time will depend upon the resistivity of the membrane that can be selected between 1011 and 1017 ohm-centimeters.

The well-focused electron beam that approaches the target with only a few volts of energy responds to the potential modulation at the scanned surface by depositing sufficient electrons to restore the original equipotential. In detail, only the most energetic portion of the beam is induced to contact the positively modulated areas, and only a portion of these contacting electrons remain at the target to erase the stored charge pattern. Typically, for modulations between 0 and 5 volts, two out of every three electrons striking the target are elastically scattered and are returned along with the unused portion of the beam to the entrance of the electron multiplier at the gun end of the tube.

If the cross-section of the return beam is examined at this plane where there is maximum dispersion of the electrons according to their component of energy normal to the tube axis, the spatial distribution of scattered and reflected electrons is that shown in figure 2. The reflected electrons are confined to the small circular area. The scattered electrons occupy the larger area, which increases in diameter as the potential of the point of scan becomes more positive. When the return beam is directed over an appropriate aperture, the

reflected electrons are removed; and the scattered electrons are allowed to enter the electron multiplier. The function of the multiplier is to amplify the video current (the scattered electron current) sufficiently so that amplifier noise does not degrade the signal. The scattered electrons entering the multiplier are two to three times the amplification of the stored target charge or six to ten times the amplification of the photoelectron current.

The experimental astrometric work was done using laboratory-design isocons. The new RCA image isocon tube, along with its associated focus and deflecting coils, is a more highly perfected electron optical design with superior performance. Nonuniformities in scanning-beam landing energies at various points in the scanning raster have been reduced to less than 0.1 volt, and improved

control of the signal separation has been achieved. The tube setup has been simplified. The image isocon tube features:

1. High signal-to-noise ratio, particularly in the dark area of a scene

2. A linear light-to-signal transfer characteristic over a brightness range of at least 100

3. An intra-scene dynamic range of 100 that retains photometric accuracy

4. High geometric fidelity in the image with excellent resolution

5. No tube adjustments needed to cover these ranges.

Two versions of the tube are available: Type 21093 contains a bi-alkali photocathode and a high capacitance target; Type 21095 has a fiber optics faceplate, an S-20 photocathode, and a low capacitance target.

Variations for slow-scan, long-integration tubes include smaller gun and separation apertures, finer meshes in the target assembly, and Elcon glass targets of 1015 ohmcentimeter resistivity.

Astronomical Applications of the Image Isocon

Figure 3 lists a number of requirements for an image-sensing system that would have wide application in astronomy. How capably the image isocon fulfills these needs is examined in the following discussion.

REQUIRED CAPABILITIES OF AN ASTRONOMICAL IMAGE SENSOR SYSTEM

1. IMAGE INTEGRATION FOR VARIABLE BUT CONTROLLED EXPOSURE LASTING FROM < 1 SEC TO SEVERAL HOURS

2. TRANSMISSION OF THE IMAGE DATA FROM ONE EXPOSURE IN A SMALL NUMBER OF SCANS

3. AN ABILITY TO DETECT POINT IMAGES IN A DARK FIELD

4. PHOTOMETRIC AND GEOMETRIC FIDELITY IN THE TRANSMITTED IMAGE CAPABLE OF QUANTITATIVE EVALUATION

5. A CAPABILITY TO RECORD SPECTROMETER DATA WITH MAXIMUM FIDELITY

Figure 3. Requirements for an integrating camera tube system for use in astronomy.

Camera-Tube Operating Cycle of the Stratoscope Television System

The experimental work has confirmed that the operating cycle established for the Stratoscope Television System does yield reproducible data when an intermittent cycle of expose-store-and-read is employed.

The first step calls for preparation of the target. This is necessary to erase any residual modulation from the storage target that remains from previous operation. It also establishes a known reference potential at the target surface. With target resistivities as high as 1017 ohm-centimeters, this erasure becomes an increasingly important factor in obtaining reproducibility since the relaxation time of the insulator is measured in days.

The tube is now ready for the optical exposure. During the extended time of light exposure, the photocathode and the electrodes of the image section that focus the charge image on the storage surface are the only parts of the isocon that must be active. Exposures as long as three hours have been made with the reciprocity between light intensity and exposure time being rigidly maintained.

For the readout of the stored charge, the image section of the tube is made inactive, and, after time has been allowed for the gun temperature to stabilize, a single readout scan of the target is made. The scattered electron current, which is proportional to the target potential modulation, is the output signal that is recorded.

