HELIOS FINAL REPORT May 1986 INTRODUCTION This document is the Final Report for the University of Iowa HELIOS Solar Wind Plasma Wave Experiment [E5a] supported under contract NAS5-11279 with Goddard Space Flight Center. This contract provided for the design,construction, testing, integration, and pre-launch activities of our instrumentation and the post-launch operations and data reduction and analysis activities. The objective of this experiment was the investigation of naturally occurring plasma instabilities and electrostatic and electromagnetic waves in the solar wind. To carry out this investigation, instrumentation consisting of a 16- channel spectrum analyzer connected to the electric field antennas, an antenna potential measurement system, a shock alarm system, and supporting electronics was designed, constructed, tested, and flown on the two HELIOS spacecraft. Five units were constructed under this contract: an engineering test model, a prototype, two flight units, and one flight spare unit. HELIOS 1 was launched on December 10, 1974, and HELIOS 2 was launched on January 15, 1976. The University of Iowa HELIOS Solar Wind Plasma Wave Experiments operated perfectly throughout the entire life of the mission which lasted much longer than the planned 18 months originally expected. The useful life of HELIOS 2 lasted until March 21, 1980 when an automatic switching of the onboard power system killed the travelling wave tube amplifier and severely damaged the spacecraft's data handling subsystem. HELIOS 2 was finally commanded off on January 8, 1981. HELIOS I continued to operate until March of 1986 when its command receiver failed. As a part of this contract, we have acquired, processed, and analyzed much valuable data from our experiments during the 11+ years of HELIOS I and 4+ years of HELIOS 2. Below we will describe the scientific objectives of the HELIOS University of Iowa Solar Wind Plasma Wave Experiment, describe the instrumentation designed and constructed to meet these objectives, discuss the significant results obtained from the analysis of the data from this experiment, describe research in progress and future research, and discuss the major problems and anomalies encountered. SCIENTIFIC OBJECTIVES The objective of the University of Iowa HELIOS Solar Wind Plasma Wave Experiment was the investigation of naturally occurring plasma instabilities and electrostatic and electromagnetic waves in the solar wind. In the solar wind a wide variety of electromagnetic and electrostatic wave phenomena can be expected in the frequency range from a few tens of Hz to several hundred kHz. These phenomena include Type III solar radio noise bursts and associated longitudinal electrostatic waves down to the solar wind plasma frequency (from about 20 kHz at I AU to 100 kHz at 0.3 AU), intense (30 mV/m) electrostatic electron plasma oscillations of the type observed with the Pioneer 8 spacecraft, electrostatic waves associated with interplanetary shock waves and solar particle emissions, whistler-mode instabilities related to anisotropic solar wind electron distributions (Tii/Ti >1), and electrostatic ion sound waves. Characteristic frequencies at which emissions or cutoffs might be expected include the upper and lower hybrid resonances, the electron and proton plasma frequencies, and the electron and proton gyrofrequencies. Wave-particle interactions can lead to many important effects in the solar wind. Of particular interest on the HELIOS mission is the investigation of instabilities which make the solar wind behave like a fluid. If the solar wind had no collective effects, the ratio Til/Ti for the protons would be 100 to I or more at the Earth, contrary to observation. It is generally believed that a plasma instability, possibly the whistler-mode, alters the proton pitch-angle distribution. Because of the large radial variation in the solar wind properties (density, magnetic field, TII/Ti, etc.) from the Earth to 0.3 AU, the stability criteria and wave phenomena occurring in the solar wind at 0.3 AU may be quite different compared to near the Earth. The spatial distribution of instabilities may influence the propagation of solar and galactic cosmic rays, slowing down times for super thermal particles can be drastically modified, waves generated in shocks can carry energy and momentum away from the shock region, and many more examples can be cited. Also of great importance is the direct observation of the electron plasma oscillations which are believed to give rise to Type III solar radio bursts through non-linear effects. INSTRUMENTATION The University of Iowa HELIOS Solar Wind Plasma Wave Experiment [E5a] was jointly planned with the University of Minnesota [P. Kellogg, E5b] and Goddard Space Flight Center [R. Weber and R. Stone, E5c] so that the combined experiments would provide complementary measurements. The frequency range of our experiment (E5a) and the University of Minnesota experiment (E5b), nominally 20 Hz to 200 kHz, was chosen to include