DEIMOS

7/13/95

[Tape One]

#.. So anyway, we are interested in doing these sort of things.... Maybe we should start with the 
detector p?? #We are making slow but steady progress on a couple of areas. We, I believe it is 
since the last one of these meetings, we glued a piece of silicon to a piece of aluminum nitride, 
which is our proposed base material substrate for our ccd package and cooled it down in a dewar 
to see how it flexes, see how good the match in expansion between the aluminum nitrate and the 
silicon. The ?? flexed about nine microns, peak to valley. #How big was it? #The size of a 2k by 
4k, so it was long the long dimension. So for a first shot at it,I thought that was pretty good. #That 
was really good! #That's right. #Because we could take the mean of that and it's just right at our 
level of spec. #I think we can probably do better, and we are going to try various experiments, one 
of which is to take and glue a second piece of aluminum... #To the other side? #.. to the first one, 
and just make it basically double thick aluminum nitride. #Oh, I see. Ok. #Stiffen up the alumi-
num nitride. # I thought you might put some similar piece of silicon on the back of the aluminum 
nitride. #Well that is also an interesting experiment, sort of balance out the forces. #There is an 
infinite number of such experiments, and I think we can certainly do better. #You glued it with 
glue that set at elevated temperatures. #That's right, it set at 120 C or something like that. #So this 
is sort of twice the temperature difference, that maybe we will actually deal with. #So that is 
another experiment that we will do with additional parts... #Ok. #..is setting them at a lower tem-
peraure. #Good, but it looks like by pulling out all of the stops we will solve that problem, #Yeah. 
# Oh, that is great, that is really good news. #Our Orbit wafer run, the first run of the proposed 
actual 2k by 4k devices is supposed to be out July 19, and we have two additional runs, the P.O.'s 
are in, but they are just on hold there at Orbit until we get this first one out and taken care 
of..#Now I thought you would get this sooner. #Well, so did I. #Given the last meeting, I thought 
this was imminent the last meeting I was at which was at the beginning of June. #Yeah, right. #So, 
it didn't happen. The last I had before this number was the end of June, it would be out by the end 
of June. But. #Right. So, do you think this will happen? #Well, it might, because we seem to be 
converging, its only a week now. #Ok. #As we move forward, it's not moving forward as fast as 
we are. And on the ccd thinning effort, we are making slow progress there, we are investigating 
several additional commercial suppliers of the mechanical thinning portion of the effort. If we 
take our ccd wafer and bond it to a handle wafer and then have the ccd wafer mechanically 
thinned then we will do a few microns final chemically thinning. We are looking at some new 
commercial suppliers for the mechanical part. #This has always been part of your plan. #Yeah. To 
have that done commercially. We just found that... we went to a meeting on wafer bonding in 
Reno, a month ago and found out that there are much better techniques for this mechanical thin-
ning now, people are designing a special million dollar machine and things for doing this thin-
ning, so we have sent some samples off to one company and we just made up a few additional 
samples of our bonded wafers to send off to a second company. So, that is basically what is hap-
pening with the thinning effort. #Well, how about your chemical thinning? #The chemical thin-
ning, we have done some simple experiments. That all seems to be pretty much in hand. #So, you 
think you are poised as soon as you have some properly mechanically thinned pieces to go ahead 
and make sort of closure tests with your chemical thinning? #Yeah. We don't have all of the 
details worked out there to do the very best job we can, one needs to get a much better character-
ization of the what the actual wafers are that you are thinning. Basically the resistivity of your 
material and then you can tune the chemicals to get optimum selectivity when you get to the epil-
ayer. #So are you waiting for that or how do you get that. #There is a company that will do those 
measurements for us, and we have been talking to them. So we will send them a test sample but 
you have to do it on the wafers you are really going to thin, so we have been waiting to get some 
of these wafers that hopefully will come out on the nineteenth. #So, really, a lot of things are 
delayed because these 2k by 4k things are not in hand. #Well, some things are but we have plenty 
to do. #And in a related thing, we have some of the aluminum nitride over in the shop and we are 
going to thread some holes into it.. We are going to do some experiments to see what we can deal 
with. #Right. #The best we can hope for is some threaded holes. We don't understand the mate-
rial, it is not in books, we have to play with it yet. #So, what are your milestones for our next 
meeting? #Well, we will continue with these tests on the aluminum nitride, to see if we can under-
stand it a little better. Exactly how good we can get it, and we are just going to continue. We are 
hoping to get some of these thinned wafers back from this company we sent them off to. Those we 
will then try doing full wafer chemical thinning on. #Is there a change that you could do that in 
the next month? #I certainly hope so, I expect that we will get these things back. The company we 
sent them off to is over in Ireland, but the doesn't seem to be a big impediment. We talked to the 
president of the company, he was real enthusiastic about the whole thing, he could see all kinds of 
other applications for adhesive bonded wafers, so he was keen to do these experiments to see if he 
could actually thin them without them coming apart in his machine. #Do you have a timetable of 
what you are doing for the next six or eight months? #No. I wouldn't say we have a timetable, we 
have a general idea of what we want to do, I guess I could write it down and put some artificial 
dates on it if you want me to. #If you could just put five lines or something and sent me an email I 
would really appreciate it, because I have kind of lost track now of what the flow of work is in this 
area for the next six months until the end of the year. So, I would just like to get brought up to 
date. #I still don't have an official ok from the people at Davis to do all of our chemical thinning 
up there, mainly just because of summer schedules, they actually haven't had a meeting, couldn't 
get everybody together up there to have a meeting of their users committee, but everybody we 
talked to said it sounds great to them, so. # It is just timing. How about Lincoln? #I haven't heard 
anything from Jerry about that other than, back when we had our meeting at site, before the last 
meeting, he said that Lincoln was going ahead starting with the design even though no contract 
has been signed because Jerry didn't have all of the check with all the consortium members. But I 
haven't had an update from his since then. #Do we need to.. somehow I really think that the 
squeaky wheel is going to get the grease here and this Lincoln thing, which is more important to 
us than maybe anybody, we just constantly sort of have to bug them on Jerry, probably, on at least 
month centers. I don't know if we have been doing that. #Well I haven't been bugging him, I think 
he is keen on getting the whole thing going as well. #Yeah, we are obviously feeling a real sense 
of urgency, I think I will just send him email and I will copy you, to find out what is happening. 
#Has Gary asked you for a update for this next SSC meeting next week? I sent you that message, 
he said he would talk about those things side? Lincoln Labs, #After the site meeting, we had more 
or less agreed that he would present something at this upcoming meeting, but in fact nothing has 
happened. #Yeah, that was going to be my next question, what about site? #Well we still don't 
have the sample that they are going to send us that they said they would send around the middle of 
July, so. #They send a 2k by 4k to Kit??b a working thin device which got killed. #What!? #Oh, 
no. When did they do that? #Apparently they put it in a dewar and tested it and then they put it in 
a dewar to take to the mountain, and when they got to the mountain it wasn't working anymore. I 
don't know the full story, I heard this about tenth hand. They have existed. It's a proof of concept 
or something. #So this is the one that we were going to get after Kit ??. #Well they told me that I 
would get it first, so maybe this is a different one. #They have got more than one working device, 
they have one less apparently now. So maybe I'll give them a call and find out what the facts are. 
