The Ohio State University College of Mathematical & Physical Sciences Department of Astronomy

Attendees: Darren DePoy, Jen Marshall, Tom O'Brien, Paul Byard, Ray Gonzalez, Dan Papallardo, Bruce Atwood, Rick Pogge, with Tod Boroson & Mark Trueblood attending via video from NOAO Tucson.

MODS CCD Detectors

Bruce reported on what he has been learning about custom-designed CCD options for MODS, this based on discussions with Dick Bredthauer and E2V people at the recent SDW2005 meeting in Italy.

The current concept for an ideal MODS CCD is a monolithic 8Kx2880 format 15-micron pixel device, in a combination of a thin, AR-coated devicesfor the blue channel, and some kind of enhanced red-sensitive device for the red channel. Current CCD fabrication lines are working with 150mm wafers (e.g., DALSA), or 125mm wafers (e.g., E2V). According to Dick Bredthauer, who has been working with Bruce and Mike Lesser on mask designs, one 8Kx2.8K device would fit on a single 150mm wafer, with other (smaller) devices filling the remaining space. If we were willing to make 7Kx2.8K devices, we could get 2 per wafer. It is generally agreed we want 8Kx2.8K.

One good bit of news is that the mask for a blue device could work with LBNL-style P-channel material for making thick or fully-depleted CCDs for the red. LBNL has its own process that does metallization (Aluminum) and processing that would be independent of the blue-device process done by Mike Lesser's lab. However, Bredthauer recommended that we consider instead using N-channel material like the Lincoln Labs devices, which have their own advantages to consider. Questions to be worked out have to do with the processing methods (e.g., compatible with Lesser's setup), etc. But the basic picture looks hopful.

The yield estimate, based on three recent runs of 4Kx4K devices on 150mm wafers at DALSA, suggests that in a single run of 24 wafers with the 8Kx2.8K design we could expect 6.6 non-shorted devices on average, which is very encouraging. If we backed off to a 7Kx2.8K format, we might expect about 14 devices. Additional runs will cost in the $70-100K range (the initial run costs this plus various one-time mask design, fabrication, and other costs TBD). What is also encouraging is that the subsequent runs could use a mix of materials, e.g., for blue and red devices. We need to consider what the best strategy would be.

There looks to be space on an upcoming run to be completed in the January 2006 timeframe for us to add a prototype 512x512 device that utilizes Bruce's multiple readout scheme. For a very low opportunity cost, this would let us prove that concept before committing to a full run for the ultimate big devices, and may even yield some small CCDs that could be useful to us for other applications beside MODS development. The timescale is good, as after the Jan 2006 run, we are looking at 6-9 months afterwards for a run of the big devices. At present we're considering 2 runs, the first with blue devices and the second with red devices (no mixed materials in the inital run). The general concensus of the Team was for Bruce to keep moving forward on this.

The other news was in discussions with Peter Pool and Paul Jordan at E2V. At present E2V uses 5-inch (125mm) wafers which could not accomodate our 8Kx2.8K device, but they are likely to be converting to 6-inch (150mm) wafers in the next year. They have expressed interest in our design, and have some good suggestions for us to consider, but the timescale and cost may not be workable. Bruce will continue discussions, as E2V is by far the largest supplier of astronomical CCD in the world now, and potentially has a lot to offer.

MODS Optics

Darren reported on a visit to SOML last week (June 20-24) to look in on MODS optics fabrication progress. Steve Miller was away, but Darren spoke with Jeremiah Valenzeula (Optical Engineer) and Roger Ceragioli (Optician). Darren showed us some surface maps he brought back showing progress on the aspheric side of the blue corrector lens. Progress has been steady but slow. On May 5 the surface error (peak-to-valley) was roughly 12-microns on the 200-300mm scales, and by June 15 was getting down to 5-10 microns. The coordinate measuring machine is still being used on >400mm scale, and shows a turned edge. Removing this effect by polishing down on smaller scales is the goal of the current work. They have about 12um to go at the outer edge. Polishing runs on a roughly 2 week cycle, with about 1 week of work, then another week to measure and begin the next cycle. They were about midway through one of these cycles when Darren visited last week. Overall, the feeling is that they have a couple of months of work remaining on the aspheric side of the corrector, then they will flip it over and begin work on the spherical side.

The spherical side is the one with the dings from the ball bearings that were dropped onto it during testing last year. They need to first remove 1-2mm of glass across the surface to regenerate it, then begin polishing to spec. Optimistically we are looking at 4-5 months remaining until we have a blue corrector, which will have significant impact on our schedule (for reference, the original schedule was for it to be delivered around the first of July, i.e., later this week). Darren and Jerry Mason will work on putting these new schedule numbers into the system to see what happens to the overall MODS work schedule.

Roger estimated that it will required 6-8 months to polish the red corrector, which has only had rough surface generation at this point. That, too, has a schedule impact that Jerry and Darren will work to assess.

As for the camera primary mirrors, they are currently being held up by the fact that they need to be worked on using the same machine that is currently committed to working on the second secondary mirror shell. It may be 3-4 months before that project will be to the point that they could start on the camera primaries and finish those up. At this point, we're looking at not seeing the final camera primary mirrors (without coating) until later this year.

In other optics news:

Paul Byard reports that HOLO-OR in Israel is getting ready to coat the bypass gratings for the IMCS system. We will have them coated in protective gold. After that they should have the 20 units ready to deliver.

JML has received the order for the blue field flattener lenses to replace the ones that were made incorrectly by SOML. We can expect these, coated, at the end of the summer.

MODS Assembly

Tom reported that the laser alignment tool for MODS is working great, some small problems, but otherwise moving ahead. He's ready to mount the multislit assembly for-real soon, drilling the holes and machining the shims. Once installed, this defines the reference axis and plane for the entire instrument.

Andy Krygier has designed a tail fixture to be used when rotating the instrument in azimuth on the mounting ring for flexure testing, as well as the motorized altitude drive. Both of these are moving along well.

Tom is getting ready to mount the invar flexures in the test camera primary mirror. The process has been worked out and it looks like it will work well. This will give us the first part of the optics system for aligning the blue channel.

Otherwise we're making lots of little parts down in the shop and keeping pretty well ahead of schedule.

Mark Derwent is away this week (he and his wife just had a baby girl in the last week), but before all that happened, he had passed along the LN2 reservoir design for the MODS detector dewars to Ralph for making the detailed construction drawings. These will be used in the bid package for fabrication. The detailed design of the "detector box" part is next.

MODS Slit Masks

Tom reports that he has made orders of test samples of the electroformed substrates for the MODS and possibly LUCIFER slit mask materials. He has contracts with two potential vendors, and should start receiving parts in 6-8 weeks.

MODS Mechanism Controls

Ray reports that he is setting up a lab system based on a Linux laptop and hooked it up to the multislit mechanism in the lab, just running 1 axis at a time until the multi-axis interface and power box is ready. All of it is running under Linux, and provides a system like the IMS software that runs on Windows for operating the MicroLYNX motors. Ray wants the lab people to start using this instead of IMS so we can begin to debug the real software as they work on the mechanisms. He has also started layout out the shared memory block that will contain all of the instrument multi-axis data, and is building things up steadily. So far, progress on the new system is excellent.

Dan has assembled the WAGO boxes for I/O, using a 2-channel A/D converter module connected to the ethernet interface/bus controller module. He has this working at a basic level well enough to hand off to Ray to start developing the software. The short-term goal is to configure one of these systems to put into the IMCS testbed utility box to replace the older "blue box" A/D units on the quad-cell readout system. This will be used for testing the IMCS system assembled on the MODS structure this summer.