II.C.5 Electronics

IRAS Explanatory Supplement
II. Satellite Description
C. Telescope System Overview
C.5 Electronics


Chapter Contents | Authors | References
Table of Contents | Index | Previous Section | Next Section

Figure II.C.10 Preamp and bias supply schematic. The JFET module is indicated by the dashed line, the dewar boundary by the dash-dot line.
larger largest

The photoconductive detector elements responded to infrared radiation by altering their electrical resistance. Figure II.C.10.a shows schematically the nominal preamplifier and bias voltage design. A matched pair of junction field effect transistors (JFETs) for each detector acted as a unity-gain source-follower amplifier, converting the high impedance output of the photodetectors to low impedance for transmission to the warm electronics outside the dewar. The JFET pairs were each suspended by Dacron threads inside a small 2 K light-tight box such that electrical dissipation in the JFETs themselves (about 200 microwatts) maintained the JFETs at temperatures of 60 to 70 K. A 3 Momega-uc metal-film resistor cemented to the JFET acted as a heater for cold starting the amplifier during ground testing and several hours after launch. The JFETs formed a differential input stage of the trans-impedance amplifier. At low frequencies the output voltage of the trans-impedance amplifier was equal to the voltage difference between the gates of the JFET pair plus any offset voltage at the input of the operational amplifier plus the voltage drop across the feedback resistor due to the photocurrent from the detector.

The detector bias was applied to one detector contact and the trans-impedance amplifier maintained the other contact at a constant DC voltage very near signal ground independent of photocurrent until the trans-impedance amplifier output saturated at about 11 volts. All detectors in a module had a common bias voltage, applied through the module frame, which are listed in Table II.C.6. An exception to this biasing scheme was module A in the 25 µm band. During testing, the frame of this module became shorted to signal ground rendering the entire module inoperative. The alternative biasing approach used for this module only (suggested by Dr. J. Houck) is shown in Fig. II.C.10.b. The bias voltage, with reversed polarity, was applied to the gate of the reference JFET. This voltage also appeared at the operational amplifier input so the output of the trans-impedance amplifier was compensated by the same amount. The net effect of this modification was to increase the gain of the trans-impedance amplifier by a factor of 1.2.

During laboratory testing the responsivity and noise of the detectors were found to depend on their history of exposure to energetic radiation, such as gamma-rays, and energetic electrons and protons. The observed sensitivity change resulting from an exposure of 0.6 Rads of Co60, roughly equivalent in dosage to a passage through a deep portion of the South Atlantic Anomaly (SAA), was about a factor of 1.2, 2, 6 and 10 for the 12, 25, 60 and 100 µm detectors. Biasing the detectors into a "breakdown" condition resulted in a large current flowing through the detector that annealed the radiation effects. The hardware provided a second bias voltage level, called the "bias boost" voltage, which effected this annealing process after exposure to SAA protons during the mission. The bias boost voltages are listed in Table II.C.6 and the effects of the bias boost are discussed in Section IV.A.7. Because the 12 µm detectors did not require annealing, the second bias voltage provided an alternative operating bias.
Figure II.C.11 Resistance vs. voltage characteristics of feedback resistor for detector 29. Measurements were made at 2 K. Solid line shows the shape of resistance vs. voltage relation adopted for processing.
larger largest

The rolloff frequency of the trans-impedance amplifier was set to approximately 80 Hz by a 0.1 pf shunt capacitor across the feedback resistor. The feedback resistors, Eltec model 102 metal film resistors, were selected from 1 × 1010 room temperature elements. At 2 K their impedance was 2.05 ± 0.1 × 1010 and varied slightly with voltage. Figure II.C.11 shows a sample resistance vs. voltage curve as measured at 2 K. A combination of three straight lines fitted to the measured points defined the shape of the non-linear resistance versus voltage relationship used for data reduction (see Section VI.A.5).