#That is a good idea. If anything is written by Lupino or by you for the SSC will you make sure 
that Garth and I get copies, Garth will get a copy in any case. #Yeah, you might also bug Lupino 
too, I don't know if he is in the middle of one of his current panics about the CHT system, but if 
you might things to move with site or if you want anything to happen you will have to start soon 
with the SSC. It will take a little while. #Well, originally I was thinking perhaps I could give some 
kind of a presentation to the SSC on this one coming up on the 22nd about our thinning stuff, and 
sort of switching gears and doing the chemical thinning up at Davis, but since that hasn't actually 
formalized yet. #It seems premature. #The next SSC meeting is actually September 18. #When 
are we reporting? #It could be at that meeting. That's in Pasadena. The next one in Northern Cali-
fornia is probably not until November. #There isn't a meeting in August? #No. September 18 and 
November 27. #Well that will be a good meeting, certainly by then, unless they have blown a hole 
over their 2k by 4k devices, I will have had one for a month and we will be able to say something 
definitive at that meeting, and I am sure that we will have our arrangements with Davis all set up. 
#These things take time to percolate through to the SSC and so it is good to have Gary give a good 
solid presentation this time around, so make sure he has some on what you are doing and what 
you are thinking of doing, as well. #Ok. #Maybe a one-two punch is the thing, but if you don't 
have the results by September I guess it would be premature. Anyway. Cause if you want the SSC 
to do something it is better to tell them and let them think about it and then get. #The thing we 
wanted to ask the SSC to do was to give site some money to work on the flatness issue. #So you 
want them to do something. #Ok, Should we go back to Deimos and NIRSPEC? #Sounds good to 
me. #Why don't you just briefly introduce this so that everybody knows what we are talking 
about. #Ok, the AO people at Keck, headed by Peter Wizinowich have concerns about moving the 
AO module on the left naysmyth platform to allow instruments like NIRSPEC to directly use that 
focus rather than take the loss that they would incur going through the AO. #The problem is that 
NIRSPEC wants to run with and without AO, and the question is how to accommodate the move-
ment that is needed to do that. #The original concept was to simply have the AO step back, and 
Tony Gluckler had come up with some conceptual idea of how this might happen on rails, and I 
guess is this thing has percolated through and they are getting closer to their design and they real-
ized that to relocate AO on it's kinematic mounts to the tolerances that they are going to have to 
which think must be in the order of microns. #God, that's incredible. #Yeah. #What is this AO 
thing? #Adaptive Optics. #I know what adaptive optics is but what is the thing itself that needs to 
be relocated I don't know what that is. #Its a ?? optical bench and horizontal optical tail, all the 
elements of ?? #Roughly how big piece of stuff. #Couple of meters. #Kind of like two tables here. 
#Ok, and the problem is that they have to locate it relative to what? #The telescope, I guess. #To 
the. #Well they would be locating the kinematic mounts on the naysmith platform, so they are 
locating relative to that platform. #And somehow or another if they don't put it back in the same 
place its calibration is different or something? #Well the problem comes at the interface of the 
instruments, where you need to locate things as was mentioned a few microns level because of the 
scale. Its f15 and its, and they are now operating to fraction limit of one micron basically, so they 
are going to going to be they.. the numbers of the Nerk two people came up with locating things at 
the focal plane to that sort of level of accuracy. #I don't understand why, all the AO does is pre-
sents a nice star image, doesn't it? So it is a piece of stuff that is out there ahead of you that makes 
the star image look sharp, right? #Yeah, I basically agree with you Harlan, I don't understand it 
myself. #What do you have to locate, what's to locate, I don't understand. #I don't know. I cer-
tainly have not gotten involved with this and don't understand the spec on the other hand, several 
people have thought about it that usually seem to know what they are doing. #I am trying to get a 
picture in my mind what it is that needs to be located and why and what difference it makes to 
anything. I don't. #We can't tell you. #The position of anything isn't even defined to a few 
microns that platform is flopping around in the breeze and all kinds of things are happening at that 
level, right? So, its meaningless even to talk about locating anything to a few microns. #I tend to 
agree with you. #I think, ??? ways?? of light rather than a great big steel structure backing around 
in the breeze. #Furthermore, you tilt the secondary and the image is going to move around. Who 
knows what is going on. #Yeah, at that level, who knows what a big structure is doing and who 
cares. So I don't understand ... #Are you concerned about this or are you convinced that these 
people know what they are doing? #This has come up often enough from the AO group and the 
AO science team that I assume they have gone through all the numbers and the claim was they 
need to locate the instruments to the adaptive optics output to the order of microns. So the prob-
lem is not moving the table, its moving the table with respect to the instruments. #Well, that's very 
puzzling, because that has to occur in any case, because we have to move NIRSPEC around, as 
you have just currently stated the problem its now a little more puzzling. #This has got many dif-
ferent aspects to it and it has been somewhat simplified in these memos. #What does that have to 
do with Deimos anyway. #The impact that it has on Deimos is that one of the solutions is that, 
instead of moving AO around, that NIRSPEC bodily moves around the other platform and coexist 
with Deimos when they want to do direct imaging mode. #Without the AO. So that NIRSPEC 
would use our platform when it wants to work without the AO. #Well I sure have to be convinced, 
if anybody wanted my opinion, anything to a few microns is just nonsense. #Right, I think there 
are two ways of discussing this question, namely, are these people thinking sanely about the prob-
lem. We are not going to resolve that question here. The other question is I think a more general 
question about whether or not Deimos should legitimately plan to get out of the way ever for any-
thing, and I think we should, I just don't think that you can squat on that platform forever. #I mean 
there is a big part of our baseline. #So I think that we should forget about this particular thing 
about NIRSPEC I think we should design ourselves to get out of the way, which we have always 
intended to do. So why don't we stop talking about the NIRSPEC connection and think about this 
is really a problem for us or whether or not it is already well in hand. #My position is that it was a 
design consideration from the beginning. #From the beginning, I think we felt that it was, this is 
scarce real estate. .. #Sandy, you mentioned getting out of the way, does that include getting off 
the platform? #No. #Ok, just rolling out of the way off the side somewhere. #Going into our stor-
age location and we have identified how we want to do that. The advantage from my point of view 
that it has right now is that we get to face up front people who might coexist with us on this plat-
form and deal with those things rather than a few years down the road, so I see some advantages 
here. #Right, ok so. I don't understand the stowage plan and was hoping to have a review of that 
today. Can we.. #I don't know that we have developed it beyond what we have seen. #Just the 
original concepts we presented in the PDR. #Well there have been several concepts floating 
around at one time or another. #The concept that we apparently have is that it would come back, 
and then move over, and sort of a diagonal parking. #Because of our size and weight we would 
plan on always staying on the platform, but that means that we need to think when we are getting 
them to set up the platform that they have an access system like what was suggested for NIRSPEC 
rails that come from storage on the naysmith level. And really the only issue is how those two sets 
of rails cross one another, and I don't see that as a big problem, in fact, as I say, I would rather 
face it up front. That's also a CARA, we obviously have to do what we need to do, but then that is 
a CARA issue as to how they access and allow other instruments to come on. #I believe we will 
come up with ways to cross those rails without jolting our instrument at all. #Ok. #Trains do this. 