most of the characteristic frequencies for plasma waves in the solar wind mentioned above. The upper frequency limit (200 kHz) is approximately the maximum electron plasma frequency expected and was also chosen to provide some overlapping frequency coverage with the Goddard Space Flight Center HELIOS Radio Astronomy Experiment (E5c) frequency range (26 kHz to 3 MHz). The lower frequency limit for the plasma wave experiments (20 Hz) was influenced by the large spacecraft spin rate (I rotation per second). Since strong electric field signals are expected at harmonics of the spacecraft spin rate due to the asymmetrical photoelectron sheath around the spacecraft, the lower frequency limit of the experiment has been chosen to be well above the frequencies of the expected spin rate interference. Although the University of Iowa and the University of Minnesota experiments cover the same frequency range, the two experiments are complementary but not redundant. The University of Iowa experiment was designed for high temporal resolution while the University of Minnesota experiment was designed for high frequency resolution. Past experience with plasma wave data from the Alouette, OGO, Injun, and IMP satellites has shown that it is very important to have both electric and magnetic field measurements so that electromagnetic and electrostatic waves can be distinguished. It is expected, therefore, that the AC electric field measurements provided by the plasma wave experiments and the AC magnetic field measurements provided by the HELIOS Search Coil Magnetometer Experiment [F. Neubauer, E4] will provide complementary data for the identification and understanding of certain phenomena. The sensor for the University of Iowa HELIOS Solar Wind Plasma Wave Experiment is the HELIOS electric field/radio astronomy antenna. The electric field antenna and the associated erection mechanism were provided by the spacecraft. The antenna is an extendible cylindrical dipole 32 meters tip-to-tip, similar to antennas used on the Alouette, OGO, IMP, and RAE satellites. In order to reduce the capacitance between the antenna and the spacecraft photo-electron sheath to an acceptable level, an insulating sleeve approximately 2.0 meters long was placed over the antenna where the antenna passes through the spacecraft photo-electron sheath region. The antenna diameter was as small as possible (0-25 inch) in order to minimize the ratio of resistive-to-capacitive coupling to the surrounding plasma. The erection mechanism was designed to minimize the antenna to erection mechanism capacitance. This capacitance does not exceed approximately 30 pf in the extended configuration. Also, special consideration was given to electrostatically shielding the antenna from spacecraft related interference, such as through the erection mechanism, deployment length indicators, etc. Each element of the electric dipole was connected to three preamplifiers which provided signals to the respective University of Iowa, University of Minnesota, and Goddard Space Flight Center experiments. The University of Iowa constructed the housing for the two sets of three preamplifiers and the University of Iowa preamplifiers themselves. The housing for each set of preamplifiers measured 4.4" by 7.4" by 1.5". The total weight for each set of preamplifiers was 0.45 kg. The main elements of the University of Iowa HELIOS Solar Wind Plasma Wave Experiment consist of (1) a differential amplifier, (2) a 16-channel spectrum analyzer, (3) an antenna potential monitor, (4) a shock alarm, and (5) a power supply. These elements are all packaged inside a main electronics box of dimensions 8.4" by 9.7" by 4.3" which weighed 2.44 kg. Details of these experiment elements, the Data Processing Unit (DPU), and the shock memory will be discussed below. We will also describe the commands and the analog and digital housekeeping data and engineering parameters pertinent to our experiment. The differential amplifier provides AC signals proportional to the potential difference between the antenna elements to the measurement electronics. It provides a high order of common mode rejection to reduce the response to interference signals present on both antenna elements. The differential amplifier is followed by a narrow band notch filter which provides approximately 30 dB of attenuation at frequencies of 20 kHz, 40 kHz, and 60 kHz (harmonics of the power supply frequency). The 16-channel spectrum analyzer is the basic element of the University of Iowa experiment. This analyzer provides relatively coarse frequency coverage and rapid temporal resolution with essentially continuous coverage of all frequencies (20 Hz to 200 kHz) and all times (using peak detectors). The lowest eight frequency channels (31.1 Hz to 1.78 kHz) use active filters spaced with four filters per decade of frequency and have sine-wave bandwidths approximately ¤ 19% of their center frequencies. The highest eight frequency channels (3-11 kHz to 178 kHz) use passive filters spaced with