James Webb Space Telescope, information, updates and discussion Options

[Redstone]

The manufacture of the JWST mirror blanks has now been completed.

Despite this milestone, the fate of JWST is still somewhat precarious, because although the scientific bang from the telescope is expected to be huge, the bucks required have increased to a staggering $4.5 billion. A Space.com article on the squeeze in NASA's space-based astronomy plans gives some background.

The JWST home page can be found here.

The Space Telescope Science Institute, which runs Hubble, also has a site here. As does ESA.

[helvick]

JWST Partner's Workshop - Dublin 11 June 2007 (Part III of III)

NIRSpec – Peter Jakobsen EADS Astrium
http://www.stsci.edu/jwst/docs/flyers/Near...ograph_2400.pdf

"A Pretty Picture is not Enough" or Imagery is Astronomy but Spectroscopy is Astrophysics.

NIRSpec is a multi object dispersive spectrograph that uses a MEMS shutter array to enable it to take up to 100 spectral samples concurrently. Sampling in one of three resolutions (R=100, R=1000, R=2700) using 2 x 2048x2048 HgCdTe Rockwell sensors. The twin detectors do not abut perfectly so there is a detector gap in the layout which mostly just causes targeting complications for R=100 sampling but causes dropout in the middle of spectra for many R=1000 samples and all R=2700 samples. These dropouts will require re-shooting the target using a different array location to recover the dropout regions.

In addition to its MOS mode it also supports an Integral Field Spectrograph mode and a classical long slit spectrograph mode.

Physically it's a monster 185kg mass measuring ~1.8mx1.4x1m 2m on a side. The prism\refraction wheel is sufficiently massive that it acts as an (undesirable) reaction wheel for the observatory.
Internally it uses an all silicon carbide reflective optics (14 reflection) optical path.
FOV is 3.4' x 3.6' with a 0.2milliarc second nominal slitwidth.

The MEMS shutters consist of 4 arrays of 365x171 micro shutters. In operation targets selection requires the opening of three shutters in a line perpendicular to the spectral spread direction – the central shutter covers the target and the shutters on either side are used for background removal. Combined with the fact that 2nd and 3rd order effects prevent the use of multiple collinear (in array terms) targets the effective maximum number of concurrent samples is ~100.

The MEMS arrays make this a very powerful instrument however manufacturing the array is extremely difficult and it is clearly pushing the limits of today's micro mechanical manufacturing expertise. The arrays are individually electrostatically latched open and reset in bulk magnetically (to closed). Manufacturing challenges mean that the arrays have clear salt and pepper effect flaws – Peter Jakobsen did not say what the error rate was but it was very obvious in the sample images he displayed. I would estimate that it was probably in the order of 2-3%. In general the fail closed flaws are more frequent than fail open flaws and happily fail-closed shutter flaws result only in aiming\planning problems as the only effect is that certain individual shutter locations cannot be used. Fail open flaws are much more problematic as any open shutter could contaminate a spectral sampling anywhere else on the same horizontal line of the array and so fail-open flaws result in the loss of an entire row of sampling locations from the MOS array (actually they reduce the effectiveness of the row immediately above and below also). Fortunately fail-open flaws can be converted to fail-closed flaws pre flight so at launch there should be no fail open flaws. Peter Jakobsen gave no indication of the expected reliability of the array. The detectors and the MEMS array are supplied by NASA.

In addition to the MEMS arrays there is an IFU image slicer with a a 0.1" resolution of 3"x3" and 5 fixed non interfering slits of width 0.1", 0.2" and 0.4".

More details here. http://www.eso.org/gen-fac/meetings/3Dspec...rribas_JWST.pdf Page 9 shows the MEMS shutter array, IFU and fixed slit layouts and the direction of dispersion relative to the detector layout. The detector gap is not shown but corresponds approximately to the mid point gap between the MEMS shutters.

Peter Jakobsen closed out with some additional comments about exoplanet spectroscopy. My notes say that he said the challenge there is that they will need to get the S/N ratio past 10^4, specifically capturing >10^10 photons into the detector in <1hr time frame. This difficulty arises (in part) because NIRSpec does not have a coronograph.