#Yeah, different requirements for bumpiness and tolerances. #Well, they also move faster. #But I 
think we can find a graceful way of doing it up front, rather than maybe fooling ourselves that we 
are not going to move. #No, we always said we would move, but I think Sandy's point is appro-
priate from our point of view, this NIRSPEC issue is somewhat peripheral, and it involves a lot of 
different items and issues at times. I did have a long conversation last night with Ian McClain who 
is the NIRSPEC PI and he is very concerned, as I was too, and Sandy will be too that we are going 
to get to the telescope at about the same time for commissioning and this is a dreadful thought that 
we, two instruments that are being commissioned and used in their early days are on that same 
platform trying to do that. So we were discussing other options and the Keck I platform is still a 
possibility, in fact it's Ian's favorite position for doing commissioning. #I see, so they might com-
mission in one spot and move. #So they may stay there for a year or so while the adaptive optics 
is getting sorted out because that is going to be commissioned on the same time scale as well.  
#Ok. #Wizinowich is talking about commissioning in 97 but there is no way they are going to be 
ready by `97. So, I am going to get back with him on friday and then we are going to set up a tele-
conference and talk with Wizinowich next week before the SSC meeting, so we can at least get 
something sorted out upline before we get into a public meeting and have this discussion again. 
#Now, I would like to say that I've never liked our diagonal plan because it looks inefficient of 
space, so I would like to put in a plug for something that is maximally efficient and moves off to 
the side parallel to the platform, leaving an absolute maximum of space that is free there, and I 
think that the reason why we have sometimes talked about diagonal plans is the existence of our 
big cooled thing with all the cables coming out of it and we didn't want to move that around, we 
were trying to move in such a way that we kept the back end of our instrument somehow... That 
box is a lot less important than it used to be. It contains a lot fewer things than it used to contain, 
maybe we just put it on a simple thing and roll it around enough. I don't think the existence of that 
valley box ought to really drive what we do with the basic instrument. Why don't we think of that 
first and then.. #We have a nice outline drawing because we have a copy of their NIRSPEC PDR, 
we now know what the instrument looks like, and a year ago we were guessing what somebody 
might present. #Now another concern I wonder about.. We are pretty heavy, what happens to the 
telescope when we move off to one side of the naysmyth platform, is this going to cause any 
imbalances or anything like that. This will be the first time that anything of this size has moved 
around. #They haven't given us any reason for concern, so I haven't had any at this point. #We 
tried measuring this with HIRES, we lifted the instrument up and put a double indicator on the 
grating and we had eight thousand pounds just hanging in the air and we were too crude to mea-
sure that deflection. #Eight thousand pounds. #We are not that heavy. #I have asked this question 
of Nelson and Wudowski and all those guys. In the beginning there was going to be twenty thou-
sand pounds of lead on both sides that you would take off as you put instruments on. No one had 
the question of what if someone shows up with a chassis and they just put the electronic chassis 
and a chair or a table there. Do you have to go get lead? No one has continued to worry about that. 
#It doesn't seem to be a problem. #But I don't know if that is written down anywhere what the 
stiffness of that platform is, which is what you want to know. [End of Tape]

Tape two

#..HIRES commissioning, finds the gains? of the guider loop had to be brought down ?? working 
fine with other instruments once HIRES went on. #That's probably occur when we put ourselves 
on the platform, now if we just move around on the platform, maybe the effect of that will be 
smaller. #I think it will be smaller.  #I don't think that that, I mean Keck weighs 270 tons and 
eight thousand pounds is four tons. #It's a small fraction. It just a percentage. #Ok. #I have a ques-
tion, not about this subject. If you are going to talk to Ian, I would ask you to ask him what kind of 
vibration their compressor puts out, this is a closed cycle helium system and it may be something 
like operating a large freezer right next to us. #Yeah, I know, actually talking about plumbing the 
telescope such that the compressor would be off the whole telescope. #That would be ideal. #It 
may not be practical because of the very high pressure cables are large so they may put it on the 
platform, but Ian is worried about that himself and so they are looking at vibration isolators and 
temperature isolators, these things consume six and a half kilowatts. #Six and a half kilowatts? 
#Yeah, in a little bit of cooling, but anyway. #We are definitely running contrary to nature. #It's a 
watt of cooling or something. I don't know. So they have to change the glycol system improve 
that, they have to double or significantly increase the capacity. #Yeah, the dewars hold fifteen 
hundred watts. #So anyway they are thinking about this and they are concerned about the vibra-
tions too, so if they were there they would have it isolated. #I guess the good news is that it would 
at some point coexist with AO on the other side. #Yes. #Otherwise typically the infrared instru-
ments are far less worried about that sort of thing. But if they are going to design it to coexist with 
their AO system.. #But if they have a high resolution spectrograph, they are going to need to be 
concerned about vibration as well. #you know in the long run, Ian's position is that they will be on 
the AO bench and they will probably want to stay there largely. They are really just driven by this 
three to five micron background. There are seventeen reflections in the AO system so they are 
worried about the background going sky high and they are still thinking about whether there are 
going to be the number of optics in there that will allow them for access to the f15 original bench? 
without the AO with fewer reflections than seventeen. #But still I guess they have to extend the 
focus out, right. #Yes, so they will have to have relay optics there.#Ok. #So we will keep in con-
tact with Ian and make sure that a rail system can coexist, I have no doubt that it can. #That is a 
generic thing that we need anyway. There will be other instruments that will want to come in and 
use that focus ?? #So who's job is it to design that? #The what? #The rail system, do you think 
that is our job? #We will probably end up being involved in it, I am sure there are some elements 
of the telescope in there too. In any case it is not a real big job. As far as stowage position, then, 
our whole structural design is developing on that. I think that by Novemberish, by the CDR we 
should have a better idea of how that.. #So you are saying that you would like it to move back and 
out and ?? put the lens up parallel with the edge in close to the edge to maximize the space. #I 
would like to. #That is just space for other instruments. #yes, #The problem with going to the end 
of the platform and going, what I think I am hearing, is parallel is that isn't access for other instru-
ments coming directly off. #No, I think this is on the side, you back away from part of the, 
#Instead of diagonal. #Yes, instead of going diagonal just laterally up against the edge. #So we are 
still pointing in to the center of the telescope, #That is correct. #Electronics? #I just had one ques-
tion on this, when Deimos is off part of the way, what requirements for access do we have to make 
th?? or other activities? If you back it close to the edge of the platform, #Too far close.Yeah, good 
question. #We will want complete access to the whole instrument. #Maybe that could be or maybe 
there is an intermediate stop position where you have access and then there is a real stow position. 
#Hm, that could be. #Something to think about. #Because we intend to have it operational when it 
is off as HIRES is now. ##So what does that mean? #So, you might want to change your grating. 
#Yeah, so you might want to stand on something instead of leaning off the side of the telescope, it 
is a long way down. #Electronics. the instrument controller design in stable at this point, before 
Terry Rickets left, we discussed the design and made up a list of about a dozen points that need to 
be ironed out, most of that will be ironed out as the project progresses and things are better 
defined. There are relatively minor things such as the turning on and off the camera, verifying the 
connector types at Keck II so that we plug into the right when we get there. Well we are in the 
progress of hiring the new EE and he should be on board about mid-August. #August first actu-
ally. #Have we selected a person? #We have. Do you want to hear? #He has some background 
with CCDs. He is familiar with uncooled types, he has used them in photometers in Los Angeles. 
