• The Kiel Electron Telescope Sensor System
• Monte-Carlo-Simulation of the KET
• KIEL ELECTRON TELESCOPE readme
  • RECORD FORMAT for the KET
  • Caveats:
  • Readme, KET90-96.DAILY
  • Readme, KET90-96.27DAYS
• References
• About this document ...

The Kiel Electron Telescope Sensor System

This figure displays in front the flight spare model sensor unit of the Ulysses Kiel Electron Telescope (KET) and in the back the electronics box. For comparison a five German mark coin is shown to demonstrate the size of the instrument. The KET is part of the Ulysses cosmic ray and solar particle investigation (COSPIN) experiment, which has been described in detail by Simpson et al. [4].

This figure shows the COSPIN sensor units as they are mounted on the spacecraft. The KET (3) is mounted below the Low Energy Telescope (LET), the Anisotropy Telescope (AT), the High Energy Telescope (HET) and the High Flux Telescope (HFT).

This figure shows a schematic sketch of the KET sensor system. D1 and D2 are 0.5 mm thick semiconductor detectors, C1 is an aerogel Cerenkov-detector and A a plastic anticoincidence scintillator. C2 is a lead fluoride Cerenkov-detector and S2 a plastic scintillator. PM1 through PM4 are photomultipliers.

Functionally, the detector system consists of two parts: an entrance telescope and a calorimeter, surrounded by a guard counter A.

The entrance telescope

is composed of a silica-aerogel Cerenkov detector C1 inserted between semi-conductor detectors, D1 and D2. Together with the guard counter A, this combination defines the geometry, selects particles with velocity beta>0.938 and determines the particle charge.

The calorimeter

consists of a 2.5 radiation length lead-fluoride (PbF2) crystal used as Cerenkov detector (C2) and a scintillator S2. In C2 an electromagnetic shower can develop. The number of electrons leaving the PbF2-crystal are counted in S2.

The KET is designed to measure electron, proton and alpha particle fluxes in several energy windows ranging from a few MeV/n up to and above a few GeV/n. The values listed in the following table are based on mean energy losses and geometry of the instrument.

This table summarizes the KET coincidence channels. The coincidence name and trigger condition are displayed in the first two columns. The particles and their corresponding energy range are listed in columns three and four. The geometrical factor and sectorization information of the coincidence channels are in columns five and six.

Monte-Carlo-Simulation of the KET

A treatment of the response functions of particle telescopes, and a number of exact formulae for multi-element telescopes have been given by Sullivan, [5]. However, the determination of the response function of rather complex telescopes like the KET instrument, makes a Monte Carlo simulation mandatory. We assume that the differential coincidence counting rate of a particle telescope can be expressed as:

where dC_i,k is the differential coincidence counting rate in channel i, J_k(E) the flux of particle species k with kinetic energy E, and R_i^k(E) the response function for particle species k in channel i, to be determined by the simulation. To get the counting rate C_i for the channel i we have to integrate over E and sum up for all particle species. In general, the response function may depend on many variables like the angular distribution of the incident particle flux, the location where a particle penetrates a detector etc. Here we assume that R(E) is a function only of kinetic energy, valid for particle fluxes which are almost isotropic over the effective opening angle, and that the simulation properly averages over all other dependencies.

The Monte Carlo simulation was performed with the CERN Library Program GEANT 3 (BRUN et al., [1]). Particles were followed down to a low energy cutoff (electrons and gamma-rays 50 keV, protons 300 keV), once reaching this cutoff the particles were considered to be stopped and to have deposited all of their kinetic energy in the traversed material. The geometry of the detectors, mountings, foils, and the relevant structure material as well as the energy resolution of the readout electronics were accounted for in the simulation.

KET geometry

is shown in the following sketch.

The model of the KET sensor unit in the simulation. Shown are three possible tracks of 120 MeV protons. Two of these tracks are triggering the channel P32 and on of them is counted in P116.

Energy loss distribution:

In addition to the count rates in all coincidence channels, the energy losses in D1, D2 and number of photons in C1, C2 and S2 are transmitted. Particle instruments are calibrated by using a particle beam provided by an accelerator (see SIERKS, [3]). In space, particles of different species, incident directions and energies are observed nearly simultanously. Such an environment cannot be simulated at an accelerator but can be simulated using an appropriate computer model. As an example the following figure shows the simulated energy distribution of isotropic protons and alpha-particles in the semiconductor detectors D1 and D2 triggering the P32 and A32 coincidence channels.

Calculated D1-D2 PHA-distribution (energy loss), using a proton and alpha distribution which is a function of energy and isotropic in direction. The solid lines show the mean energy losses. Marked on these lines are the expected energy losses for 35, 40, 50, 70 and 120 MeV/n. The dotted lines display the electronics thresholds. In comparison the following figure shows a PHA distribution measured in space:

Measured P32-protons and A32-alpha-particle D1-D2 matrix in January 1991. The entries below the dotted line could be identified as background random coincidences (HEBER, [2]). As is discussed in detail in HEBER, [2], the proton and alpha-tracks in that matrix are well described by the Monte-Carlo simulation. One important result of the analysis are the determination of the geometrical factors as a function of energy:

Response functions R^Sim_i(E) for isotropic protons and alpha-particles. The rectangular boxes indicate the response functions expected by using the Sullivan [5] theory. For the low energy channels this theory is a good approximation.

	G/(cm2 sr MeV/n)	E /(MeV)/n	Sigma;E/(MeV)/n
protons K3
S	57.8	81.0	26.1
alpha-particles K33
S	52.7	78.2	24.0
protons K34
S	83.3	190	42
alpha-particles K29
S	23.4	164	25
protons K12
1	230	315	160
2	500	700	390
3	1370	1250	420
4	1500	950	490
alpha-particles K31
A	220	315	150
B	480	700	390
C	1320	1250	420
D	1500	950	490

KIEL ELECTRON TELESCOPE readme

This readme was last modified March 10, 1998 and shall explain the file structure of the KET files provided for the Ulysses Data System (UDS). The previous sections "The Kiel Electron Telescope Sensor System" and "Monte-Carlo-Simulation of the KET" explain the instrument. Only count rates are provided for the UDS, because the geometrical factors may change in future. The geometrical factors provided in the previous section should be checked against the values on our homepage:

All Ulysses data system files (UDS files) have the name

Herein YY is the year (eg. 90) and DOY the day of the year (eg. 365 of year 90 is 31.12.90).

      IMPLICIT REAL(K)
               WRITE(40,'(6I5)') IYEAR,IDOY,IHOUR,IMIN,ISEC,ICOV
               WRITE(40,'(10G11.3)')
     1             K1,K21,K22,K23,K24,K25,K26,K27,K28,P4
               WRITE(40,'(10G11.3)')
     1             K3,K34,K12,K10,K2,K33,K29,K31,K30,K13
               WRITE(40,'(10G11.3)')
     1             K14,K15,K16,K17,K18,K19,K20,E4,K11,K32
               WRITE(40,'(6G11.3)') 
     1             D10,D20,C10,C20,A01

      IMPLICIT REAL(K)
	       WRITE(40,'(I2,1X,I3,1X,I3,1X,8G10.3/10X,6G10.3)')
     1         IYEAR,IDOY,ICOV
     1         K3,EK3,K34,EK34,K12,EK12,K10,EK10,
     1         K33,EK33,K31,EK34,K30,EK30

      IMPLICIT REAL(K)
               WRITE(40,'(4I3,2G13.4)') IYEAR,IDOY,IHOUR,IMIN,E4,EE4