Gabe wrote that three gratings were required. Actually we can make do with two, provided that one of them is suitable for full-beam use. Ken pointed out that (June 8, 1999) we might envisage projects which require long-slit and higher resolving power: this would argue for a "small matched pair" with one large grating as an option for later.
Phase B preliminary
The following text applies to the 1200 g/mm case but is indicative:
The gratings required, will be a full aperture blank, and two half pupil blanks. The prescribed grating from the optical design document is equivalent to the Spectronic Instruments (which includes Richardson Grating Lab) catalog master item 35-X-X-280. This is a 1200g/mm, 17.5 degree blaze angle. The size of the master is 154 x 206 mm.
Quotations were obtained in July 1998 for both a standard size full size grating, and the two half size replications on the user supplied blanks. Copies of these quoations, were faxed to Nigel Douglas at that time, and copies are attached to this report. This quotation is now out of date, as they are valid for only 60 days. In summary a full size (standard blank) grating, of 135 x 165 x 30 mm, Zerodur blank, was quotated at $(us)3895.00 plus shipping, with delivery time of 12 weeks. The two "half pupil" gratings, where quoted as, "replicated on user supplied blank" basis, for a blank size of 130 x 80 x 20 mm, at $(us)2200.00 each. Delivery time was 4 to 6 weeks. [Note the cost of the blanks would need to be added to this.]
Now, with the adoption of version 18 of the system design, I would advise a change in these grating dimensions. The above prices are an indication of prices for budgeting purposes, but a new quote will be requested before an order is raised, by the person responsible for the procurement of the gratings.
Because of the addition to the system, that allows for operation with the insertion or removal of the dichroic beamsplitter for the H-aplha arm, this has brought about a designed translation of the grating axis center in the x-plane of 3.15 mm. This will equalize the "shift" of the beam footprint on the grating surface, as the axis moves by 6.3mm when the beamsplitter is inserted. The extra area used on the grating, and the proposed mechanical mounting method, will require a blank of increased dimensions.
Reference is made to the Zemax raytrace relative "footprint" plots, and an allowance of 5 mm for the mechanical clamp space in three places. The required clear aperture is 130 x 160 mm , and a mechanical blank size of 140 x 170 mm. Blank thickness could remain at 30 mm, which represents an apsect ratio of 7 to 1. This new dimension will allow for the lateral shift, and pupil shift, with the insertion of the beamsplitter. [ note: the 140mm is the groove length, and the 170mm is the ruled width]
Likewise the face dimensions of the half pupil gratings will need to be increased, to allow for the new arrangement. The new dimension are 85 x 140 x 24 mm. This will provide a clear aperture of 130 x 80 mm. Again reference is made to the Zemax "footprint diagrams". [ note: the 140 mm is the groove length, and the 85 mm is the ruled width.]
From the previous quoations and advice from Richardson Lab, repilication right to the edge is possible with our blanks. (in fact this is the cheapest option). However as a specification for the blanks they require a minimum bevel on the front face (grating surface) of 1.5mm face width at 45 degree. The rear surface face has a minimum bevel at 45 dergree of 0.5 to 1.0mm. This fact will determine the lost area of the grating pupil for the two beam system, for the adopted concave arrangement. This will be approx, a dead area strip width, of 0.8 mm for the two gratings.
If the decision is made to return to the convex roof - half pupil arrangement this should not vary the costs, but a special shape of the blanks where the gratings join will be required to minimize the loss of light in this area. The losses will be greater than for the concave arrangement.
In summary then, exact quotations will be sourced at time of order, and when the mechanical drawings are available. From the above data, a replication price of approx $(us)3000.00 for the full size grating, and $(us)2300.00 each for the half pupil gratings can be assumed. Delivery time is approx 4 to 6 weeks for a user supplied blanks. The procurement of blanks for these gratings is an additional cost.
A quote for these special blanks could be sourced from Spectronic Instruments (Richardson Grating Lab), when the drawings are available. Or they can be sourced from another party. Documents and quotes held on file:
Quote standard size full size grating. 15 July 1998
Quote half pupil gratings, on user supplied blanks. 27 Jul 1998
Zemax grating "footprint" plots.
case A: TNG system, with full grating, and beamsplitter in beam.
case B: TNG system, with full grating, and beamsplitter out of beam.
case C: TNG system, with half grating to upper camera, and beamsplitter out of beam.
case D: as for case C, with half grating to lower camera.