four filters per decade of frequency and have sine-wave bandwidths approximately 11% of their center frequencies. A detector and log compressor in each channel rectifies and logarithmically compresses signals from the filter and produces an output DC voltage approximately proportional to the logarithm of the signal intensity in each filter channel. The dynamic range of the compressors extends from 10 [iV to IV RMS. The log compressors provide three outputs each: (1) a peak output, (2) an average output, and (3) a short time constant shock output. The peak output provides a DC voltage (0.00 to 5.10 Volts) proportional to the logarithm of the peak signal observed since the preceding reading. The peak detector is reset after readout by a signal from the DPU. The average output provides an RC average output with either (1) a time constant T which depends on the bit rate and science format mode or (2) a forced short time constant, Ts,. Which time constant is used is determined by the shock alarm channel number. The forced short time constant for the average output is selected when Bit 0 of the shock alarm channel number is commanded to level "I". The short time constant shock outputs provide short-time-constant determinations of the logarithms of the field strengths to the shock mode memory. One channel of short time constant outputs (called the rapid sample output) is continuously available in the normal science data. The channel selected for this rapid sampling is determined by the shock alarm channel number. The output from this channel is sampled 16 times in one spacecraft rotation by the DPU and stored for later readout. On HELIOS 1 the selected channel was equal to the shock alarm channel number. On HELIOS 2 provisions were made to sequentially step through all 16 channels in 16 consecutive readout periods when Bit 1 of the shock alarm channel number was commanded to level "I". A Shadow Blanking Pulse was provided by the DPU to the experiment to open the input to the eight lowest frequency compressors for one eighth of a spin period twice per rotation to reduce the effects of large voltage pulses generated as the antenna sweeps through the spacecraft shadow. This blanking pulse could be eliminated by command. The average potential, Vavg, and the difference potential, Vdiff, between the antenna elements are measured with operational amplifiers in the main electronics package. The frequency range of the Vdiff measurement is from 0.2 to 10.0 Hz. The dynamic range of the Vdiff measurement can be controlled by command to be either t8.O, ¤2.0, ¤0.5, or ¤0.125 Volts. The dynamic range of the Vavg measurement is ¤20.00 Volts. These data are needed for a complete understanding of the antenna operation in the plasma surrounding the spacecraft. A shock alarm signal is to be sent to the spacecraft when the plasma wave experiment senses electric field noise associated with an interplanetary shock wave or other transient events. The shock alarm circuit permits the selection of any one of the 16 spectrum analyzer channels, by command, for shock identification. The threshold field strength required to trigger the shock alarm has 16 levels and is controlled by a four bit binary counter. The shock threshold level, which we call Shock Value E, is reset whenever any shock channel command is received. After reset, the threshold is advanced upward every time the threshold is exceeded. In each case the new threshold level is set to the peak value of the electric field strength observed. This procedure, therefore, results in the storage of the "best" event observed over the shock search interval. Inhibit lines from E5b and E5c prevent the threshold from advancing while the E5b impedance measurement is on or while E5c is calibrating. The shock level value is available on two outputs. The first is a four bit parallel binary word sent to the DPU. The DPU sends this value and other data to the spacecraft data handling system as serial words. The shock level value is also sent in the form of a single analog voltage (0-00 to 5.10 Volts) directly to the spacecraft as an engineering parameters The higher voltages represent higher values of the shock level. This line was included as a backup to provide some useful data via the engineering subsystem if there were ever a major failure in the science data handling subsystem. The experiment power supply provides regulated voltages of ¤6 Volts and ¤12 Volts to the experiment electronics. The power supply operates at a frequency of 20 kHz and is synchronized by a 40 kHz synchronization signal provided by the spacecraft. The total power consumed by our main electronics package and our two preamps is 2.66 Watts at a nominal input voltage of 28 Volts. The Data Processing Unit (DPU) for the Plasma Wave Experiment is separate from the main experiment electronics and was provided by Goddard Space Flight Center through an industrial contractor. The purpose of the DPU is to (1) provide interface connections between the experiment and the