There are several ways in which MIPS behaved differently than was expected, pre-launch. The 24 micron array worked exactly as anticipated, except that it was a factor of ~2 more sensitive than expected. There are only 22 bad pixels in this array; the bad pixel mask (pmask) appears in Figure 6.7 (and on the archive website) and the bad pixels are listed individually in Table 6.5 (the numbering starts with 1,1 in the lower left corner). There is some contamination on the pick-off mirror (POM) that feeds MIPS, and the scan mirror position does not repeat position exactly from campaign to campaign. In both cases, the 24 micron array is the most affected; see section 7.1.1.
The 70 micron array suffers from thermally-activated high resistance in a cable connection that is external to the instrument, but feeds it. This has resulted in a fixed-pattern noise on one side of the array (side B, on the right) that renders that side inoperable. The signal-to-noise ratio and sensitivity are much better on the other side (side A, on the left). Also, one readout on side A (located in the lower right corner of that side, next to side B) is inoperative. The ''good side'' of the array (side A) is not as sensitive as was anticipated. Observers should have planned at least 3 or 4 times redundancy to ensure good data. Bad pixel masks (pmasks) showing these effects are shown in Figure 6.8 and are available in fits format on the archive website.
For data taken in IOC/SV or the first 4 MIPS campaigns, the bias setting of the 70 micron array was set at a level that attempted to increase the sensitivity of both sides of the array, but a consequence was that the stimulator flashes left latents in the data. Data taken in MIPS-5 or later have a different bias setting and therefore weaker stim latents. See section 7.2.2 below for more information.
One block of 5 contiguous pixels in the 160 micron array exhibits anomalous behavior attributable to a thermally-activated short in cabling external to MIPS but well inside the CTA. In data oriented as observers receive it, this block is located in the bottom row; it is the third group of 5 pixels in from the left. The array is not as sensitive as was anticipated. Observers should have planned at least 3 or 4 times redundancy to ensure good data. Bad pixel masks (pmasks) are shown in Figure 6.9 and are available in FITS format on the archive website.
During the early part of the mission, 160 micron signals from K stars were detected to be about a factor of five stronger than expected. Review of the instrument design revealed a weakness in the stray light control that results in a short-wavelength (1-1.6 micron) light leak in this band. See discussion below in section 7.2.6.
Table 6.5: Bad pixels in 24 micron array.
Pixel # (x)
Pixel # (y)
1
127
1
128
2
127
2
128
3
127
3
128
4
128
42
128
52
86
52
87
53
85
53
86
53
87
53
88
54
86
54
87
100
101
101
100
101
101
101
102
102
101
112
25
Figure 6.7: Bad pixel mask for 24 microns; black is bad pixels.
Figure 6.8: Bad pixel mask (pmask) for MIPS-70 data. Pink is bad pixels; red is bad half of the array; purple is non-linearity; green is noisier top row and first column. Black is good pixels.
Figure 6.9: Bad pixel mask (pmask) for MIPS-160 data. The center stripe has no detectors (see SOM for details), and the ''bite'' in the lower right is the inoperative readout.
Table 6.6: BCD bit mask definitions for 24 microns found in bbmsk.fits or ebmsk.fits files.
Bit
Condition
0
Incomplete or questionable row-droop correction (rowdroop)
1
No row-droop correction applied (rowdroop)
2
Hard saturated (satmask)
3
Read-2 correction could not be applied (rowfluxcorr)
4
Soft saturation corrected and slope value replaced by difference value (desatslope and satmask) (Note: replacement only matters for those pixels that are soft-saturated, e.g., those for which the first difference exceeds a threshold).
5
Latent-image flag. (Note: this bit is set for ebmsk but not bbmsk files in photometry mode. It is set for bbmsk files in scan mode.)
6
Droop removed using questionable value (droopop). (Note: if the pixel is saturated or actually missing from the downlink [bits 13 and 14 respectively], the droop correction will be unreliable.)
7
Flat field applied using questionable valule (flatap)
8
Flat field could not be applied (flatap)
9
Radhit detection (radhitmedfilt)
10
11
Pixel masked in pmask - bad hardware state (satmask)
12
Non-linearity correction could not be computed (slopecorr)