Electrical Characteristics of Survey Arrays
Table II.C.6
Effective Wavelength (µm) 12 25 60 100
Nominal bias (volts) 3.27(A)
2.50(B)
1.50 0.160 0.185
Boosted bias (volts) 2.50(A)
2.00(B)
11.00(A)
7.00(B)
1.00 1.00
Nominal/Low Gain 7.18 5.98 10.8 13.4
Nominal/High Gain 0.107 0.109 0.102 0.100

The rest of the telescope electronics processed the signals from the visible and infrared preamplifiers, transferred data to the spacecraft onboard computer for storage and subsequent transmission to the ground station, and received commands from the spacecraft and distributed them to the telescope systems. This data-processing was split into two major elements: the analog electronics and the digital electronics. More details are contained in Langford et al. (1983) and Long and Langford (1983).
Figure II.C.12 Focal plane array infrared channel data flow.
larger largest

The analog signal path for the infrared detectors was entirely DC-coupled. Figure II.C.12 shows a functional diagram of the components of the analog electronics for one infrared detector channel. In order to maintain a negative voltage at the output of the analog electronics, the DC offset voltage at the output of the trans-impedance amplifier could be changed by 17.8 mV steps using an 8 level commandable offset with level 4 corresponding to no change. The commandable offset levels utilized during the mission are listed in Table II.C.3.

Nuclear pulse circumvention circuitry prevented sharp pulses from cosmic rays and charged particle hits on the detectors in the SAA and in the polar horns from contaminating the infrared data stream. The output of the trans-impedance amplifier was fed to an integrator and to a pole-zero amplifier which flattened the frequency response to 450 Hz to improve the operation of the circumvention circult. The output of the pole-zero amplifier went to a differentiator and to a Bessel filter which delayed the signal by about 150 µs. The differentiated and integrated signals led to a comparator which opened a switch to prevent the unwanted, fast rise-time pulses from passing through the system. The integrator raised the minimum threshold to blank the unwanted spike as the DC voltage from the trans-impedance amplifier increased. The track/hold capacitor clamped the input to the gain amplifier to a fixed level while the switch was open. Further details of the design and performance of the pulse circumvention circuitry can be found in Emming et al. (1983) and Long and Langford (1983).

The Bessel filter boosted the trans-impedance amplifier output by a factor of two. An additional amplifier could increase the system gain by software commandable factors of unity (low gain), of 5 to 12 depending on the wavelenght band (nominal gain), and of ten times nominal gain (high gain). All survey scans were made using the nominal gain except in those areas, typically close to the Galactic plane. Some of these areas were rescanned using low gain (Section III.D). The overtall nominal gain for each infrared channel and the ratios of the nominal to low and high gains for the different detector modules are listed in Table II.C.3. Finally. 12 dB/octave low-pass filters with cutoff frequencies of 6, 6, 3, 1.5 Hz for the 12, 25, 60 and 100 µm bands, respectively, limited the frequency response and reduced high frequency noise. The outputs of the low-pass filters were fed into multiplexers and then to a 16-bit analog-to-digital converter operating at 125 µV per data number for subsequent processing by the digital electronics.

The visible detector data flow was similar to the infrared data flow, except that the trans-impedance amplifier was AC-coupled to a MOSFET preamplifier. For both infrared and visible channels, the pole zero amplifier, integrator, differentiator, comparator, track/hold calacitor, switch, gain amplifier and low-pass filter were contained in a single miniature hybrid circuit.

Under low background conditions the limiting noise in the analog electronics chain was the Johnson noise of the 2 × 1010 omega-uc feedback resistor. At a temperature of 2 K this noise level was roughly 1.6 µV Hz.

The digital electronics processed the digitized infrared and visible detector data, collected telemetry information from various sensors located on the telescope, received and executed commands issued from the onboard computer and transmitted the formatted telemetry, infrared and visble detector data to the onboard computer. The infrared detectors were sampled at 16, 16, 8 and 4 Hz at 12, 25, 60 and 100 µm, respectively. To minimize spacecraft data storage requirements, the digital electronics compressed each 16-bit infrared detector value to an 8-bit value representing the difference between the successive 16-bit numbers. For details of compression scheme, see Appendix II. 1.

During star crossings, two of the eight visual detectors were sampled at a 500 Hz rate. The onboard digital electronics determined the visual magnitude of the star and its crossing time, passed this information to the spacecraft attitude control software to update the satellite pointing and recorded the data for subsequent use in the attitude reconstruction. The digital electronics also measured 108 temperatures, voltages and pressures to monitor the health of the telescope. These and other housekeeping data were multiplexed, digitized and formatted for transferral to the onboard computer.


Chapter Contents | Authors | References
Table of Contents | Index | Previous Section | Next Section