spacecraft, (2) provide A/D conversion of all scientific analog data, and (3) provide buffer storage, as necessary, to adapt to the spacecraft data transmission format. Experiment E5a has 50 analog output lines which must be sampled by the DPU. These lines are the 16 spectrum analyzer peak outputs, the 16 spectrum analyzer average outputs, the 16 short time constant shock outputs, the rapid sample output, and the Vdiff Output. For real time data transmission of the normal science data these outputs are sampled in a basic data block consisting of 16 6-bit Peak samples, 16 6-bit Average samples, and 8 8-bit Vdiff or Rapid Sample samples. The 16 Peak spectrum analyzer outputs are sampled simultaneously to within 0.25 seconds. The 16 Average spectrum analyzer outputs are also sampled simultaneously to within 0.25 seconds, but these Average samples are taken at a different time than the Peak samples. The Peak and Average sampling of a given channel are equally spaced in time between successive data blocks. The DPU A/D converter converts these samples into digital words and stores them in a buffer storage for later transfer to the spacecraft telemetry. The DPU also provides a signal to our experiment which resets the Peak detectors immediately after the 16 Peak samples are obtained. 16 Vdiff or Rapid Sample samples are acquired at 16 equally spaced angles covering one complete spin. Two consecutive data blocks are required to transmit either the complete 16 Vdiff or 16 Rapid Sample samples. The spin synchronized sampling alternates between Vdiff and Rapid Sample in each two consecutive data blocks. The time required to transmit each basic data block is determined by the spacecraft bit rate and telemetry format. It varies from 1.125 seconds to 576 seconds. Sampling of data for shock mode storage occurs concurrently with the real time data sampling described above. When a shock or other large transient event occurs, data from both before and after the shock are stored in the spacecraft shock mode memory for later readout. Our experiment has 17 analog output lines which are sampled by the DPU for the shock mode memory read-in. They are the 16 short time constant shock outputs and Vdiff. For the shock mode data transfer these outputs are sampled in a basic data block consisting of 16 6-bit short time constant spectrum analyzer samples and four 8-bit Vdiff samples. The sampling rate for the 16 short time constant spectrum analyzer samples is such that all the samples occur within the time to transmit each basic data block. The times to transmit a basic data block are 0.281 seconds, 0.141 seconds, and 0.070 seconds, respectively, for the three shock mode memory read-in bit rates. The Vdiff samples are equally spaced in time, both within a basic data block and between successive data blocks. Thus the time between Vdiff samples is one quarter the basic data block transmission times just listed above. Spin synchronized sampling of the Vdiff data is not used in the shock data because the sampling rates obtained in the shock mode data are sufficiently high to permit direct real time resolution of spin effects. The onboard shock mode memory contained on each HELIOS spacecraft has a capacity of 219 (524,288) bits. Each shock mode data block consists of 1152 bits acquired from the shock mode experiments (the magnetic field experiments, E2 and E3, the search coil magnetometer experiment, E4, and the plasma wave experiments, E5a and E5b) plus relevant spacecraft and engineering data. 1120 bits from each block (the 32-bit synchronization word is not stored) are stored in the memory which can hold 468 full blocks. The memory is divided into three parts, A, B, and C. Parts A and B each hold 108 blocks while part C holds 252 blocks. Read-in rates to the memory are 4096 bps, 8192 bps, and 16,384 bps. Thus the times to transmit each 1152-bit block are 0.281 seconds, 0.141 seconds, and 0.070 seconds, respectively, for the three bit rates. The shock mode data are read continuously into either part A or part B of the memory which are cyclic. Once the part is filled, new data are written over beginning at the beginning of that memory part. When a shock alarm signal is received from either E2 or E5a (the choice is determined by a spacecraft command) the data are immediately routed to part C of the memory which is filled just once. From then on until another shock alarm signal is received, the shock mode data are read into either part B or part A, whichever one was not being used just before. The above shock mode memory read-in processes continue until commanded to stop. Then the shock mode memory data are transmitted to the ground. Data from part C of the memory are data immediately following the last shock alarm triggering and the data from either part A or part B are the data immediately preceding the last shock alarm triggering. The remaining part of memory data (either part B or part A) contains the data acquired just before the memory was commanded to stop read-in. The time