Subsections

• 2.4.1 Signal flow in ISOPHOT
• 2.4.2 Cold Readout Electronics (CRE)
• 2.4.3 Detector Interface Electronics (DIE)
• 2.4.4 Readout timing control

2.4 Data Collection with ISOPHOT

The analysis and reduction of ISOPHOT data require a basic knowledge of the instrument itself and the instrument procedures performed during an observation. The way ISOPHOT data were taken is given in the following section. This section is of importance to understand the structure of ERD products (Chapter 6).

2.4.1 Signal flow in ISOPHOT

In this section we introduce the elements related to the ISOPHOT signal flow. For the understanding of the instrument output related to an astronomical observation, systems on both the cold and warm side of the cryostat were of importance, see also Figure 2.4:

Figure 2.4: Main systems for data collection.

• The cold focal plane unit (FPU) inside the ISO cryostat (payload module) described in (Wolf 1994, [57]; Lemke et al. 1993, [35] and Lemke et al. 1994, [36]) containing
  • detector subsystems with the cold readout electronics (CRE),
  • the change wheel (CHW) system,
  • the focal plane chopper (FPC), and
  • the internal reference sources (FCSs).
• The warm external electronics unit (EEU) in ISO's service module containing
  • the central processing units,
  • the readout timing control and voltage supply units, and
  • the detector interface electronics (DIE) of the processing subsystem.
  Telecommands were processed by the central processing electronics. The electronics for the CHWs, temperature control, bias supplies, etc. were in the EEU. The central processing unit also generated the instrument telemetry. It combined the science data from the detector with relevant housekeeping information on the selected subunits. In case PHT was the prime instrument, but no exposure was ongoing, extended housekeeping telemetry was generated.
• The ISO on-board data handling system (OBDH) in ISO's service module which communicated with the instrument EEU. The instrument telemetry coming from the EEU was combined with the pointing data from the attitude and orbit control system (AOCS) and put in telemetry blocks. Each telemetry block or record covers a time period of 2s. The telemetry of a whole revolution is stored in a TDF (Telemetry Distribution File).

2.4.2 Cold Readout Electronics (CRE)

The ISOPHOT detectors were either Silicon doped or Germanium doped photoconductors. The function of the Cold Readout Electronics (CRE) unit was to pre-amplify and sample the photo-currents generated in these detectors. To reach the maximum efficiency, the CRE was integrated on the chip containing the detector. For more details, see Wolf 1994, [57] and Wolf, Grözinger & Lemke 1995, [58].

Figure 2.5: Schematic diagram showing the detector - CRE amplifier chain, the capacitance is also denoted by in this document.

The detector/CRE chain in Figure 2.5 represents an integrating amplifier with capacitive feedback. Each detector pixel was connected to one input channel of a CRE. In case the detector was receiving IR photons from an astronomical or internal source, the voltage V at the output of the CRE increased as a function of time. The voltage increase per time interval was dependent on the amount of the current through the detector which is in turn dependent on the number of photons falling on the detector. In the remainder of this manual this voltage increase or slope (in V/s) will be referred to as photo-current or signal. For the observer the signal is a fundamental concept: once signals from calibration standards have been measured, it is possible to relate the signal to a power from the source in the filter band.

After the amplifier stage there was a sample & hold (S&H) stage. By applying a sample pulse the voltage was clamped and while the front-end continued to integrate, the clamped signal was read out through a multiplexer (MUX). In that way the readout did not disturb the integration.

Since the output voltage should stay within a limited range, the voltage was reset by short circuiting the capacitors using both the ground as well as the reset switch. A reset pulse was applied in addition to a sample pulse after a number of desired voltages had been sampled. The readout associated with this reset is called the destructive readout, the other sampled voltages are called non-destructive readouts. The time between two reset pulses (in seconds) is called the fundamental integration time or reset interval (RI). The duration of a reset interval is denoted by . All readouts collected during one reset interval are part of one integration ramp or simply ramp. In Figure 2.6, the readout data stream is shown schematically.

The voltage of the first readout after reset is not zero but starts at an arbitrary voltage level. This level is called the reset level, since it has more or less the same value during a measurement. Variations were caused by instabilities in the amplifier electronics.

Figure 2.6: Schematic diagram showing CRE output for a ramp of 7 non-destructive readouts (NDR=3).

2.4.3 Detector Interface Electronics (DIE)

The detector interface electronics (DIE) acquired the output signals from the CREs and processed these for the telemetry transmission. It also served as a multiplexer thus reducing the number of output lines. For redundancy reasons there were two DIEs in the EEU, each interfacing with the following groups of detectors:

• DIE1: C200, P1, P2, and SL
• DIE2: C100, P3, and SS

The DIE subtracted a commandable offset from the CRE output voltage and amplified the difference with a commandable gain factor. Note that for PHT the gain was a constant unlike other instruments for which the gain was used to adjust the dynamical range. The dynamic range was adjusted by selection of the proper reset interval (see Section 2.4.4).

The DIE also converted the analogue signal from the CRE to a digital signal. This AD conversion produced a 12 bits value per readout which means that each readout has a digitalisation resolution of 4096. The fact that there are several readouts per ramp increases the resolution in the signal.

2.4.4 Readout timing control

To cover a dynamic range equivalent to the flux density range of astronomical sources (a few mJy to hundreds of Jy) within the available commanding bits in the telecommands, the design of the readout parameters was based on an exponential scale with base 2. Only a power of 2 readouts per reset interval was commanded, and the duration of a reset interval (in seconds) had also to be a power of 2. Since there were 4 bits available for commanding, the reset interval could have 16 values ranging from 1/256, 1/128, 1/64,128 s. Per ramp there were readouts. The last one, the destructive readout, was generally not reliable, since it was disturbed by the reset. The number of non-destructive readouts per ramp was determined by the parameter NDR, it is equal to .

The parameter DR was defined to set the chopper plateau time. DR was defined such that is equal to the number of destructive readouts (or number of ramps) per chopper plateau.

The chopper dwell time or duration of a chopper plateau can be derived from:

(2.1)

The commanding of PHT imposed at least 4 ramps per chopper plateau, thus DR for chopped observations. The maximum plateau time is 128 s. The first ramp of a chopper plateau is affected by the chopper transition and therefore unreliable. DR was always commanded, both for staring and chopped observations. In case of staring observations, DR was commanded such that s. After 128s a virtual chopper transition (with physical chopper throw ) occurred.

For high readout rates the amount of data to be transferred was too high to be matched into the ISO TM format. In order to avoid telemetry overflow, the parameter data reduction (DAT_RED) was introduced. The value of DAT_RED is an integer indicating that only the first ramp of a sequence of DAT_RED ramps should be transmitted. For example, DAT_RED = 4 means that only the first of every 4 ramps is transmitted. DAT_RED = 1 indicates no reduction. Observations of bright sources had a DAT_RED parameter larger than 1. The value of DAT_RED depended not only on NDR and RI, but also on the number of pixels per detector array.