Case E: WHT system, with full grating, and beamsplitter in beam
Contact
For special requests, Gabe's contact at Richardson Grating Labs is Glenn Hallit (hallit@spectronic.com). For efficiency calculations N.D. contected John Hoose (hoose@spectronic.com).
Current situation
The following text applies to the 600 g/m case:
The change to the 600 line grating arrangement will allow a change in the face dimensions of the two half pupil gratings. Zemax footprint diagrams indicate the dimensions could be, 140mm (groove length) x 75 mm (ruled width). These plots are on file.
Specification
Two 600 g/mm gratings, blazed at 8.63o e.g. Jobin-Yvon 510-13-XXX or Spectronic Instruments (Milton-Roy) 35-53-*-260, to be placed in a concave mounting. John Hart (email 29 July 1998) wrote: We have had a reply from Spectronic about replicating gratings to the edge. There is no problem in doing this. They also gave us info about edge chamfer requirement. It is clear that we could have a concave V mount with sub mm gap. They quoted a price of about A$3500 for each replication onto a customer furnished substrate. (This was for 1200g/mm gratings but will be better for the 600g/mm case as the concave angle is smaller).
Performance
Calculated efficiency at m=1 (26.6o incidence) is 82% (email N.D. June 7, 1999). Usually the manufacturer will confirm this calculation if requested.
Here are the calculated efficiencies for other orders for Aluminium (501nm):
m p s average 4 0.0006 0.0010 0.0008 3 0.0014 0.0008 0.0011 2 0.0020 0.0004 0.0012 1 0.7694 0.8917 0.8306 0 0.0663 0.0120 0.0392 -1 0.0853 0.0015 0.0434(thanks to R McPhedran, who used the code of D Maystre).
The worst unwanted reflection is 4% at m=0 and at m=-1 - these may require some baffling (email N.D. May 31, 1999).
Our own optimisations had been done with the older code of McPhedran and Botten, which only seems to work for INFINITE conductivity. We actually optimised angles for the 1200g/mm grating, and for reference we show how the 600g/mm compares (we include here the Milton-Roy 588 g/mm grating blazed at 9.5o, which would be even better but which is apparently not available in large enough size):
AD=35 wavel=501
INF alu (assuming R=8306/9010)
1200g/mm 17.5 0.69281 0.96626 0.8295 0.7647
600g/mm 8.6 0.9010 0.8306
[600g/mm 9.4 0.9171 0.8454 n/a]
[588g/mm 9.5 0.9125 0.8412 small only]
[588g/mm 9.2 0.9185 0.8467 n/a]
those marked [ ] are not available for various reasons but
indicate performance that might have been achieved if the AD
could still have been changed.
It is proposed that the `single arm' mode, with one grating covering the whole aperture, retain the 1200g/mm grating. this gives the instrument increased versatility without any sacrifice in generality.
Prices
Gabe indicated in the phase B report that the optical workshop at Stromlo would make blanks from stock Zerodur for $3000 (one large) or $3000 (two small). A standard full-size grating was quoted by Richardson at $6490 plus shipping, replication onto a customer-supplied blank $3670.
Thus: total cost (May 1999) of grating pair for concave mount:
$10400 plus shipping
total cost (May 1999) of one large grating, assuming we fit
the mount to a standard-size grating, $6490 plus shipping,
on our own substrate $6670.
We are currently (Jan 2000) looking for fresh quotes from Richardson and Jobin-Yvon. The following specs are based on the mechanical sketch of the grating, 12540-A3.
(Here as appropriate for Richardson, similar for Jobin-Yvon) grating type: 35-53-*-260 (600g/mm) we require two gratings, of dimensions (in your convention) 140 x 80 mm that is, the groove length is 140 and the ruled width 80. thus, these gratings could be produced by slicing a single standard grating (table 1) of size 154 x 206. we have designed our mount for a thickness of 24 mm, but this can be changed. information requested: 1. what is the standard thickness for grating type 35-53-*-260 2. what is the price of this grating in standard sizes (we cannot find a copy of the price list #35-2046PL, is it not on your web site?) 3. would you be prepared to slice a grating as described? cost? 4. what is the costs of replicating 35-53-*-260 onto two substrates of dimension 140x80x24? (we received a quote from you last year of $2000 for this - is this still valid?) 5. would you quote for manufacture of the substrates?
On Jan 5, 2000 John Hoose quoted $US2,560 each for the half-size (154x206) 600 g/mm gratings required, including substrate fabrication. The total bill of $10,000 is thus about that expected.
The gratings were delivered to the AAO around September 7, 2000. At the same time we received efficiency curves confirming the good performance.
Document Date: Sep 8, 2000