periods covered by the data in part C of the memory are 70.9 seconds, 35.4 seconds, and 17.7 seconds, respectively, for the three shock memory read-in rates. The time periods covered by the data in part A and part B of the memory are 30.4 seconds each, 15.2 seconds each , and 7.6 seconds each, respectively, for the three shock memory read-in rates. The University of Iowa HELIOS Solar Wind Plasma Wave Experiment has five Shock Alarm Channel Number commands, three Vdiff Gain commands, and a Shadow Blanking Override command. Command 006 resets all four bits of the Shock Alarm Channel Number to level "O" and turns off the Shadow Blanking Override Command. Command 371 sets Bit 0 of the Shock Alarm Channel Number to level "1" and also sets the average output of the spectrum analyzers to the forced short time constants Command 027 sets Bit 1 of the Shock Alarm Channel Number to level "1" and on HELIOS 2 causes the Rapid Sample electronics to sequentially step through all 16 channels. Command 350 sets Bit 2 of the Shock Alarm Channel Number to level "1". Command 216 sets Bit 3 of the Shock Alarm Channel Number to level "1". Each of these five commands also resets the Shock Value E to zero. Because the noise level amplitude in each channel is sufficient to raise the Shock Value E one step above zero, the Shock Value E will immediately jump at least one step above zero unless it is inhibited. Thus these commands can be used to generate an artificial shock for a shock mode memory read-in if E5a is controlling the triggering of the shock mode memory (spacecraft command 275) and if E5a is not being inhibited by E5b or E5c during their calibrations. Command 161 resets both Vdiff Gain bits to level "O". Command 237 resets Vdiff Gain bit 0 to level "1". Command 140 resets Vdiff Gain bit I to level "1". Vdiff Gain = 0 is the least sensitive gain setting and covers the range from -8 Volts to +8 Volts. The Vdiff Output is linearly proportional to the input but it has a DC offset such that 0 Volts in equals 2.50 Volts out. Since the 8-bit Vdiff output has a conversion ratio of one data number per 20 millivolts, the data number corresponding to 0 Volts in is 125. Each higher gain step is four times more sensitive than the gain step below it. Command 223, the Shadow Blanking Override command, removes the shadow blanking of the lowest eight spectrum analyzer channels discussed earlier in the spectrum analyzer section. The University of Iowa HELIOS Solar Wind Plasma Wave Experiment has three Analog Housekeeping words and one Digital Housekeeping word in the engineering data transmitted along with the normal science formats plus one Digital Housekeeping word stored in every block of the shock mode memory data. The three Analog Housekeeping parameters are the Shock Value E, the Low Voltage Power Supply Monitor, and the Antenna Average Potential. The Shock Value E (SHOC-E) is found in Subcom Channel B-040. This parameter is an analog voltage proportional to the logarithm of the peak intensity measured by the selected shock channel since reset. The Low Voltage Power Supply Monitor (LVPSM) is found in Subcom Channel C-038. This parameter is an analog voltage equal to 2/3 of the regulated +6 Volt power supply line in the experiment. The Antenna Average Potential (V-AVG) is found in Subcom Channel C-039. This parameter is an analog voltage proportional to the average antenna potential, but it has a DC offset such that 0 Volts in equals 2.50 Volts out (Data number = 125). It measures the antenna potential from -20 Volts to +20 Volts. The Digital Housekeeping word, given the designation DSEDB4, is found in Subcom Channel B-009. Bit 1 of this word is equal to Shock Alarm Channel Number Bit 3. Bit 2 of this word is equal to Shock Alarm Channel Number Bit 2. Bit 3 of this word is equal to Shock Alarm Channel Number Bit 1. Bit 4 of this word is equal to Shock Alarm Channel Number Bit 0. Bit 5 of this word is equal to Vdiff Gain Bit 1. Bit 6 of this word is equal to Vdiff Gain Bit 0. Bit 7 of this word is a logical "I" if the Shadow Blanking Override is on or if the E5a shock alarm triggering is inhibited by E5b or E5c. The Digital Housekeeping word (designated 5E) in the shock mode memory data blocks contains the binary representations of the Shock Alarm Channel Number and the Shock Value E. The Shock Alarm Channel Number is given by bits 0-3 of 5E. Bits 0-3 of 5E are equivalent to bits 3-0 of the Shock Alarm Channel Number. The binary representation of the Shock Value E are given by bits 4-7 of 5E. Bits 4-7 of 5E are equivalent to bits 3-0 of the binary representation of the Shock Value E. MAJOR PROBLEMS AND ANOMALIES ENCOUNTERED Shortly after the launch of HELIOS 1 it was discovered that one of the two antenna elements which make up the electric dipole antenna did not extend properly and was electrically shorted to the spacecraft structure. The resulting antenna configuration was therefore an electric monopole, the spacecraft body and associated booms acting as a ground