He is a real hot analog electronics designer, and this comes from all of his references. #Dave, how 
old is this fellow and how much experience does he have beyond school. #He graduated with his 
masters in 86, I am going to say, I could be a year plus or minus and he has been working for high 
tech electronics firms ever since. He has worked for two particularly, first he worked for Teldyne, 
not Teldyne Gurley I gather, but an instrument company called Teldyne and he designed a million 
dollar plus test instruments some of which are still in use and still being sold and by all reports 
fairly successful. When his supervisor, or I guess the president of Teldyne went to photo whatever 
they are called, he recruited Chris to come with him about a year later. #What is his motivation in 
wanting to make a career change? #He got laid off. The company he was working for had not 
released a new product for a couple of years, and in that business he could see that things were not 
going well. Just as we were ending the interview process he was laid off and hired back right 
away incidentally as a part time employee. #No benefits and that sort of stuff. #How many appli-
cants did we have? #Thirty-five. #And how many interviews did you.. #Six interviews for a short 
list. So you had a pretty good selection. #We did, we ended up, this guy was far and above any-
body else, we weren't real sure what we would do if this guy wouldn't take the job. Our number 
two candidate was a fair distance back there. #I want to say that I have grave concerns since I 
have been here the way the hiring is done, technical people on the faculty are pretty much 
excluded from the whole process and I think the track record has been very poor over the last ten 
years, quite a number of technical people have been hired that turned out to just be a complete 
bust and have been later fired and so I have some concern. #We did have a faculty member on the 
selection committee. Mike Bolty. #Mike Bolty has no technical experience at all, academic kid 
essentially in the business and new and everything. #The real problem is the way we use our engi-
neers is very different than most people are used in research and large industry, typically what will 
happen is that a person will graduate and will very quickly develop a specialty and stay pretty 
much within that specialty. What we are doing is we are hiring an electronics engineer to be in 
this to be more or less a generalist. So it is tough and we did make a mistake. We think we have 
got a pretty good candidate this time, we know that by all reports is an excellent analog designer 
and in my mind at least, designing in electronics and analog circuits is far more difficult than 
doing digital, in fact, one of the references I talked to said that this guy was a hot analog designer 
and so I asked him how do you he would do digitally? The guy came back and said, why would 
you use him as a digital designer, this guy's too good. #Our problem is that we need him for 
everything. #We need somebody who has a very broad background. Terry Rickets who had the 
job before spent about twenty years developing into the way he did his job, which he did very 
well. Last time we tried to essentially clone Terry, and that didn't work. So this time I am backing 
up a bit, and saying, what I want is a hot designer. #My comments expand over the last say ten 
years. You have only been on board a short time and you stubbed your toe once maybe, but there 
have been a lot of toe stubs, maybe a dozen or so over a ten year period and the underpinnings are 
systematic in the way the administration of this place is operated does not involve those members 
of the faculty with technical expertise. Typically I am never involved for example and Lloyd Rob-
inson was virtually never involved. #Lloyd was involved in the last selection. #In the last one, ok, 
but in the past.. #In this one he was not involved. #I am just expressing a sort of global concern 
that I have had for a long period of time. # I think I also have a concern but it is not yours. My 
concern is that the advertisement and hiring process sort of throws out a job announcement into 
the universe and hopefully somebody good out there will see it and bite and then come back. It 
has happened that with our announcements just as happened now, you get one really good person 
that you feel you have confidence in and you ask yourself, what would have happened if we didn't 
have that person who replied. #I think. #It's just thin. #It is hard to find somebody. I did hit all of 
the observatories. and we did ?? on the world wide web. #That's the best you can do. #Do you 
have a feeling that our salary differential is making this .. #We are somewhat lower than industry 
jobs, but industry jobs are disappearing rather rapidly, so we are becoming kind of attractive. #We 
are such wonderful people too. #And the diversity of the job is very interesting to a number of 
people. #One other statement, it's a simple thing and that is getting the faculty available to sit in 
one place for six interviews is very hard, when we hired Dave, I think we did a good, but Joe 
Miller had to have a stand-in when he couldn't be there, so then it is hard to compare notes, Well 
how did you think about that person and if you didn't sit there it's a practical consideration. #That 
?? functioned very well in getting people together. #I have used that as a good example. But pre-
viously we have had that problem, now you are probably more available. #Next time you are 
going to be on my list. #You are going to be sorry Harlan. What is this person's name? #Chris 
Wright. #So our game plan is to give him a bunch of design jobs right of the bat and see him go. 
Ok,. #What else have you been doing for Deimos. #I have gathered together a bunch of informa-
tion in the catalogs on calibration lamps. #My favorite subject.  #Nice folder, getting thicker. I 
guess we will want to discuss that off-line. #I need a deadline from you as to when you need me to 
sit down and really deal with this, because it is getting constantly pushed off. #There is sort of two 
tiered decision there. We have got to determine if we are going to be using hollow cathode lamps 
for any reason. #No. #Ok, that means that we can put this decision off for quite some time because 
the electronics control just turns on and off relays. So we can put any power supply on the other 
end of those relays that we want for lamps, and then it becomes a question I guess for housing and 
mounting the lamps and power supply. #But we will need to know how many but I think that has 
been fairly stable, of course I have forgot the number. #No, there is an issue here of whether or not 
we need one or two lamps of each type, and that relates to how bright they are and how they are 
going to illuminate the hatch cover, and where the hatch cover is sitting. How far in front of the 
focal plane it is and maybe what it is made of. #In talking to Tom Bede, he said they used penray 
lamps, and he said they used two of them but they stopped them way down with some sleeves, 
they have got plenty of brightness they just use two so they get more even illumination on them. 
#At least for the mechanical design, if we assume we are going to use two of each, we can always 
back up to one, I assume and then if it becomes a problem, we can always .. #Let's assume two, I 
would like to get by with one, because I think its true that if we use two we need two power sup-
plies, isn't that how it works? #Yes. #It's going to double all of our costs. #Sandy, do we have 
someone who is basically in charge of how to do the calibration and secondly, if so, do we have a 
write up of what the general process is going to be? I know we talked about scattering off the 
dome, scattering off a sheet near the secondary, scattering off a cover, but as far as I know, no sin-
gle metod has been identified as the preferred method. #That's right. I am worrying about this. 
#Are you.. #There are tests going on at Mt. Hamilton to try and understand the callibration pro-
cess and problems of the Kast red ccd up there, which seems to be a good prototype problem ccd 
to work with because it has bad fringing. #I guess the thought is the tough problem is what to do 
to get the illumination to go in the spectrograph properly and we'll supply as many lamps located 
wherever they need to be in order to make that happen? #Thats not how I am looking at it now 
Harlan, I am looking at it from a more fundamental point of view, like a physicist would look at it, 
what is this device doing, what is going on in this device, how is the light behaving and why are 
we seeing fringing problems and I don't think we... We have identified at least two possible con-
tributions to fringing and what we need to do is to quantify them separately... #The fringing is a 
different issue, I am talking about the issue of how do you supply the light to the spectrograph. 
