Three companies received a call for quotation for CFH12K-like filters 30cmx30cm (covering the central 36 CCDs of the MegaCam mosaic) and 30cmx35cm (the central square plus two additional CCDs on East and West of the mosaic).
Barr proposed only mosaic filters, while OCLI and
REOSC proposed monolithic filters. REOSC performances are given for
300x300 filters. The Schott substrate they plan to use is not available
in larger size with the adequate quality. Performances are the same for
the two other companies.
I remind you that we won't go through the
option of buying four additional CCDs to make the full 40 CCDs mosaic and
still have four spares. It means tha we'll probably have a 9x4 square mosaic
(and 300x300 filters will be ok).
Here is a table summarizing the answers to
our call and the comments coming from the MegaPrime project (last row of
the table). Wavelengths are in nm...
|
|
|
|
|
| Geometry | mosaic
- 1 central square of 150x150 - 6 rectangles around of 75x150 |
monolithic for broad band
mosaic for narrow band |
monolithic |
| U' Transmission | 75% min. MOSAIC | average > 60% | average > 65% |
| CWL | 350 +/- 3 | 350 +/- 6 | 350 +/- 6 |
| CWL uniformity | +/- 3 | +/- 3 | +/- 3.5 |
| FWHM | 64 +/-3 | 64 +/- 6 | 64 +/- 6 |
| B Transmission | 80% min. MOSAIC | average >60% | average > 80% |
| CWL | 430 +/-5 | 430 +/- 10 | 430 +/- 10 |
| CWL uniformity | +/- 5 | +/- 5 | +/- 6 |
| FWHM | 97 +/- 5 | 64 +/- 6 | 97 +/- 10 |
| V Transmission | 80% min. MOSAIC | average > 70% | average > 75% |
| CWL | 545 +/- 4 | 545 +/- 9 | 545 +/- 9 |
| CWL uniformity | +/- 4 | +/- 4 | +/- 6 |
| FWHM | 88 +/- 4 | 88 +/-9 | 88 +/- 9 |
| R Transmission | 85% min. MOSAIC | average > 85% | average > 82% |
| CWL | 648 +/- 6 | 648 +/- 13 | 648 +/- 13 |
| CWL uniformity | +/- 6 | +/- 6 | +/- 6.5 |
| FWHM | 130 +/- 6 | 130 +/- 13 | 130 +/- 13 |
| I Transmission | 85% min. MOSAIC | average >85% | average > 85% |
| CWL | 830 +/- 9 | 830 +/- 19 | 830 +/- 19 |
| CWL uniformity | +/- 9 | +/- 9 | +/- 9 |
| FWHM | 193 +/- 9 | 193 +/- 13 | 193 +/- 19 |
| Z' Transmission | 70% min. MOSAIC
|
average > 85% | average > 85% |
| Cut. WL | 850 +/- 10 | 850 +/- 20 | 850 +/- 20 |
| Cut. WL uniformity | +/- 10 | +/- 10 | +/- 10 |
| HAlpha Transmission | 70% min. MOSAIC | average > 85% MOSAIC | average > 70% |
| CWL | 657.5 +/- 1 | 657.5 +/- 1 | 657.5 +/- 1 |
| CWL uniformity | +/- 0.5 | +/- 0.5 | +/- 4 |
| FWHM | 10.0 +/- 0.5 | 10 +/- 1 | 10 +/- 1 |
| H Alpha Off Trans. | 70% min. MOSAIC | average > 85% MOSAIC | average > 70% |
| CWL | 642.8 +/- 1 | 642.8 +/- 1 | 642.8 +/- 1 |
| CWL uniformity | +/- 0.5 | +/- 0.5 | +/- 4 |
| FWHM | 10.0 +/- 0.5 | 10 +/- 1 | 10 +/- 1 |
| Price | 0.83 | 2.26 | 1.00 |
| Delivery schedule | > 1 year
(as per CFH12K experience) |
T0 + 4 months | T0 + 5 months |
| Comments | -Mosaic for all filters
|
- very expensive
- mosaic for narrow band |
- monolithic for all filters
- performance still reasonable - narrow band CWL uniformity a problem though |
As a preliminary conclusion, we would go (as seen from the project) for REOSC for the broad filters, and discuss with them what can be achieved for the narrow band filters. The problem for the narrow band filter by REOSC is that the CWL (central wavelength) uniformity is not good enough, moving by +/- 4nm over the whole field. It's much easier to achieve the spec (+/- 0.5) with small elements inserted in a mosaic.
We made some simulations of mosaics. If we can avoid them, we should forget them! There are a lot of potential diffraction sources at the gap level, and the shadow on the focal plane is both large and deep.
Oh, I forgot to mention that we can afford REOSC (it's close to the amount we have in the MegaPrime budget for the filters), and definitely not OCLI. It would make sense to have them making the filters, as they are building the WFC and, more important, coating it. It will be very important to have the U' filter starting where the coating is good enough in the blue (as you know, the broad band coatings have a very sharp transmission curve on the blue side...).