PDS_VERSION_ID = PDS3 LABEL_REVISION_NOTE = "2006-10-16,Thierry Semon(UoB)" RECORD_TYPE = STREAM RECORD_BYTES = 80 PRODUCT_ID = CALIBRATION_DESCRIPTION PRODUCT_CREATION_TIME = 2006-10-16T08:00:00.000 DATA_SET_ID = "GIO-C-NMS-4-HALLEY-V1.0" DESCRIPTION = "NMS CALIBRATION DESCRIPTION" END Calibration and data analysis (Meier, 1992) ------------------------------------------- The calibration of the flight unit was performed with the toggle 0 mode with a N2+-ion beam at 700 eV beam energy and a N2-neutral beam at 700 eV. This calibration leads to a mean detector gain for the complete detector. In order to expand this calibration to all species the following correction factors have to be applied: _ A relative correction factor for the individual pixels is given by the deviation of the local gain to the mean gain. _ Because the gain decreases at high count rates a correction factor for the non-linearity was introduced. _ For other ions than N2 species dependent correction factors have to be applied: o For neutrals the ionization cross section of the electron bombardment in the ion source is species dependent. o The instrument transmission is a function of the ion mass and the toggle mode. A transmission correction factor des- cribes the difference of transmission for a certain mass to the transmission for N2 in toggle mode 0 (standard). o Not only the transmission but also the detector gain depends on the ion species. The detector gain has to be corrected for the species dependent yield of the detector. The different factors are discussed below in detail (from Meier,1992) Mean detector gain Guni(GS): --------------------------- The gain of the detector depends on the high voltage between front and back of the microchannel plate. This voltage assumed 15 discrete values (gain steps, GS) between hex 1 (gain ~10^-2) and hex F (gain ~10^6). For each of these values a mean gain factor (unified gain: Guni(GS)) was determined for the bare detector. Relative correction factor for the individual pixels ---------------------------------------------------- Each pixel has an individual gain which is taken into account by the relative gain REL(GS, i) (GS: gain step, i: number of pixel). Correction factor for the non-linearity (NL(GS,i,AN,T)): ------------------------------------------------------- The detector gain depends on the count rate AN(i) and decreases with increasing count rate. The correction factor NL(GS,i,AN,T) depends on the count rate, the gain step GS, the pixel number i and the detector temperature T. Effective Gain -------------- The effective gain Geff (GS,i,AN(i),T) of each pixel is calculated by: Geff(GS,i,AN(i),T)=REL(GS,i)*Guni(GS)*NL(GS,I,An,T) [1] Detector background ------------------- The background is the result of residual gas in the instrument. The thermal noise is already filtered by the DPU. The background is measured and transmitted separately and is deducted from the spectra. Mapping of the mass lines ------------------------- The counts are added over five pixels around the center of the mass line. The overlap of adjacent peaks is taken into account. The corresponding uncertainty is ideally less than 1%, in general less than 5%. Toggle correction krel(M,Toggle) -------------------------------- Periodically the mass spectra were shifted on the detector (toggle mode). The main reason was the fact, that due to the four MCP's which are aligned side by side, some masses fall onto the gap in between and could therefore not be evaluated. Due to the grids in front of the detector, a systematic difference in the peak height between shifted and unshifted spectrum can be observed. As a function of mass and instrumental mode (neutral or ion mode) the peak height of the two types of spectra were adjusted. The toggle correction can be expressed by a 3rd degree polynomial: Krel(M,Toggle0)=-2.873*10^-5*M^3+2.532*10^-3*M^2-8.109*10^-2*M+1.9166 Krel(M,Toggle1)=-3.368*10^-5*M^3+3.012*10^-3*M^2-9.616*10^-2*M+1.9293 [2] GS/HG-effect ------------ To enlarge minor peaks the NMS records every second spectrum at a higher MCP voltage. The difference corresponds to three gain steps. Again a systematic offset could be observed. A scaling factor was derived to determine the true particle counting rate of the regular and of the enlarged spectrum. The GS/HG effect results from a gain shift during the time between the laboratory calibration phase and the flyby at comet P/Halley. Except for the detector number 3 the corresponding factor is compatible with 1. Relative detector sensitivity ----------------------------- The detector