Title: MIPS 24, 70 and 160 µm Internal Stim Calibration

Analysis by: Doug Kelly (6 June 2001)

(Revised on 3 March 2003; April 16, 2003)

Summary: The following is a summary of the relationship between brightness of the MIPS stimulators and the commanded stim DAC values. At 24 µm we only measured brightness vs. DAC on CE-1, whereas for the Ge stims we have more extensive data (although taken with the stims in manual mode, rather than in flash/auto mode). These measurements were taken during CTA testing. The 70FldA and 70FldB stim brightnesses are equivalent, although the illumination patterns are a bit different. The median 160FldA stim brightness is about 0.65 +/- .05 times the brightness of 160FldB. The Ge stims are covered first; the 24 µm stims are addressed near the end of the report.

70 and 160 µm Flood Stimulators

The internal stim brightnesses are controlled by patchable constants (DAC settings) in the flight software. The nominal goal is to define a set of default values for the stim DAC settings and to use the same values throughout the mission. In practice, the values we define at the start of the mission will almost certainly have to be modified as we learn how the responsivity of the arrays behaves in a radiation environment, as we better define typical background brightnesses, and as we determine how best to deal with issues of latency and saturation.

The data used for calibrating the 70um and 160um Flood stim brightness came from the 70um and 160um full-well linearity tests run on day 01-230. The data were reduced using mips_sloper with the linearity correction turned off, with the cosmic ray rejection threshold set to 5-sigma, and with the high saturation thresholds set to 62500 at 70um and 61000 at 160um. 50 DCEs were collected at each DAC setting. The first 10 DCEs were ignored, and averages were determined over the remaining 40 DCEs, first to determine a mean slope image and then to calculate the mean value over the array. All data were taken with 70FldB (identical in brightness to 70FldA) and 160FldB (about 50% brighter than 160FldA).

Since the rapidly changing responsivity of the Ge detectors is an unavoidable reality, these curves should be viewed only as an indication of how stim brightnesses vary with DAC setting, and not much weight should be given to the actual calibration of the stim brightnesses.



The 70um flood stim datasets are summarized here, and the data are shown graphically below. The first plot shows the brightnesses on a linear scale, while the second plot uses a log-linear scale. The data include both 10s and 3s exposures and show both the mean value for the array and the array maximum. The maximum comes from a bright illumination spot about halfway between the center and the lower left corner of the array (see an image of the flood stim illumination pattern below).



The 160um flood stim data can be found here, and the data are shown graphically below. The first plot shows the data on a linear scale and the second on a log-linear scale. There are three data sets, one with 10s DCEs, one with 3s DCEs, and one with 3s DCEs and rapid resets (every 1.5s). The plots show both the mean and the maximum value on the array. An image of the flood stim illumination pattern is shown below.


The brightnes of the stims appears to be quite repeatable day-to-day, and between CEs. Below are figures showing stim brightness vs. DAC setting for experiments run on different days and different CEs. The first figure compares results for the 70FldB stim, with the data from day 01-226 taken using CE1, and those from 01-230 with CE2. Following that is a similar figure for the 160FldB stim. Each experiment (70 and 160) on day 226 followed an anneal, Vreset optimization, 1/2 hour settling period, 1hour of readnoise data, and one hour of dark data (w/ stimflashes). On day 230 each experiment immediately followed an anneal, Vreset optimization, and 1/2 hour settle.



Images of the illumination patterns can be found at the following links for the 160um flood stim and the 70um flood stim, . Note that these old 70um images are rotated 90 degrees ccw w.r.t. the orientation eventually adopted for the 70um images.

Histograms of the 70um flood stim illumination patterns are shown below.


24 µm Flat and Flood Stimulators

The nomenclature for the 24 µm stims can be confusing. There are 4 stims total, 2 being controlled exclusively by CE-1, and the other 2 being controlled exclusively by CE-2. For either pair, the first stim is referred to as either the A stim or the Flat stim (i.e. "A" = "Flat"). The second stim is referred to as either the B stim or the Flood stim (i.e. "B" = "Flood"). This might not be so bad if, in fact, the "Flood" stims didn't provide a more even illumination pattern than the "Flat" stims. Sigh.

Nominal 24 µm stim DAC Settings


Median Brightness (DN/s) vs. CE, Low DAC Setting.

24 µm Flat / A: Brightness (DN/s) vs. DAC (CE-1)
The illumination pattern has a strong horizontal gradient in the displayed
images. Brightness levels from the left, center and right of the images are
given to "illuminate" the gradient.

24 µm Flood / B: Brightness (DN/s) vs. DAC (CE-1)
The Flood stimulator has a relatively even illumination pattern, so only
a single value (median) is given vs. DAC.



Back to log sheet