This camera system uses a 1-second vertical sweep time for the 500-line raster. Recording of the video output can be on magnetic tape, by direct visual display, or on photographic film.

The single slow-speed readout scan removes 97 to 98 percent of the stored charge at the target as compared with 70 percent for the normal scan velocity. A reduction of the beam current to one-thirtieth its normal value provides a finer beam-spot diameter and an improved modulation transfer function (MTF).

Camera systems intended for quantitative photometric use, particularly at slow readout rates, must have very well stabilized voltage and current sources if the performance capabilities of the sensor tube are to be fully realized.

To provide an image signal with the most favorable signal-to-noise ratio from the object of interest, it is necessary to know something of the signal intensity. This foreknowledge permits the exposure to be adjusted so that the charge stored at the target from this part of the scene is at the maximum established by the target capacitance. Overexposure leads to distortion of the photometry. Overloads in various amplifiers, the transmitter, the receiver, or the display system must also be avoided by careful calibration of permissible ranges for the signals.

Sensitivity and Photometric Fidelity for Point Images

Point sources such as diffraction-limited stellar images present a special problem for scanning sensors. This results from having a limited area of positive charge surrounded by a large region at cathode potential. This negative coplanar grid repels the beam, making it necessary to have between five and ten times as much stored charge density in an isolated 50-micrometer spot than in an extended area in order to obtain an equivalent output signal near the threshold illumination level.

The means adopted to overcome this problem in the operation of the image orthicon and image isocon was to bias the target 1 or 2 volts positive before initiating the readout scan. The beam now lands at all points of the raster, and the point image produces the normal signal amplitude relative to the background.

In the Stratoscope, the philosophy adopted for determining the appropriate size of the stellar image at the photocathode of the camera tube is that indicated in figure 4. The 50-percent response point of the isocon modulation transfer, which is at 10 cycles per millimeter for the high target-capacitance

Table 1 lists a number of performance capabilities of a high target-capacitance isocon in the Stratoscope camera. The maximum exposure to charge the target to 80 percent of its full capacity is indicated. This ensures that the transfer characteristic remains linear. When the 50-percent MTF of a 50x50 micrometer point image is used, the maximum signal-to-noise ratio is 138. For a minimum signal-to-noise ratio of 10, the elemental target charge is 400.

In the telescope tests carried out at Princeton, a 16-millimeter-diameter aperture and a focal ratio of f/250 showed that a 9±1 magnitude, diffraction-limited, stellar image could be recorded with a 100-second exposure. Figure 5 shows the monitor display and 16 successive line traces of an oscilloscope display recorded during one scan of the isocon. The star image covers 7 or 8 lines in this instance.

The equivalent condition for a 900millimeter-diameter aperture at f/250 is that a

13±1 magnitude star image would be recorded with a 1-second exposure. The corresponding situation for 103 a G photographic film used in Stratoscope II with a 900-millimeterdiameter aperture and f/100 focal ratio is that a 10±1 magnitude star would be recorded with an exposure of 1 second. The new isocon design, optimized for the slow scan, is expected to give improved capabilities.

Spectrometer Data Recording with the Integrating Television Camera

An astrometric problem for which an integrating television-camera system should be helpful is that of measuring the spectra of faint sources. Preliminary results were obtained by using a spectrometer having a dispersion of 1 angstrom per millimeter attached to the 23-inch refractor in the Princeton Observatory.

By orienting the spectrum so that the dispersion axis is at right angles to the scan lines, it is possible to integrate the output of each scan line and to use this signal as one point in the recording of signal intensity versus wavelength. In this test, it was not convenient to use cylindrical optics to spread the spectral image to cover the full horizontal width of the raster. Instead, only one-fiftieth of the horizontal dimension was illuminated, and a gating circuit was used to select and to integrate the signal from only that portion of the line which contained image information (about three times the spectrum width).

Figure 6 shows the oscilloscope display from the 500 active scan lines covering a 70-angstrom portion of the spectrum of Arcturus (magnitude 0.2). The single isocon readout scan was made following an 8-second exposure. The line integrator was not employed in this instance.

Figure 7 shows the display from a single isocon readout following a 176-second exposure of a third magnitude star with the line integrator accumulating all of the signal that passed through the gate.

The conventional intensity-recording technique for this spectrometer employs

pulse-height discrimination of the output from a photomultiplier. Counts are taken for intervals of 10 to 15 seconds at successive points 1 angstrom apart. The total time is equivalent to the camera tube exposure. The signal-to-noise ratio of this camera tube display is within a factor of three of that