sensitivity does not only depend on the MCP voltage applied but also on the incident velocity and chemical structure of the projectile. This can be described by a factor Y(X+, E, MHV) whereby X+ is the ion species, E is the initial energy plus the acceleration in the sensor and MHV is the detector high voltage (or gain step). A list for different species can be found in the archive (specific detector yield data) normalized to the flyby velocity of 68.37 km/s and to the gain step hex F(used for neutrals) and hex C(used for ions). This relative detector sensitivity is 1 for N2+ at 700 eV. Absolute sensitivity -------------------- P(M) is the experimentally determined count rate which was normalized to the uppermost gain step (hex F) and which is corrected with respect to the gain and the non-linearity. P(M)=Sum (An(i)*Guni(hexF)/ Geff(GS,I,An(i),T)) where the sum is over 5 pixels i which contribute to Mass M [3] AN(i) is the effective count rate after separation of mass peaks and deduction of background. Absolute ion sensitivity ------------------------ The absolute sensitivity for ions Sion gives the relation between the count rate P(M) and the effective ion density n(X+) of a certain ion species X+. n(X+)=P(M)/(krfel(M,Toggle)*Y(X+,E,MHV))*Sion [4] Sion is given by the following expression: Sion = 1/(t*Guni(hexF)*Q*e*A*tau*v = 8.05*10^-5 cm^-3 [5] with: t = 0.161 Transmission of N2+, 700 eV, Toggle 0 (standard) Guni (hex F) = 2.104 106 Unified gain at GS = hex F Q = 2.95 1013 C-1 Absolute sensitivity e = 1.6022 10-19 C elemental charge A = 0.36 x 5.0 mm2 Area of the entrance slit tau = 62.96 ms Integration time v = 68.373 km/s Ion velocity relative to the spacecraft Absolute neutral sensitivity ---------------------------- Sneutr connects the count rate to the density of the neutral gas: P(M)/(krel(M,Toggle)*Y(X(M)+,E,MHV))*Sneutr(X,M,IE)=n(X) [6] with Sneutr(X,M,IE)=S*pi*a0^2/(sigmatot(X,IE)*delta(X,M,IE)) [7] X is the neutral species. X(M)+ is the ion with the mass M, which is created by the electron bombard- ment from X. Sneutr(X,M,IE) is the neutral sensitivity, if ions of the neutral species X created with the electron energy IE generate the mass line M. sigmatot(X,IE) is the ionization cross section for a species X with the electron energy IE. The cross section is given in units of pi*a0^2 (= 8..80 10^17 cm^2, a0 is the Bohr radius). The factor delta(X;M;IE) is the fraction of the total ions created from the species X which fall onto mass M. S has been determined for the two electron energies 18 eV (LeV) and 90 eV (HeV)as: SHeV = 230 cm^-3 SLeV = 362 cm^-3 All archived data were taken in the HeV mode The following data analysis has been performed on the archived data: -------------------------------------------------------------------- - The mass peaks have been separated - The background has been subtracted - The count rates have been normalized to the gain step hex F according to [3] - The ion density has been calculated according to [4], but with a specific yield for N+2. - The neutral densities have been calculated according to [6] and [7] but with a specific yield for N2. It is the task of the user of the archive to calculate species dependent yields according to the assumptions made for the composition of the cometary coma, that is to choose the correct relative detector sensitivity (for ions and neutrals) and the ionization cross section and fraction for a specific ion in the case of neutrals. Correlated and uncorrelated uncertainties of the data ===================================================== Uncorrelated uncertainties due to counting statistics: ------------------------------------------------------ with P=Number of counts, C Calibration factor, n density in cm-3 (see formula [4]) z It follows: In the case of neutrals C~100. It then follows: , For ion densities C~10-4, that means Correlated uncertainties ------------------------ z Due to an unexplained difference between mass and energy analyzer absolute ion densities have an uncertainty of a factor 2.25. z The detector yield has been determined as a function of the ionized species with an uncertainty of ~2 % (1 digit) z The overlap of neighboring peaks gives an additional mean uncertainty of < 3 % (Meier, p15) z Other uncertainties which are independent of the mass give an additional contribution of 10 %. (Toggle Effect, Unified Gain, background correction, etc.). For absolute densities the correlated uncertainty amounts to app. 15 % for neutrals and >50 % for ions. For relative densities (e.g. mass18 /mass 19) the correlated uncertainty is ~10 %.