dichroic specifications (Vik Dhillon, 12/12/2000)
I requested quotes from 3 different companies:
The best specification and cheapest quote has come from CVI Technical
Optics - £6288.6 +delivery in 8-10 working weeks. The specifications
of the dichroics are:
- Dichroic 1:
Cut point at 387.79nm
Clear aperture: 118mm x 89mm
Substrate dimensions: 128mm x 99mm x 10mm
Coating side 1: >99% reflectivity (average) at 300nm-(start of cut)
and >95% transmission (average) at (end of cut)-1000nm.
Coating side 2: <2% reflectivity (average) at 335-1000nm
(i.e. a similar specification coating to that
provided on the lenses)
- Dichroic 2:
Cut point at 551.43nm
Clear aperture: 94mm x 70mm
Substrate dimensions: 104mm x 80mm x 10mm
Coating side 1: >98% reflectivity (average) at
(start of dichroic 1 cut)-(start of dichroic 2 cut)nm and
>94% transmission (average) at (end of cut)-1000nm.
Coating side 2: <2% reflectivity (average) at 335-1000nm
(i.e. a similar specification coating to that
provided on the lenses)
- Durability: ultra-hard oxide-based coatings, which meet our temperature
specifications of -10 to +30 degrees C and humidity specifications of
0% to 100%.
- Polarization: assumes randomly polarized light
- AOI: 45 degrees
- Substrate material: UV grade fused silica
- Scratch-dig: 10-5
- Wedge: <5 arcminutes
- Chamfered edges
- Surface figure: lambda/4 over whole surface, lambda/10 over
a beam footprint diameter (approximately 20mm)
- COST: 6308.58 GBP
Here are the final curves for dichroic 1
and dichroic 2 in WORD format
and an accompanying email from Helmut.
Note that these are theoretical curves (see below for actual measured
curves).
Here is a plot showing the SDSS filter band-passes
and the dichroic cut-points, with an explanatory email (to Tom).
Just prior to coating, I noticed that CVI had made the substrates the wrong
dimensions - the optical engineers had not received the final set of
dimensions from Helmut. This caused a slight delay whilst they recut
the substrates to the correct size. The coated dichroics were finally
delivered in December 2001. Here are the measured curves for
dichroic 1
and dichroic 2 in EXCEL format.
Here are the measured curves for
dichroic 1
and dichroic 2 in ASCII format.
Here is an accompanying email from Helmut.
Here is a plot showing the dichroic curves together with the filter
bandpasses (plotted using the program "dichroic_curves").
The cut-point of dichroic 1 falls at 380nm and the cut-point of dichroic 2
falls at 558nm. Whereas dichroic 2 is within the 1.5% manufacturing
tolerance quoted by CVI
(it is actually 1.2% off),
dichroic 1 is not (it is 2% off specification). CVI offered either to
recoat dichroic 1 or give me a discount on it if I keep the original. I
looked at the resulting throughput, as shown in this plot (computed using the
program "throughput").
The u-band throughput is approximately 7% below the value it would have been
had dichroic 1 been on specification. Given the importance of the u-band to
the ULTRACAM project, the great time and expense we have put in trying to
keep it at a maximum, and the much greater band-pass and efficiency of the
g-band, we were unable to accept dichroic 1 and sent it back to
CVI for recoating. The recoated dichroic arrived in the middle of January 2002
and the cut-point falls perfectly at 387nm. Here is the final curve for
dichroic 1 in ASCII format.