Principal Investigator:     R.E. Vogt
 
      The following section on instrumentation has been extracted
      from the NSSDC documentation for the Voyager Cosmic Ray
      Subsystem (Reference_ID = NSSDCCRS1979).
 
      Instrument Overview
      ===================
 
      As its name implies, the Cosmic Ray Subsystem (CRS) was
      designed for cosmic ray studies [STONEETAL1977B]. It
      consists of two high Energy Telescopes (HET), four Low
      Energy Telescopes (LET) and The Electron Telescope (TET).
      The detectors have large geometric factors (~ 0.48 to 8
      cm^2 ster) and long electronic time constants (~ 24
      [micro]sec) for low power consumption and good stability.
      Normally, the data are primarily derived from
      comprehensive ([Delta]E[1], [Delta]E[2] and E) pulse-
      height information about individual events. Because of the
      high particle fluxes encountered at Jupiter and Saturn,
      greater reliance had to be placed on counting rates in
      single detectors and various coincidence rates. In
      interplanetary space, guard counters are placed in
      anticoincidence with the primary detectors to reduce the
      background from high-energy particles penetrating through
      the sides of the telescopes. These guard counters were
      turned off in the Jovian magnetosphere when the accidental
      anticoincidence rate became high enough to block a
      substantial fraction of the desired counts. Fortunately,
      under these conditions the spectra were sufficiently soft
      that the background, due to penetrating particles, was
      small.
 
      The data on proton and ion fluxes at Jupiter were obtained
      with the LET. The thicknesses of individual solid-state
      detectors in the LET and their trigger thresholds were
      chosen such that, even in the Jovian magnetosphere,
      electrons made, at most, a very minor contribution to the
      proton counting rates [LUPTON&STONE1972]. Dead time
      corrections and accidental coincidences were small (< 20%)
      throughout most of the magnetotail, but were substantial (>
      50%) at flux maxima within 40 R[J] Of Jupiter. Data have
      been included in this package for those periods when the
      corrections are less than ~ 50% and can be corrected by the
      user with the dead time appropriate to the detector (2 to
      25 [micro]sec). The high counting rates, however, caused
      some baseline shift which may have raised proton thresholds
      significantly. In the inner magnetosphere, the L[2]
      counting rate was still useful because it never rolled
      over. This rate is due to 1.8- to 13-MeV protons
      penetrating L[1] (0.43 cm^2 ster) and > 9-MeV protons
      penetrating the shield (8.4 cm^2 ster). For an E^-2
      spectrum, the two groups would make comparable
      contributions; but in the magnetosphere, for the E^-3 to
      E^-4 spectrum above 2.5 MeV [MCDONALDETAL1979], the
      contribution from protons penetrating the shield would be
      only 3 to 14%.
 
      The LET L[1]L[2]L[4] and L[1]L[2]L[3] coincidence-
      anticoincidence rates give the proton flux between 1.8 and
      8 MeV and 3 to 8 MeV with a small alpha particle
      contribution (~ 10^-3). Corrections are required for dead
      time losses in L[1], accidental L[1]L[2] coincidences and
      anticoincidence losses from L[4]. Data are given only for
      periods when these corrections are relatively small. The
      energy lost in detectors L[1], L[2] and L[3] was measured
      for individual particles. For protons, this covered the
      energy range from 0.42 to 8.3 MeV. Protons can be
      identified positively by the [Delta]E vs. E technique,
      their spectra obtained and accidental coincidences greatly
      reduced. Because of telemetry limitations, however, only a
      small fraction of the events could be transmitted, and
      statistics become poor unless pulse-height data are
      averaged over a period of one hour.
 
      HET and LET detectors share the same data lines and pulse-
      height analyzers; thus, the telescopes can interfere with
      one another during periods of high counting rates. To
      prevent such an interference and explore different
      coincidence conditions, the experiment was cycled through
      four operating modes, each 192 seconds long. Either the
      HETs or the LETs were turned on at a time. LET-D was cycled
      through L[1] only and L[1]L[2] coincidence requirements.
      The TET was cycled through various coincidence conditions,
      including singles from the front detectors. At the expense
      of some time resolution, this procedure permitted us to
      obtain significant data in the outer magnetosphere and
      excellent data during the long passage through the
      magnetotail region. Some of the published results from this
      experiment required extensive corrections for dead time,
      accidental coincidences and anticoincidences
      ([VOGTETAL1979A], [VOGTETAL1979B]; [SCHARDTETAL1981];
      [GEHRELSETAL1981]). These corrections can be applied only
      on a case-by-case basis after a careful study of the
      environment and many self-consistency checks. They cannot
      be applied on a systematic basis and we have no computer
      programs to do so; therefore, data from such periods are
      not included in the Data Center submission. The scientists
      on the CRS team will, however, be glad to consider special
      requests if the desired information can be extracted from
      the data.