Sun Angle In Degrees

Fig. 6-16. Plot of the Output From Representative Photocells Versus Sun Angle for Adcole Digital Sun Sensors

The fine bits in Fig. 6-14 are used by an interpolation circuit to provide increased resolution. Straightforward addition of encoded rows to the pattern is not possible because thé 0.53-deg angular diameter of the Sun from near the Earth would blur the output from adjacent bits. This effectively limits Gray 'code transitions to a 1/2-deg LSB. By combining the output of 2 or 3 offset LSB patterns in an interpolation circuit, 1/4- or 1/8-deg transitions are obtained.

Two-axis sensors consist of two measurement components mounted at right angles, yielding a 64- by 64-deg or 128- by 128-deg FOV as shown in Fig. 6-17. Full 4w sr coverage for the two-axis sensors is obtained by use of five or more 128- by 128-deg sensors. (See Sections 2.1 and 7.1.) Onboard logic for selecting and telemetering data from the illuminated Sun sensor in multisensor configurations is based on monitoring the output of the ATA photocell and selecting the sensor with the highest output signal (effectively the smallest angle relative to the optical null or boresight angle).

6.1.4 Fine Sun Sensors

Increasingly stringent attitude accuracy requirements, such as for IUE, MAGSAT, or SMM, imply Sun sensor absolute accuracies of several arc-minutes to 5 arc-seconds and even better relative accuracies. Resolutions of less than I /8-deg LSB, the practical limit of the device shown in Fig. 6-14, to an LSB of 0.1 arc-second can be achieved by electronically combining the output current from four offset photocells beneath a reticle pattern as shown in Fig. 6-18. The SMM fine Sun sensor, shown here, consists of an entrance slit composed of 72 pairs of alternately opaque and transparent rectangles of equal width (0.064 mm); a 1,5-cm spacer; an exit slit composed of four offset reticle patterns, each with 68 alternately opaque-transparent rectangle pairs; and four photocells, one beneath each pattern [Adcole, 19771. The entrance and exit slits are separated by a vacuum to reduce the effect of spectral dispersion on the accuracy.

The Sun sensor output is periodic, with a period which can be adjusted to meet the accuracy and FOV requirements, e.g., a 2-deg period for SEASAT with a ± 32-deg FOV and a 1-deg period for SMM with a ± 2-deg FOV. The fine Sun sensor is combined with a digital Sun sensor to resolve ambiguities in the output angle. Two sensors are mounted perpendicular to one another for two-axis output. The sensor operation is described more fully in Section 7.1.

. 6.2 Horizon Sensors

Gerald M. Lerner

The orientation of spacecraft relative to the Earth is of obvious importance to space navigation and to communications, weather, and Earth resources satellite

payloads. To a near-Earth satellite, the Earth is the second brightest celestial object and covers up to 40% of the sky. The Earth presents an extended target to a sensor (3.9 sr at a 500-km altitude) compared with the generally valid point source approximations employed for the Sun (7X10-5 sr) and stars.* Consequently, detecting only the presence of the Earth is normally insufficient for even crude attitude determination and nearly all sensors are designed to locate the Earth's horizon. (The detection of the presence of small planetary bodies, such as the Moon from a near-Earth orbit, is, however, sufficient for coarse attitude determination.) Horizon sensors are the principal means for directly determining the orientation of the spacecraft with respect to the Earth. They have been employed on aircraft and were used on the first U.S. manned flights in the Mercury and Gemini programs [Hatcher, 1967]. In this section we describe the requirements imposed on horizon sensors, outline die characteristics of several generic types, and describe the operating principles of horizon sensor systems in common use.

As described in Chapter 4, the location of the horizon is poorly defined for a body possessing an atmosphere because of the gradual decrease in radiated intensity away from the true or hard horizon of the solid surface. However, even a body possessing no atmosphere, such as the Moon, poses a horizon sensor design problem due to variations in the radiated intensity. To illustrate an extreme case, a detector triggering on the lunar horizon in the 14- to 35-/im infrared spectral region

* Betelgeuse, the star with the largest angular radius, subtends 6x 10"14 sr.

will experience fiftyfold variations in radiance (120°K to 390° K.) between illuminated and unilluminated horizons. As illustrated in Fig. 6-19, if the radiation integrated over half the sensor field of view (FOV) is just above threshold at a cold Moon, the horizon location error at a hot Moon is half the sensor FOV because the sensor would then trigger at the edge of the FOV. Lowering the threshold or decreasing the sensor FOV may not be possible because of the low intensity of emitted radiation relative to noise for practical detectors."Thus, for the lunar horizon, a different choice of spectral region (the visible) is frequently employed to provide a sufficient radiation intensity with a small FOV.

Earth resources, communications, and weather satellites typically require pointing accuracies of 0.0S deg to less than a minute of arc, which is beyond the state of the art for horizon sensors. However, Earth-oriented spacecraft frequently employ autonomous attitude control systems based on error signals from horizon sensors with accuracy requirements of 0.5 to 1 deg. Thus, although payload requirements may not be met by current horizon detectors, control requirements are easily met, and significant cost savings may be realized by increasing the accuracy of horizon sensors to meet payload and control accuracy requirements simultaneously and thereby avoid the necessity of flying star sensors.

In Chapter 4, we showed that the position of the Earth's horizon is least ambiguous in the spectral region near 15/tm in the infrared. Most horizon sensors now exploit the narrow 14- to 16-/im C02 band. Use of the infrared spectral band avoids the large attitude errors encountered on Mercury, Gemini, and OGO due to spurious triggerings of visible light (albedo) horizon sensors off high-altitude clouds [Hatcher, 1967]. In addition, the operation of an infrared horizon sensor is unaffected by night or by the presence of the terminator. Infrared detectors are less susceptible to sunlight reflected by the spacecraft than are visible light detectors

Was this article helpful?

0 -1

Post a comment