#They go together, why don't we discuss this off-line, because as part of my little project here to 
understand fringing better, I would like to use some of Brian's time to do some calculations, and I 
was meaning to come to you and discuss this anyway, so why don't we have a longer discussion? 
#Ok. #One other point that comes to mind on calibration lamps. Apparently with the HIRES, 
Steve?? used the hollow cathode that Lorie? Margon? lamp? and apparently they built in two 
complete systems because they were very hard to get to and they just used one, and the other one 
is there for when the one. #In case we would ever want it? #I got the impression it was, if during 
the middle of a run, if you can't get to these lamps, it might be nice to have a second one there, 
available to just turn on. #You are just saying that this is a redundancy this is insurance, so we 
should have the same...#As insurance. I see, so you are saying that having two lamps is not a bad 
idea. #It depends on where they are. #It depends on whether a technician can get in there and 
change one in ten minutes or if it's going to take a crew of five people a day to get things open and 
get to them. #What do you think about that. #I think in HIRES it would be not five technicians, 
but maybe three. #How about Deimos? #Oh, it's not going to be that hard. HIRES is in an awful 
position, where we ended up putting that for various reasons. #Is there a technician on duty at all 
times, as opposed to a night assistant or somebody who doesn't necessarily know about the inte-
rior of the instrument? #The first part of the night until about one o' clock in the morning. #There 
will be a technician soon signing up, we will be on on the latter part of the night as well, so all 
night in the future, yes. # That's good news. #It's driven a lot because of remote observing, is part 
of the justification, but we also get the advantage of having a technician there too. #You need two 
people up there anyway so there might as well be a technician. #I'd say you could change it as 
easy as changing a grating. So whatever you want to put down, one person, one hour. That would 
be... #This is a grating into the slide? #Yeah. #Not just changing the grating position. #I'm sorry, 
manually changing the grating one for another you have sitting next to you in an envelope. So, I 
don't know how much you can afford, what does that second system cost in hardware, a thousand 
dollars? #Depending on if you go with the Osram type its about probably fifteen hundred dollars, 
if you go with the Penray style lamps you are talking more about five hundred. #But we have sev-
eral kinds. So that's another planning decision we might want to make I mean it is nice to have 
some kind of lamp in the system just to see where you are, but you might decide that you don't 
have to duplicat every kind. The use of the penray is good, I really thought we were going to have 
to use the Osram lamp. And by the way the penray lamps are too bright in Low Res. #They say 
the they stop them down with sleeves and I assume that they could probably make sleeves with 
smaller slits or .. #Too bright is good. #Yeah. #That depends on what you are illuminating, where 
the lamps are placed, but that is I think beyond what I will be dealing with. Further on in electron-
ics, the CCD controller of course is waiting for the new engineer and further specifications from 
the project. Some issues that I have noted that need to be addressed are we are very near the end 
of what IO capabilities we have from our controller boxes, so I recently recieved some catalogs 
and we are looking at these opto 22 modules as possible route for expanding our IO capabilities 
box. #This is for the instrument controller? #Yes. This is for bits to turn things off and on and to 
read information back.. Jack came over and was looking at getting more position information 
from his grating selects and we looked at the number of bits we had available and this one thing 
would have taken up all of the bits if we used them as individual bits, I think it is time to at least 
plan for the ability. #Expansion, and we are talking a two channel instrument, or a one channel 
instrument, one beam or two beam. #This is one beam, we would have to do it for each beam. #Is 
this a big deal, or is this. #No, it does impact the number of ports we will need on the terminal 
server. That again is something I am looking into. Apparently, Galil has some expansion for an 
Opto22 also, so I will look into that. Talking with Terry there is one RS232 available to go out to 
an Opto 22 controller. The problem is the Opto 22 don't mix analog and digital. So we'd need to 
pick that very carefully, whether we'd need the analog or the digital IO. #Another issue is with the 
current design that Terry has for monitoring the power supply of the UPS's we'd need to get infor-
mation from the manufacturer and make sure that what he is designed in fact will work with the 
outputs that they are supplying. And then I want to order a fan cooling unit. We are going to put 
these in each of the compartments with the instrument controllers and the ccd controllers. We 
would like to order that and then try to figure out a way to sense if the fan is actually working. We 
thought of some scenarios where you can turn the fan on and be drawing current and the blade's 
not turning and that sort of thing and Terry came up with the idea of a couple of different pressure 
sensors in various orientations to try. #How about a little strip of paper and a ccd camera to just 
monitor...#Will that work in all of our orientations? #One interesting thing is you could shine 
some LED's through the fan blades. #There's a radiator right in front of the fan. So we have a cou-
ple of ideas of what to do. #??? # Well the problem there is we won't have much room next to the 
heater. So at any rate that's just something that we want to work on at this point. #How many fans 
total do you see? #The way the Eric and Jerry got together and they got a coolant chamber and 
two fans on it and there will be one of those in each of the four isolated instrument compartments, 
or controller compartments, #and then twice that for two beams? #No, that's two per beam. #I am 
having a hard time thinking just one beam. #And that is about it for electronics. #I am wondering 
at what level we have to worry about vibration from those ??? the compartments for the electron-
ics ?? isolated in the ?? room?? #Yeah, where are these fans located, are they on the instrument or 
off. #All the electronics and all are on the instrument in sort of a ring around the instrument, I 
imagine we could put that whole fan assembly on ?? #We can pretty easily take that high fre-
quency above a few hertz and there is usually not a lot of energy in the lower stuff. #Have you 
asked that because you saw something in HIRES? Because we are duplicating that system. #Yes 
and no, where I can think of fans in site at HIRES are not as close or tightly coupled to... #The one 
by the ccd is about seventeen inches away and its... #But it's not on the optical bench. #It hangs 
from the ceiling. #It's not but it's got those wires that we tried to isolate. Have you seen any wig-
gling there? Ok, good. I think we can probably isolate these too. #Camera and optical tools.[End 
of Tape]

Tape three

#...about a week ago and found that everything is going along just fine and there is no anticipated 
problems in the production schedule of either of those items as of yet, of course they are not very 
close to delivery time yet, so things always look great when it is a long way up the hill. At any 
rate, I think the news is good there. We are about ready to ship the calcium fluoride, the big bool 
that we own. We are just waiting for the bureaucracy to generate the paperwork to release that 
item and we will ship it. The people at Optibac are ready at any time to put the curbs on that 
blank, I haven't released those curbs yet because we have the problem with the camera design has 
to be updated to the meltsheet data and it is kind of chancy to release the curves hoping that I can 
give up those degrees of freedom. What I am going to do is wait on into the fall and see how the 
workload in the optical shop is going and if we get desparate for something to do then I guess I 
will take a chance and release those curves and get the polishing of that bool. At any rate we want 
to get that blank out of the shop, because as long as we own it we are responsible if it happens to 
fall in half and as soon as we get it to Optivac then they are responsible. Let's see, we finished the 
Johns Hopkins lenses in mid June and so that job is not holding up anything in the shop and after 
that we started to rough and grind and we have now polished this ten inch test calcium fluoride 
piece. The idea there is that... what we are going to do is .. we put a big dome curtain on the thing 
and we found a concave ten inch diameter lense that we own and we are going to match those two 
and then we are going to couple them and do mechanical tests and thermal abuse tests and so on 
and so forth to test our whole coupling process. To test this silgard 527 material and make sure 
that we really can use that as the couplant. We are at a point now where Girard is experimenting 
with different polishing methods on the calcium fluoride to see what combination of pitch hard-
ness and diamond grit and carrier and so on gives the best finish on the piece of that size. We will 
probably be ready to couple I'd say within a week or maybe ten days, something of that order. 
