an increase in the gamma-ray integral decreased the frequency. This signal was recorded on Channel 5 of the tape. The KBr crystal was illuminated by a 12-volt incandescent light and its transmission was measured by a photocell whose output current controlled a converter free running at. 30 ko such that the frequency decreased as the transmisaion of the crystal decreased. This signal was recorded on Channel 4 of the recorder. For the two Lil detectors, the output of the corresponding converter was recorded on the tape, axcept for a 150-ysec interval beginning 10 usec after gamma-ray arrival at the canister, when the magnitude of the detector current was recorded directly on the tape. This was necessitated by the inability of the recorder system to record pulses short enough to resolve the rapid variations of detector current expected during the early part of the neutron pulse. To accompliaeh this, both the converter output and Lil detector out- put were brought to a gated mixer. The detector output was taken to Input 1 of the gate of mixer through a high-pass filter to attenuate the slowly changing components of the detector-output current. The converter output was taken to Input 2. (See Figure 2.3.) Input 2 was gated off from plus 10 psec to plus 160 usec, relative to gamma-ray arrival . time, and mixed with Input 1. The result of all this is that the output of the gated mixer consists of direct detector output when it is changing rapidly and converter output when it is changing slowly. Both the Li'l and Lil detectors were treated in this way and recorded on Channels 3 and 2, respectively, of the tape. The sixth channel of the tape was used to record the output of a 32-ke crystal-controlled oscillator, providing an internal time standard. 2.3 MAGNETIC-TAPE RECORDER Instrumentation applications of this kind are faced with the problems of loss of reliable telemetry data transmission because of ionization of the atmosphere, limitation of the data channels available, and the inadequate high-frequency response available with standard RDB voltage-controlled subcarrier oscillators. The purpose of the magnetic recording system was to overcome these problems by providing data storage for six channels of in- formation, time delay between the collection of data and the transmission of data, and reducing the frequency components of the data to, in effect, extend the frequency response of the voltage-controlled subcarrier ogcillator. The magnetic recording system consisted of a two-speed recorder with the electronic components required for recording, erasing, timing, and playback. A block diagram is shown in Figure 2.8. The inputs to the recording system were supplied by encoders and consisted of: Channel 1. A pulse input positive for 10 x.sec, negative for 150 usec and 7 volts amplitude in each direction. Channel 2. The output of a converter binary, which had a no-signal frequency of 500-cycles, square wave, and a full-signal frequency of 50 kc, square wave, with a level of 7 volta, peak to peak. Channel 3. Same as Chamrel 2, except from a diffarent converter binary. Channel 4. The output of a converter binary with a zero input frequency of 30 ke and a full signal frequency of 500 cycles. Channel 5. The output of a pulse-controlled oscillator that had a zero-input frequency of 30 ke and a full-signal frequency output of 500 cycles. Channel 6. The output of a 32-ke crystal-controlled oscillator used to determine the exact speed-reduction ratio between record and playback, as well as an accurate time base for evaluation of the data. 23