FCS calibration tables, see Chapter 5. The FCS measurement provides the detector responsivity R:

$\begin{displaymath} R = \frac{{\langle s_{FCS} \rangle}\cdot{C_{det}}} {P_{FCS}{\cdot}A_{FCS}}~~~~~~~~~~A/W,\end{displaymath}$ (7.14)

where

$\langle s_{FCS} \rangle$ is the weighted mean signal of the FCS measurement (in V/s),
C_det is the detector capacitance (in F),
P_FCS is the in-band power on the detector from the FCS (in W/mm),
A_FCS is the aperture area of the FCS measurement in ${\rm mm^2}$ .

The in-band power P_target is obtained by:

$\begin{displaymath} P_{target} = \frac{\langle s_{target} \rangle C_{int}^{det}}{R}~~~~~~~~~W,\end{displaymath}$ (7.15)

and

$\begin{displaymath} {\Delta}P_{target} = \frac{{\Delta}\langle s_{target} \rangle} {\langle s_{target} \rangle} P_{target}~~~~~~~~~W,\end{displaymath}$ (7.16)

where P_target and $\langle s_{target} \rangle$ can refer to any filter in a multi-filter sequence taken with detector P1, P2, or P3. C_int^det is the capacitance of the detector.

Caveat: In principle, inhomogeneous illumination by the FCS (see section 7.4.3) affects the calibration involving apertures. The FCS power on the detector is not proportional with aperture area for a given electical power applied to the FCS. Although the correction method has been implemented in OLP Version 7, the present version of the Cal G files provide no correction. The investigations are ongoing, and non-unity correction factors are expected in a future version of the calibration files.

Ancillary data required:

The PHT-P FCS calibration tables in units of ${\rm W/mm^2}$ are stored in Cal G files PP1FCSPOW, PP2FCSPOW, and PP3FCSPOW for the three P detectors, see section 13.10.
The values of the FCS aperture area in ${\rm mm^2}$ from the aperture ID are hardcoded in the software. The values used are listed in section A.1.2. These areas were obtained from on-ground measurements of the instrumentation.
The correction method for inhomogeneous FCS illumination of the apertures assume the following Cal G files (see section 13.13): PP1FCSAP, PP2FCSAP, and PP3FCSAP.

7.4.3 Signal to-in band power conversion for C detectors

Detailed description: section 5.2.3

The detector flat-field or pixel-to-pixel variation in the PHT-C responsivity is found to be filter dependent. Correction for the pixel-to-pixel variation is in principle taken care of by the FCS measurement which determines the responsivity for each individual pixel. However, for a multi-filter observation the FCS measurement is done in only one filter. Therefore a correction is necessary that relate the flat field obtained from the FCS measurement with those in the other filter bands.

Firstly, the FCS measurement is used to determine the responsivity of each pixel R(i) for the filter in which the FCS measurement is taken:

$\begin{displaymath} R(i) = \frac {\langle s_{FCS}(i,f) \rangle \cdot C_{int}^{det}} {P_{FCS}(p,f) \cdot {\Gamma}(i,f) \cdot {\chi}(i,f)}~~~A/W,\end{displaymath}$ (7.17)

where

f indicates the filter, i the pixel number,
$\langle s_{FCS}(i,f) \rangle$ is the FCS signal per chopper plateau,
C_int^det is the detector capacitance,
P_FCS(p,f) is the in-band power (in W) for a given FCS heating power p,
${\Gamma}(i,f)$ is the illumination matrix, and
${\chi}(i,f)$ the detector flat field correction characterizing the filter-to-filter response.

The illumination matrix ${\Gamma}(i,f)$ is needed to correct for the fact that the FCS illumination is not flat but varies from pixel to pixel in the C100 and C200 arrays. This correction is applied to each pixel in the array with respect to a reference pixel for which the FCS calibration tables apply (section 4.5.4).