plane. Although this was not the intended antenna configuration, the effects of this failure were not particularly serious. Monopole antennas of this type have been successfully used on several previous spacecraft plasma wave and radio wave experiments. The primary detrimental effects for the HELIOS 1 Plasma Wave Experiment were a loss of 6 dB in the electric field sensitivity because of the shorter antenna length and an increase in the noise level of the 178 kHz channel by about 25 dB because of an interference signal conducted into the experiment from the shorted antenna. It was also thought that the reduced common mode rejection caused by the asymmetrical antenna configuration would result in larger interference levels, particularly from the spacecraft solar array. However, comparisons with the HELIOS 2 spacecraft, for which both antennas extended properly, showed that the background noise voltages were essentially the same in all except the.178 kHz channel. One disadvantage of a monopole antenna is that the effective length cannot be estimated as well as for a dipole antenna because of the complicated geometry of the spacecraft body. To calculate electric field strengths from the HELIOS 1 data, we assumed that the spacecraft body acted as a perfect ground plane, in which case the effective length is one half of the length of the monopole element, or leff = 8.0 meters. The assumption is justified mainly on the grounds that the spacecraft body and associated booms have a rather large capacitance to the surrounding plasma which should maintain the spacecraft potential essentially constant with respect to the local plasma potential. At higher frequencies, above the electron plasma frequency where plasma effects are not important, the effective length is probably somewhat smaller, by about 15-20%, because of the finite size of the ground plane. Because of the large range of electric field strengths encountered in the solar wind, uncertainties of this magnitude are not considered serious. The calculated electric field strengths also rely on the assumption that the wavelengths are longer than the antenna length. In most cases, specific tests based on spin modulation measurements of the antenna pattern and Doppler shift estimates can be performed to verify this assumption. The power flux for auroral kilometric radiation and electric field strengths for electron plasma oscillations observed by HELIOS 1 as it passed through the Earth's bow shock were in good quantitative agreement with previous similar measurements from the IMP 6 and 8 spacecraft. This showed that the assumptions regarding the effective antenna length were reasonable. Below I kHz, the noise levels on HELIOS I and 2 were noticeably higher than what we had found on IMP 6 and 8. The rapid increase in the HELIOS noise levels below I kHz is caused by interference from the solar array. Because of the spacecraft rotation, large voltage transients, as large as 70 Volts, occur as the solar panels rotate into and out of the sunlight. These voltage transients are coupled to the antenna through the plasma sheath surrounding the spacecraft and produce strong interference over a very broad range of frequencies. This same type of interference also occurs at low frequencies in the IMP 6 and 8 electric field measurements but is more intense on HELIOS 1 and 2 because of the higher spin rate (1.0 rps) in comparison to the spin rates of IMP 6 and 8 (0.083 and 0.4 rps). The sensitivity of the HELIOS 1 instrument was about a factor of two poorer than that of the HELIOS 2 instrument at all frequencies (except at 178 kHz where it was a factor of 16 poorer) due to the factor of two difference in antenna lengths. The sensitivities of both HELIOS instruments were poorer at all frequencies than the IMP 6 and 8 sensitivities mainly due to the fact the IMP spacecraft had longer antennas. The effective antenna lengths on HELIOS 1 and 2 were 8.0 and 16.0 meters, respectively, while the effective antenna lengths on IMP 6 and 8 were 46.2 and 60.9 meters respectively. Since during much of the time, especially during solar minimum, many of the receiver channels remained near their noise levels, it would certainly have been desirable to have had longer antennas on HELIOS. A highly unusual and unexpected interference problem was encountered on HELIOS 1 when the S band telemetry signal was switched from the medium gain to high gain telemetry antenna about 10 days after launch. When the telemetry transmitter was switched to the high gain antenna, a very intense broadband interference occurred, with an increase in noise level of nearly 50 dB in some channels. This interference was also accompanied by a number of other dramatic effects indicating a major disturbance in the plasma around the spacecraft, the most notable being a large increase in the E Z 100 eV electron flux detected by the solar wind plasma experiment and a charging of the electric antenna element to -30 to -40 Volts with respect to the spacecraft structure. Subsequent