Things are going kind of slow in the optical shop now because as you all know Dave's wife Dorie 
has a baby and Dave's done taken a low profile. That's fine because we are not in any real rush, at 
the moment anyway. #Who is going to do the coupling tests? #Well Girard will probably do the 
coupling and I imagine he and I will do some initial poking around, if the mechanical engineers 
are interesting in doing any tests we welcome your input. #How do we support this? #You mean 
financially or mechanically? #Mechanically, I mean now we have a doublet that is part calcium 
fluoride and part glass. #Right, well as long as it just sits there it can sit in a face-up attitude like 
that with the concave sitting on top of the convex and it could just sit in the shop. #Right. #Now 
the kind of test, the first mechanical test you do would be to push on the thing with a known load 
laterally and measure the displacement so you get some feeling for what the spring constant is and 
then you also want to remove that load after a period of time and see if the restoring force brings 
it back or whether there histeresis. Once you find out elastic constant numbers, then you can cal-
culate what you think is going to happen when you turn the thing on its side, hold the calcium flu-
oride and just let the other one just droop under its own weight and so we will see what happens 
there, then we will probably turn back flat and we will put it in the freezer and take it down slowly 
and hold it at a low temperature and bring it back up and see if any separation has occurred on the 
coupling material, and just kind of proceed in that test. #Should we try any tests with the mechan-
ical cell? #With a cell? #yeah. #Well, I don't have any particular tests with the cell that I want to 
do, but the mechanical engineers might have one, so great, that's fine. #Also, I'd like Erich to be 
involved in the mechanical tests if you would, ?? filled in some work before and it was good, I 
would just like Erich to be involved. #That would good to have Erich involved. #I have one ques-
tion. I guess Girard is kind of unsure of what the glass is actually made of. Do you think we could 
determine pretty close order what the thermal expansion coefficient is. #yeah, I talked to him 
about that, the zeroth order thing to do is to use some of Bill's index oils and try to match the 
index to a fraction so we can really narrow down what the glass is, it is probably some kind of 
flint with a faily high coefficient. If we can isolate what the glass type is we can look at a catalog 
to find out what the cte is. That would be a way to go about it. I suppose if all else fails we can 
actually try to measure the cte. Typical optical glasses that are not plurophosphates and so on all 
have roughly within twenty to thirty percent they have got the same cte, and #Is that all you need 
to know? #Yes. Actually Jack has a pretty good idea of perhaps measuring it using a profilometer. 
#Why do want to know it so precisely? #I'd like to know how much it is changing with respect to 
the calcium choride.. #You want to know what the differential is #..what the thermal stresses are 
in the couplant. #Because that is what will pop your couplant loose. #So the problem is by bad 
luch we could have in this piece of glass something that matched the calcium floride too well. 
#No, that is not possible. #Even fluorophosphates don't have as high a coefficient ... #Could you 
give us some numbers. #You mean off the top of my head? #Yeah. Calcium fluoride is high, is that 
right? #Calcium fluoride is typically about three times higher than a typical glass. #Ok, so the 
point is if the glass is going up or down by ten or twenty percent, relative to a factor of three, it is 
kind of a small effect. That's how I saw it, but maybe you would disagree. #I am just worried, 
because I think we are talking about a three thousandths thick couplant. #Yeah. #If your radius 
differential changes as three thousandths of an inch you have a forty five degree shear angle on 
your couplant there and I don't know how far you can push that couplant before it starts to sepa-
rate. #What does this got to do with knowing the cte of the glass? #I'd like to know it is actually 
shearing. #you can calculate all of the differentials, I had intended to do that of course and the rea-
son we use this concave  it saves a lot of money, we don't have to make a concave. We'll order 
zerodure and made one, but... #We just kind of want to know what it is. #Sure, absolutely, I want 
to know what it is too and we will find out. #There's a lot of things that we can think of as we go 
along too. I'll share some photographs, there is a lot of strain in this thing, you should pass those 
down there. #Right, that calcium fluoride may explode on us when we cool it or heat it or what-
ever it's.. the good part about it is that it was free.. #We got what we paid for.# The bad part about 
it is that it was free but its got all kinds of crystal planes and stress in it and God knows what, it's 
terrible. #And it's great because we know this is as bad as it is could ever be, we will never work 
with anything like this. #Never, never anywhere near as bad as this. #So but we may invent ?? 
harnesses and stressing textures as we go along, invent tests. #Girard generated this thing and put 
a polished surface on it. #great. #.. and it didn't come apart in the generator. #And it is giving 
Girard the chance to play with polishing material. #We think it is just a very good exercise to go 
through for a couple of months here doing what we can think of doing, so that when we get the 
real 25,000 bucks a piece bools, we will be comfortable with them. If the thing falls apart I will 
call up Rob Sparrow and tell us to ship us another one, a better one. Anyway, that will all be going 
on for the next month or more and I'd be glad to have Jack and Erich and Walt in it. Let's see here, 
we have got that and this. Another thing we have been doing we have been casting around to find 
a source of iron for the iron tools and what we need there is an iron that doesn't have porousness, 
an iron that doesn't have sand in it, an iron that machines nicely so we can machine the aspheric 
shapes on the grinding tools and we think we found an iron. We bought a rod of the stuff that is 
about five inches in diameter, yea long, and we can cut baloney slices off it and do tests, and the 
first test we are going to do is a combination of a test that will test the grinding characteristics of 
this iron and also test the wear rate, and what we are going to do is we bought some LAK -7 
which is the glass that has the most critical aspheric on it in Deimos and we are going to put a hel-
lacious big aspheric on this four and a half inch piece of about fifty thousandths of an inch of 
amplitude and we are going to determine what the relative wear rate is we will start with a known 
shape on the tool and grind the aspheric in and then we will measure the tool again, and the idea is 
if we can figure out the wear rate then we can differentially correct by making the initial shape of 
the aspheric tool slightly wrong in such a way as just as it finishes the aspheric grinding it comes 
to the right shape and kisses the part off with a matching tool. So that test is also going to be in 
progess, we are kind of doing these things in parallel and that will be going on in the next month 
or so. #Question Harlan. #Yeah. #We have used iron for tools before. Are we doing something 
different here than we did before? #Yeah, what we've done before is there is a lot of iron kicking 
around the machine shop and the optical shop old tools and we have just kind of cast around and 
we need a piece about ten inches in diameter, oh, here's one right here. And what's happened is 
that sometimes we get a piece that works fairly well, we don't have any idea what it's origin is 
and we can't duplicate it. Other times we've gotten iron that has sand in it and it scratches the 
glass or iron that has a lot of porousness and that tends also to scratch the glass because little 
pockets of abrasive get stuck adn then they release. #Right. #This time what we are trying to do is 
identify an iron that really works and we have a source of it so we can continue to buy it in the 
future.#And this business of the wear rate and the differential correction is also a new approach? 