Secondly, the power on each pixel for any filter f' can be derived from the mean signal per chopper plateau $\langle s_{target}(i,f') \rangle$ :

$\begin{displaymath} P_{target}(i,f')\,=\,\frac {\langle s_{target}(i,f') \rangle \cdot C_{int}^{det}}{R(i) \cdot \chi(i,f')}~~~~~W,\end{displaymath}$ (7.18)

and

$\begin{displaymath} {\Delta}P_{target}(i,f') = \frac {{\Delta}\langle s_{target}... ...e} {\langle s_{target}(i,f') \rangle} P_{target}(i,f')~~~~~W.\end{displaymath}$ (7.19)

Ancillary data required:

The PHT-C FCS calibration tables in units of W/pixel are stored in Cal G files PC1FCSPOW and PC2FCSPOW. The tables give the mean in-band power from the FCS averaged over all pixels, see section 13.10.
The illumination matrices which give the FCS illumination for each pixel normalized to the mean pixel illumination are stored in Cal G files PC1ILLUM and PC2ILLUM, see section 13.12.
The filter-to-filter flatfield correction for C100 and C200 are stored in Cal G files PC1FLAT and PC2FLAT, see section 13.14. For each pixel the correction is normalized to the mean correction of all filters in the C100 or C200 detector.

7.4.4 In-band power for raster maps

Detailed description: None

In case of a raster map, two FCS measurements are collected: one at the beginning and one at the end of the map. It is assumed that differences between the derived responsivities of the two FCS measurements can be attributed to a detector responsivity drift which is linear in time. In order to compute the detector responsivity at the mean time of the plateau signal a linear interpolation in time between the two FCS signals is performed for the P detectors, and a linear interpolation in time between the two responsivities R(i) is performed for the C detectors (cf. eqn. 7.18).

Ancillary data required:

None

7.4.5 Default responsivities

Detailed description: None

It can happen that one or more FCS measurements are of insufficient quality for the derivation of a reliable detector responsivities. Such an event occurs e.g. in case all ramps/signals during the FCS measurement were rejected due to saturation or glitches.

For all chopped observations using the PHT03, PHT04 and PHT22 AOTs, the detector responsivity that can be derived from the chopped FCS measurement is assumed to be too unreliable for further use. In the current version of PHT-OLP the default responsivity is used for the in-band power calibration.

The default responsivity is time dependent and depends on the orbital phase of an measurement. For each measurement the orbital phase is read from ERD keyword TREFCOR2 which gives the orbital phase corresponding to the mid point of an AOT, and the default responsivity is determined by linear interpolation in the Cal G file which gives the default responsivity versus orbital position for a given detector.

Derive_SPD will also compute the FCS1 responsivity from the FCS measurement and the results are stored in the designated Cal A file. To give the observer some information on the discrepancies, the ratio:

$\begin{displaymath} r(i) = \frac {R_{FCS}(i)}{R_{default}(i)},\end{displaymath}$ (7.20)

is stored for each pixel i in the SPD header with keyword RESPRi.

Ancillary data required:

Default detector responsivities are stored in Cal G files PPRESP, PC1RESP, and PC2RESP for the P, C100 and C200 subsystems, respectively, see section 13.11.

7.4.6 Dependencies on mission dates

Detailed description: None

During the mission two main events took place which seriously impacted on the ISOPHOT calibration:

the change in FCS1-TRS2) emission properties (cf section 3.5 in revolution 94. This event caused the FCS emission for the long wavelength detectors (P3, C100 and C200) to become about a factor 2 higher than before for the same electrical power. Consequently, new uplink and FCS power tables had to be constructed after extensive in-orbit measurements of the FCS/TRS2 properties.
an extra curing procedure was introduced approximately halfway in the ISO science window for detectors P3, C100 and C200 as of revolution 150 onwards. The extra curing was performed during the handover period between the VILSPA and Goldstone ground stations. The procedure restored the nominal detector performance whereas, without the procedure, both detector noise and responsivity continue increasing.

As a consequence, the following calibration files have different entries depending on the revolution date:

FCS power Cal G files (PP3FCSPOW, PCxFCSPOW see section 7.4.1) are different before and after revolution 94.
orbit dependent default responsivities Cal G files (PPRESP, PCxRESP see section 7.4.5) are different before and after revolution 150.

Ancillary data required:

Information on the time dependence of ISOPHOT Cal G files is stored in Cal G file PTIMEDEP, a detailed description is given in section 13.2.

Next: 7.5 Quality flags for Up: 7 Data Processing Level: Previous: 7.3 Signal Processing

ISOPHOT Data Users Manual, Version 4.1, SAI/95-220/Dc