Schott filter glass for blocking comes in square 165mm sheets, so that is about the maximum size of the filters. Ours are actually trimmed down to 162mm, which leaves very little around the clear aperture, but that was the manufacturer's choice. Filters can also be thicker or thinner, but the camera focus will shift, which may result in the focus not being reachable by the current dewar focus mechanism. The IMACS mosaic has "wing chips", which are intended to be used for closed-loop flexure control. If you want to use these, the filter needs to be larger. So, imagine that the clear aperture is a square with the corners radius'ed off by a circle. Then the filter sizes for direct imaging are (edge to edge and diameter corner cutoff (all in mm): width/no-wing width/with-wing max-diameter (mm) (mm) (mm) 143 156 188 In short, if you want the smallest filter possible, a square filter with a clear aperture of 143mm-square would work. What we did was have a high spec for the 143mm-square aperture, and requested that the filter transmit out to a clear aperture of 156mm for the wing chips, but with no tight specifications in the extra region. Our filters cost anywhere from $4K to $9K, depending on the band. Other types of filters have come in as high as $14K. possible filter manufacturers are: Barr Associates http://www.barrassociates.com/ Omega http://www.omegafilters.com/ Custom Scientific http://www.customscientific.com/ Andover http://www.andovercorp.com/ Custom is probably cheapest. Omega has never given me a quote. Barr is the only company that can measure inhomogeneity of the filter substrate, Barr made all of our filters so far. I have never tried to quote from Andover, for no particular reason. For filters which do not require a blocking substrate (some interference filters), almost any optical manufacturer can quote. Below are two sample requests for quotation. The first for a blocking- substrate filter, the second for order-separating interference filters. ----------------------------------------------------------------------- Dear XXX, I would like to get a quote on a filter to be used in the IMACS astronomical spectrograph, which will be placed on the Magellan telescopes at Las Companas, Chile. Barr Associates has already made eight filters for this instrument. The specifications are as follows: Physical size: 165 mm x 165 mm square (nominal) filter may be edged smaller Edge finish: Sealed, no frame Thickness: 10 mm minimum, 12 mm maximum Optical Thickness: Same as 12 mm of Ohara PBL25Y at peak wavelength Path length homogeneity: 1/4 wave (425 nm) over 25 mm Clear Aperture: 156 mm x 156 mm square surface outside of C.A. must not leak out-of-band light Scratch-dig: 60-40 Central wavelength: 425±3 nm Full Width, half maximum: 8 nm Peak transmission: >50%; >75% desired Out-of-band blocking: 10^{-5} or better The operating environment is one of two cameras: Long camera, f/4.00 beam, 5.50 degrees maximum tilt of central ray Short camera, f/2.04 beam, 3.12 degrees maximum tilt of central ray The filter will be used with both cameras. Filter quote should include: * Price * Graphic plots of filter passband in f/4.00 and f/2.04 beams * FWHM in an f/4.00 beam, and wavelength shift at 5.50 degrees * FWHM in an f/2.04 beam, and wavelength shift at 3.12 degrees * Peak transmission * Out-of-band blocking Please send quotations to: .... ----------------------------------------------------------------------- OCIW would like to get a quote for two filters to be used in the IMACS astronomical spectrograph. Because of the short delivery time and loose specification on out-of-band blocking, we are assuming that these filters will be interference coatings on plain glass. The specifications are as follows: S1) The passband wavelengths are: Filter #1: 4800 A - 7800 A Filter #2: 6300 A - 9500 A S2) Physical size of optical face 165 mm x 165 mm square. The exact size is not critical, as these filters are intended to be potted into custom frames. S3) Maximum thickness is 12.8 mm. S4) Edges to be neatly beveled. No frame. S5) Optical thickness must be equal to 12 mm of Ohara PBL25Y. From the Ohara catalog, the indices of refraction are: n(PBL25Y @ D =5892A) = 1.581 n(PBL25Y @ A'=7682A) = 1.573 More specifically, the thicknesses for the two filters are given by the following formula thickness = 12 mm ( n{PBL25Y} - 1 ) / ( n{substrate} - 1 ) or Filter #1 thickness = 6.972 mm / ( n{substrate @ D } - 1 ) Filter #2 thickness = 6.876 mm / ( n{substrate @ A'} - 1 ) Note that because of requirement S3, the minimum index of refraction of the filters is respectively nD=1.581 and nA'=1.573, so BK7 is not a possible substrate material. S6) Substrate material is to be chosen by the vendor, consistent with ease of manufacture and index of refraction given by requirement S3 and S5. S7) Clear aperture is 156 mm x 156 mm square. This is a 95% clear aperture. S8) Total path length inhomogeneity due to substrate index inhomogeneity and surface figure is 1/4-wave or better over 25 mm. S9) Scratch-dig 80-50. S10) An Out-of-band blocking of 1/100 of the peak transmission over 3400 A - 11000 A is desired. An out-of-band blocking of 1/20 is minimally acceptable. S11) Delivery date is 2003-01-15. Filter quote should include: * Price * Substrate material and filter thickness * Graphic plots of filter passbands * Graphic plots of filter out-of-band blocking * Delivery date Please note that the delivery date is very close.