#Yeah, it is something that we have known that we could probably do for some time, we have just 
never got around to doing it. Usually we have been making aspherics when we are in the heat of 
battle when we have got to get the job done, we don't have time to experiment, and this is a nota-
bly different situation here, we do have a few months were we can do these experiments. #This 
cast iron that you got for the test is small, but you are going to get something bigger. #Yeah, that 
same iron is available in a variety of sizes, we just bought a small piece to save money.. #Neil said 
before he left that we should get it and let it sit for six months and age. Is that a good idea do you 
think, because if so we should get the bigger.. #Well, I really don't know. Neil seemed to feel that 
that was the best thing to do with iron rather than trying to have it anneiled, I am not enough of an 
expert on metals to know. Maybe, Dave do you know? #I was going by Neil's experience. #I 
guess what we will do is as soon as we are really comfortable and sure that that is the iron we 
want there is no reason not to go ahead and order the parts and let them sit around, I guess it won't 
hurt anything to anneal them. #But you are not going to use these five inch ones on the larger 
aspheres. #No, that is just to do this particular test. Let's see know. Dave has mapped out the tool-
ing, the various bases and aluminum backers and polishers and grinders and so on. He's got a 
whole laundry list of that, I think his list is finished. I don't know whether he has released that to 
the shops, #Not to the shops, Carol is actually drawing them all out. #Oh, Carol is drawing them, 
Ok, so that's where we are there, at some point not too long from now we can release them into 
the shops and the shops can begin to construct those various tools and bases as they have time. 
#So that is something that will keep the shop busy. #Not a big deal, it will not keep the shop 
busy.#That's just a good back up project when the machinists don't have anything else to do, they 
can throw those on and #Does that include our aspheric tools? #No, these are backers and polish-
ing tools, such as for example if you are going to polish a piece you need to hold it, you need a 
base of some kind. And then the question is can you use that base also maybe to put pitch on and 
use it as a polishing too. Part of the job is for the somebody, is this case it is Dave Hilyard to map 
out what all the bases and polishing tools and grinding tools are and to get those ordered, make 
sure you don't have duplication because they are nontrivial in cost, if you have a piece twelve 
inches in diameter and you start with a three or four inch thick piece of aluminum and you have to 
hog out all the metal and so on, there some expense there. So we try not to duplicate any of that. 
Anyway, that is going along in a routine way I think now and unless we find some unanticipated 
thing I don't see any problem there. I think it is a straight forward thing. #Yeah. #Let's see, 
another thing that needs attending to that is very major and that I haven't done anything on is this 
full question of qualifying the profilometer. The profilometer is sort of limping and was just 
barely good enough to get the Johns Hopkins lenses finished but there tolerances relative to what 
we need on Deimos were fairly lax, so I still have to address the question of qualifying the profilo-
meter and unfortunately that is both a technical as well as a political issue. The politics of it is the 
part that is stickly. #I thought you had an agreement that you would inherit this device and you 
would have total control over it. #Well I have had a lot of agreements they sort of come and go, 
Sandy depending on the time of day and what's political and I get .. let's say at the moment I don't 
have any agreement that I am sure will in fact will be what happens. #This is largely related to the 
optical beam, or it relates generically to the whole machine room. #It's just the political issue, the 
way the politics with this place runs, #But what is happening with the profilometer now? #Well 
right at the moment it is just sitting there. #Nobody is using it. #Nobody's using it as far as I know 
but for all that I know somebody will walk in there tomorrow and pull the head off, I don't know. 
I have no idea from day to day and week to week what is going to happen there. #At some point in 
the near future, you and I and Joe and Dave Hilyard when he is back, and whoever else is 
involved need to gather together and determine what needs to be done and who is to do it and the 
schedule. #Is there an optical beam there now? #The optical beam is up at Mt. Hamilton, nothing 
will happen to it. We are still building parts. #We are still waiting for materials. #But the optical 
beam is not going to be mounted until we have this meeting. #And when do you think this meet-
ing will take place? #We are more or less waiting until Dave comes back because obviously he is 
a very key part of this. #Right, and do we have any notion of when that might be. #Yeah, this 
week, the meeting could occur next week, #as soon as next week. #It is not anything that is going 
to go off very much farther because some people want to get this optical beam project going too. 
Again, nothing is going to happen until we all gather together and talk about this thing and deter-
mine who is to do what and what the schedule is and what the impact is. #Ok. #As usual, this is 
teh first time I have heard of this and you know I am completely out of the loop and nobody sees 
fit to talk to me about anything so it is very difficult for me to conduct myself over there. Well 
Dave certainly is aware of what is going on. #Well he may be and you may be and Miller may be 
and fifteen other people may be, but I'm not and that's.. #As a result of this meeting Harlan, start-
ing then and thereafter, you will be solidly locked in to the loop. #Well, I doubt it. I think the 
approach that I am going to take is that I am going to assume that the profilometer is a completely 
unknown thing until I actually begin to need it to do work on Deimos and at that point I'll play 
like a shortstop and field the ball that's hit and hope it doesn't roll between my legs. #I gather you 
have a raft a tests that you are well aquainted with and intend to run. #Well, I have the tests that I 
have been doing for a number of years and ways of proceeding, but I don't have any control over 
the thing. #When would you like to start running these tests if life were paradise and you could 
arrange things to your liking. #Well, #How soon would you start. #Probably never. There is no 
way to answer your question Sandy because the profilometer is not under my control and it never 
will be under my control as long as the department is administered the way it is administered and 
that is just the way it is. #I asked a schedule question. I would like to know from your standpoint, 
you would like to do this work. #I guess I will start a few days before I begin to need the profilo-
meter to try to do these aspherics in Deimos and. #Only a few days?, I would think you would 
start well in advance of that. #But if I start well in advance, for all I know somebody will come in 
here and tear the thing apart or change it or they won't.. I don't know what is going to happen. 
#What would you like to do though, what kind of lead time would you like to give yourself. #That 
depends on whether I can control the thing or not. #Yeah. I said.. #I don't want to fix it today and 
then not need it for a month and then somebody comes in a week later and tears it apart or changes 
it. I don't want that. #Your concern is very clear. So I said assume the life is a paradise, what 
would you do? You would give yourself a lead time of a couple of months, or what. #If life were a 
paradise I would have absolute control over that thing, not a thing would ever happen in there 
without my ok. #I understand. # Next thing I would do is go in and rip that error system out, it is 
completely ignorant to have an error system rattling and banging in an enclosure where you are 
going to measure millionths of an inch. And I'd do a lot of things to that profilometer if it were 
under my control. But it isn't. #Can I ask the question of the schedule yet once more? #I can't 
answer it, I am not being obstreporous, I just can't answer it. because I don't have any control. 