investigations of these effects have led to the conclusion that the interference and electron heating are caused by a multipacting breakdown of the high gain antenna feed. This breakdown is essentially a transit time resonance for electrons accelerated across the gap in the antenna feed and occurs when the secondary emission coefficient is sufficiently large to cause a rapid buildup of electron density in the gap. Since the resonance condition is highly sensitive to gap spacing, the antenna feed on HELIOS 2 was redesigned with a larger gap spacing to eliminate this problem. As a result of this modification, no similar interference problems were observed on HELIOS 2. Because of the multipacting problem with the high gain antenna, useful plasma wave data could initially be obtained only by using the medium gain antenna. Since operation with the medium gain antenna resulted in a significant bit rate penalty to the other experiments, this mode of operations was quite limited, usually consisting of only a few 8-hour passes per week. Fortunately, shortly after the first perihelion passage, the multipacting interference completely disappeared, and it has not reappeared except for a short period during the second perihelion passage. 'The disappearance of the multipacting breakdown is not understood in detail but is generally - thought to be caused by modifications of the surface properties (secondary electron coefficient) of the high gain antenna feed due to the high temperatures encountered near perihelion or to the wearing down of sharp edges on the feed by the continuous bombardment of electrons. ACKNOWLEDGEMENTS We are extremely grateful to the many individuals and groups at Goddard Space Flight Center without whose support this project could not have been so successfully accomplished. We especially want to thank Dr. James H. Trainor, the U.S. Project Scientist, Gilbert Ousley and Charlie White from the HELIOS Project Office, James Kunst and his integration support team, Earl Beard from the Information Processing Division, and Pat Corrigan and his team from the Operating Satellites Project. 3-0 Plasma Waves Data Processing Unit 3.1 General Description The Data Processing Unit (DPU) for the Plasma Waves Experiment is separate from the main experiment electronics and is provided by GSFC through an industrial contractor. The purpose of the DPU is to (1) provide interface connections between the experiment and the spacecraft, (2) provide A/D conversion of al-1 scientific analog data, (3) provide buffer storage, as necessary, to adapt to the spacecraft data transmission format. 3.2 Data Format (Real Time Modes) Experiment 5a has 50 analog (0-5 volt) output lines which must be sampled by the DPU. In Figure 4 these outputs are PO, pi, ..., P15 , AO, Al, ..., A15, SO, S1, ..., S15, sN, and Vdiff. For real time data transmission these outputs are to be sampled in a basic data block consisting of 256 bits, broken down as follows: (1) 16 6-bit Peak Samples (PO' Pl' ''' P15) 96 bits (2) 16 6-bit Average Samples (AO, All ..., A15) 96 bits (3) 8 8-bit Vdiff or sN samples (spin synchronized) 64 bits BASIC DATA BLOCK SIZE 256 bits The 16 peak spectrum analyzer samples above are sampled simultaneously to within 0.25 seconds. The 16 average spectrum analyzer samples are also sampled simultaneously to within 0.25 seconds, but these average samples are taken at a different time than the peak samples. The peak and average sampling of a given channel are equally spaced in time between successive data block. The DPU A/D converts these samples and stores them in a buffer storage for later transfer to the spacecraft telemetry. The DPU also provides a signal to the experiment which resets the peak detectors immediately after the 16 peak samples are obtained. The 16 Vdiff or SN samples in (3) above occur at sixteen equally spaced angles covering one complete spin. The spin synchronized sampling alternates between Vdiff and SN in each 2 successive data blocks. The SN channel number is determined by the channel number of the shock alarm circuit. The time required to transmit each basic data block is determined by the spacecraft bit rate and telemetry format. The basic data block time and corresponding normal time constant T for the average field strength measurements are given in Table 2 for each spacecraft bit rate and telemetry format. 