PDS_VERSION_ID = PDS3 RECORD_TYPE = STREAM OBJECT = TEXT PUBLICATION_DATE = 2005-02-24 NOTE = "Description of MB velocity calibration" END_OBJECT = TEXT END Velocity calibration The interpretation of acquired Moessbauer (MB) spectra is impossible without knowing the drive velocity precisely at any given time. Moessbauer drive velocity calibration for MIMOS II is rather straightforward and done in three different ways, thus ensuring redundancy. Prior to flight each individual drive system was calibrated by measuring in backscattering mode an alpha-iron foil standard. A maximum drive velocity was preset by firmware. Fitting the acquired Moessbauer spectrum using the well-known parameters of the alpha-iron foil then yielded the real velocity. This procedure was repeated at different temperatures. During the mission, the magnetite CCT (Compositional Calibration Target) was measured in several runs to verify the functionality of MIMOS II. The well-known Moessbauer parameters of magnetite can be used for velocity calibration again. These kinds of measurements have been done already in the lab with the flight units as a function of temperature, to be used as reference for the measurements on Mars. The primary method for velocity calibration is the internal reference target and detector configured in transmission measurement geometry. The reference target is a mixture of alpha-Fe(0) (metallic iron, 30% enriched 57Fe) and alpha-Fe2O3 (hematite, 95% enriched 57Fe), and its Moessbauer spectrum is measured automatically during each backscattering measurement. Each component of the reference target has well-known Moessbauer parameters, so that fitting of reference spectra enables velocity calibration for each individual measurement done in backscatter geometry, ensuring that the actual drive velocity is always well defined, regardless of prevailing environmental conditions. Note: The text above is an excerpt of Klingelhoefer et al. (2003), JGR, Vol. 108, No. E12, 8067, doi: 10.1029/2003JE002138. Taking the differential signal into account Instrument performance tests with the JPL-produced flexprint cable for connecting MIMOS II electronics and sensor head were done late during the ATLO period. The length of the cable and therefore a large impedance in the grounding wire had unforeseen effects on the instrument performance: The drive system started "ringing" at low temperatures (~ 200 K). To avoid this problem during Mars surface operations, the amplification in the feedback gain of the drive system was reduced by about a factor of 10 in case of Opportunity and a bit lower on Spirit. The effect on drive system performance is negligible on Spirit's MB sensor head, but must be taken into account on Opportunity's MB sensor head. Prior each MB integration on Mars, sometimes also after the integration, the drive error signal or differential signal is measured. It is defined as Nominal_Drive_Velocity - Real_Drive_Velocity = Differential_Signal. Thus to attribute the real velocity to corresponding channels in the MB spectra one needs to subtract the differential signal from the nominal velocity. The nominal velocity is a triangular waveform which starts and ends at negative v_max - channel 1 and 512, respectively - and reaches positive v_max between channels 256 and 257. However, this channel assignment (folding point) is subject to small variations and needs to be determined using a variable in fitting routines. The measured differential signal as contained in raw data from the instruments is biased and needs to be centred on a zero line. The measured offset is completely due to artificial offsets of the ADC system. The mean values of the differential signal and of the real velocity are exactly zero. The formula Differential_Signal_centred(x) = Differential_Signal(x) - Sum{Differential_Signal(i)}/512 yields a centred differential signal. Here x denotes a particular channel number, the sum goes over all channels i from 1 to 512. The differential signal comes in counts per channel. The counts are proportional to drive velocity. The proportionality factor to convert counts into a velocity in mm/s is 0.008 +/- 0.002 per cycle. Differential signals are almost always taken as a sum over 250 cycles so the conversion factor becomes 0.008/250. Effects of diurnal temperature changes Not only rocks and soils to be investigated on Mars, but also the MER MB sensorheads are subject to the Martian diurnal temperature changes. The behavior of some of the electrical and magnetical components of the MB drive system is temperature dependent. Although the value for maximum drive velocity preset by the software does not change during one measurement covering a range of temperatures, the real maximum drive velocity decreases with decreasing temperature in a linear fashion. From the third release onwards we provide individual velocity scales for each temperature window, taking this temperature dependence into account. Ideally different temperature windows should be fitted individually. However in the case of low counting statistics summing of temperature windows to enhance spectral features might become a desirable option. In this case we recommend using the velocity for the "average" temperature window, i.e. the temperature window covering the center of the encountered temperature range and/or with the longest integration time. Summing of spectra will result in broadened lines. Therefore Moessbauer parameters derived from summed spectra should only be evaluated in context of long term measurements on other units. However, summing of spectra has only a negligible effect on the derivation of area ratios and Moessbauer parameters of individual phases. Disclaimer: All calibrated released data are subject to change according to the analysis progress by the MB team. In the first release we used a velocity scale derived from laboratory calibrations of the drive system, which was confirmed by initial measurements on Mars. The differential signal, an average for each rover, was taken into account for both rovers. Drive velocity and, in particular, the differential signal are subject to small variations due to temperature changes and different orientations of the MB sensor head. In general these variations have very little if any effect on the derivation of MB parameters. However, we encourage users of the released data to perform their own calibration for individual data sets as described above, using available reference spectra and differential signals.