3-3 Data Format (Shock Mode) Sampling of data for shock mode storage occurs concurrently with the real time data sampling described in the previous section. When a shock occurs data both before and after the shock are stored in the spacecraft shock mode memory for later readout. Experiment 5a has 17 analog (0-5 volt) output lines which are sampled by the DPU for the shock-mode readout. In Figure 4 these outputs are SO' s1, ..., S15, and VDiff. For the shock mode data transfer these outputs are to be sampled in a basic data block consisting of 128 bits, broken down as follows: (1) 16 6-bit Short Time Constant Spectrum 96 bits Analyzer Samples (Sol Si, ..., s15) (2) 4 8-bit VDiff Samples 32 bits BASIC BLOCK SIZE 128 bits The sampling rate for the 16 short time constant spectrum analyzer samples is not critical except that all 16 channels must, of course, be sampled within the time required to transmit each basic data block. The VDiff samples are equally spaced in time, both within a basic data block and between successive data blocks. Spin synchronized sampling of VDiff is not used in the shock data because the sampling rates obtained in the shock mode data are sufficiently high to permit direct (real time) resolution of spin effects. The time to transmit each basic data block is determined by the spacecraft bit rate and telemetry format. The approximate data block times for the shock mode sampling are given in Table 2 for each shock mode bit rate. T0 is the duration of shock mode data stored in part C of the shock memory. 4.0 E5a Commands 4.1 Command 006. Cmd 006 resets all four bits of the Shock Alarm Channel Number to level "O" and turns off the Shadow Blanking Override Command. 4.2 Command 371 Cmd 371 sets Bit 0 of the Shock Alarm Channel Number to level "1". The command also sets the average output of the spectrum analyzers to the forced short time constant. 4.3 Command 027 Cmd 027 sets Bit 1 of the Shock @m Channel Number to level "1". 4.4 Command 350 Cmd 350 sets Bit 2 of the Shock Alarm Channel Number to level "1". 4.5 Co=and 216 Cmd 216 sets Bit 3 of the Shock Alarm Channel Number to level "1". Each of the above five commands also resets the Shock Value E to zero. Because the noise level amplitude in each channel is sufficient to raise the Shock Value E one step above zero, the Shock Value E will immediately jump at least one step above zero unless it is inhibited. Thus these commands can be used to generate an artificial shock if E5a is controlling the input to the shock memory (Cmd 275) and if E5a is not being inhibited by E5b or E5c during their calibrations. 4.6 Command 161 Cmd 161 resets both V-Diff Gain bits to level "0". 4.7 Command 237 Cmd 237 sets V-Diff Gain Bit 0 to level "1". 4.8 Command 140 Cmd 140 sets V-Diff Gain Bit 1 to level "1". V-Diff Gain = 0 is the least sensitive gain setting and covers the range from -8 volts to +8 volts. The V-Diff output is linearly proportional to the input but it has a DC offset such that 0 volts in-equals 2.50 volts out. Each higher gain step is four times more sensitive than the gain step below it. 4.9 Command 223 Cmd 223 removes the shadow blanking of the lowest eight spectrum analyzer channels. 5.0 Analog Housekeeping Data 5-1 Shock Value Level E (SHOC-E) Channel B040 This channel is an analog voltage proportional to the logarithm of the peak intensity measured by the selected shock channel since reset. 5.2 Low Voltage Power Supply Monitor (LVPSM) Channel C038 This channel is an analog voltage equal to 2/3 of the regulated +6 volt power supply line in the-experiment. 5-3 Antenna Average Potential (V-AVG) Channel C039 This channel is an analog voltage linearly proportional to the average antenna potential, but it has a DC offset such that 0 volts in equals 2.50 volts out. It measures the antenna potential from -20 volts to +20 volts. 6.0 Digital Housekeeping Data Channel B-009 (DSEDB4) 6.1 Bit 1 is Shock Channel Number Bit 3 6.2 Bit 2 is Shock Alarm Channel Number Bit 2 6-3 Bit 3 is Shock Alarm Channel Number Bit 1 6.4 Bit 4 is Shock Alarm Channel Number Bit 0 6.5 Bit 5 is V-Diff Gain Bit 1 6.6 Bit 6 is V-Diff Gain Bit 0. 6.7 Bit 7 is a logical "1" if the Shadow Blanking Override is on or if the E5a shock alarm system is inhibited by E5b or E5c. 7.0 Housekeeping Data in Format 6 The 5E word in Format 6 contains the binary representations of the Shock Alarm Channel Number and the Shock Value E. The Shock Alarm Channel Number is given by bits 0-3 of 5E. Bit 0 of 5E is equivalent to Bit 3 of the Shock Alarm Channel Number. Bit 3 of 5E is equivalent to Bit 0 of the Shock Alarm Channel Number. Bit 4 of 5E is equivalent to Bit 3 of the Binary representation of the Shock Value E. Bit 7 of 5E is equivalent to Bit 0 of the binary representation of the Shock Value E. 8.0 Thermistor Locations 8.1 E5a Box The thermistor in the E5a experiment box is located approximately in the middle of the box on the shock alarm board. 8.2 E5 Preamps The thermistors in the preamps are attached to the walls of the connector compartments. TABLE 1 Filter Center Bandwidth Type of Short Number Frequency Filter Time Constant 0 31.1 Hz +-19% Active 0.2 sec. 1 56.2 +-19% Active 0.1 sec. 2 100.0 +-19% Active 0.05 sec. 3 178.0 +-19% Active 0.05 sec. 4 311.0 +-lg% Active 0-05 sec. 5 562.0 +-19% Active 0.05 sec. 6 1.0 KHz +-19% Active 0.05 sec. 7 1.78 +-19% Active 0.05 sec. 8 3.11 +-11% Passive(UTC) 0.05 sec. 9 5.62 +-11% Passive(UTC) 0.05 sec. 10 10.0 +-11% Passive(UTC) 0.05 sec. 11 17.8 +-11% Passive(UTC) 0.05 sec. 12 31.1 +-11% Passive(UTC) 0.05 sec. 13 56.2 +-11% Passive(UTC) 0.05 sec. 14 100.0 +-11% Passive(UTC) 0.05 sec. 15 178.0 +-11% Passive(UTC) 0.05 sec.