#When does the schedule say that we are supposed to start the aspherics? #We don't have a sched-
ule for that because. #I can't believe that. #Well Dave might have it, but it is so far in the future 
that I have no idea. #We are supposed to have delivery right around the end of the year.. #Of the 
glass? #yeah, I think there is a one month time when you are going to do some weeks on the opti-
cal prescription, and just about february we should. #It's months and months from now though, 
and the glass order may slip and so it is on into the future. #I hasten to point out that the camera 
and production of the optics is on a critical path. So. #Ok. go on, we have heard about the profilo-
meter. #Ok, on camera and optical tooling, I think that's what I have got. #What is the situation 
with the collimator? #You want to do that now? #Sure. #Brian has been working on some soft-
ware for the purpose of mapping the detailed footprints of the various beams as they go through 
Deimos to see what they look like and whether they interfere with any of the mechanical struc-
tures or anything. We just need to know what the paths are in detail to kind of cross check against 
the output of the mechanical engineering software. So he has been doing that. Part of that has 
been motivated by need to know what those footprints look like on the collimator. The blank is on 
order and I don't know when it is going to come in. #August 11th, it should go to Kodak#And I 
presume that Dave is going to start swinging that blank when it arrives here and he'll probably 
grind it, polish it. My predilection would be to try to do that collimator without the use of the pro-
filometer, it is a concave, there is no reason we need the profilometer but that is Dave's call, if he 
wants to test with the profilometer, he can. #And the optical tolerances on it are quite a bit less 
than the camera optics. #Yeah, that's right, it is a fairly straightforward thing for him to make and 
I don't anticipate any particular problems. #If he doesn't make it nicely it will introduce optical 
distortions for astrometry, that is my main worry. #He'll do a good job, it's a straight ?? make I 
don't have any anticipation that there is any fly in the ointment there. #It is hard to get an optician 
to make something un nice. #It is not going to be near as hard to make I don't think as that spher-
ical mirror was on the hextet blank on the HIRES. That was a harder thing. #Let's see here. As far 
as the hole size in the collimator and the question of putting the sleeve with the davidson and so 
on I had a conversation with Erich and with Dave and I understand that the mechanical details of 
the mounting are fairly complicated back there and it may be inconvenient to do what I had envi-
sioned and that's ok. I am not married to the idea to having the Davidson there, but we do need 
some way to line up the.. first of all to know where the optical center is of the collimator and 
where its axis is relative to the mechanical dimensions of the collimator and secondly, some way 
to survey it into the Deimos instrument and put it in the proper place and know that it is there and 
to be able to check that it is there from time to time and I am amenable to any method that is com-
patible with the mechanical design, so I think the kind of shoes is on your guys' foot to make a 
suggestion as to how you want to do that and then we will go from there. #Yeah, we thought per-
haps we might put a reference flat on the front surface of the mirror outside the clear aperture. 
#Outside the clear aperture. #And then we could measure it from there knowing several spots.. we 
know where that is from the optical axis. #You mean to polish a flat out there, somehow or other. 
#That would be done by Kodak, not here, and I asked Dave to get a quote from Kodak. #What 
would we do, polish the mirror first, and then send it to Kodak and have them polish ... #No, 
Kodak would have to do while they were figuring it. #During the generation process. #Wait a 
minute now we are going to do the figuring here. #I am sorry, I used the wrong term. #All that will 
do is that will put a face flat on the mechanical blank, but where the optical surface ends up rela-
tive to where Kodak puts the generating is... #That may be different, and we will just have to mea-
sure what the difference is. #We would be able to measure that don't you think? #When we are 
finished, yeah, we will, sure. #So then we will have a subreference, which is this flat surface. That 
is our thinking. #What is this a rim around the outside? #Outside the clear aperture but inside the 
blank's diameter. #Usually the way that is done is in the reverse order. What you do is you first put 
the optical surface on and then, having done that, you put that optical surface on a rotary table, 
line it up so its axis is coincident with the rotation axis, and then you bring a grinder in and take 
you face cut. Then you know that your face cut is square to the optical axis. #Yeah, that seems 
more logical to me, why don't we do it that way? #We don't have any way to do it here. #It 
requires a trip across country. #We would have to ship it back. #Is that the quote that Dave got? #I 
don't know that he has gotten one, but I asked him to get one. #To do it afterwards? #No, to do it 
while they were generating it. And we realized it at the time, while we were doing it that this flat 
would not be perpendicular to the optical axis, but we thought we could be able to measure what 
the difference is. #Well we may end up being lucky, if we are lucky it'll come in pretty close and 
we won't have a problem there. I guess we will just kind of play it by ear, and we will just proceed 
and see how things end up and we will do in the end what we have to do to make it right. #I am 
comfortable with that. #Well we are proceeding with the flat idea, if it does have to take another 
trip across the country I guess we can do that if we have to. #This sort of highlights the point that 
we never have had a collimation report, or a collimation plan from end to end. We have had a cou-
ple of meetings, we have thought about it, Jack took a stab at what he was thinking, this is some-
thing that has clearly evolved with time, you wrote your thing before we started to do the 
tolerance document, which I think with Eric's help had some impact on how we thought we 
would do alignments. So, I don't know, I am a little concerned about this, is it time to think about 
it? #We are not as concerned. #You're not as concerned? #Because we know what the hardware 
looks like. What we don't want to do is put the hardware in the wrong place, so if we have already 
designed the hardware and had the review, I think we are happy on that day, maybe we want to 
open it up again. #This is the hardware for the cell. #The collimator cell itself. #Another one sits 
in front of the glass and now you have to measure the glass through that hardware. #I think that 
Sandy is going beyond that, saying a collimation report for the entire instrument. #End to end, 
you mean all of the elements. #Because even your document, which began to think about this 
question stopped at the critical point of where we put the grating in. Didn't get past the grating. 
#So that would be something that Brian would help us with. #I don't know, I think Harlan should 
give us the benefit of his thinking. I have a feeling Harlan that you have a lot of ideas in the back 
of your mind and nobody knows what they are. #Well. #Like for example this idea about the 
Davidson ??? [End of Tape]

Tape four

#..if we had thought about it from the very beginning.. #Well, I thought I have been talking about 
that for years. Jack certainly knows about it he used it all over HIRES and probably several instru-
ments before that. #But we couldn't figure out get your head in there, we could get your auto col-
limator. ??? is way inside this instrument. #Haven't they got those eye extenders and ccd things 
with fiber optics and whatnot. #Now you are talking about a different piece of equipment, it might 
be less ??? #See I never knew that is was supposed to be attached to the collimator mirror, I just 
thought we just had to look through there. We could look through there. #Is that a true statement, 
you are worried about access for looking at it? #Yeah. #You can buy ccd cameras that you can put 
in, you know, this big, these days. #That is what I just said, eye extenders. #But with a collimator, 
you actually have to get in and turn things on it. Its a manual instrument. #You can buy autocolli-
mators that do what you are describing. You don't have to take the Davidson that we own and 
make it into a modern instrument. #No, we can buy another Davidson that's completely auto-
matic. It costs money, I am trying to get ?? done cheap. #How much does it cost? #I think they are 
fifteen thousand dollars. #You are looking at a screen and you are turning knobs here to focus, run 
through the focus and checking all of these things along the way. We don't have one of those 
things, but we'd love to have one. We had been using just a person putting their eye there and 
turning a knob for a long time. It's a cost saver. #The nineteen fifties model. #Yeah, but you know 
then we start talking about shipping things cross country, and so I mean its a trade off. I begin to 
wonder whether we are making the right decision here. #I guess it all starts at the budget, when 
you put in what you really want to do and something comes back and says, that costs too much, 
make it cost less. #That area of the hole is kind of difficult, as Eric explained the mounting to me, 
it is kind of dynamic in there it isn't just a big hunk in there at all, there is a dynamic part, a coun-
terweight. In retrospect I can see that to put the Davidson in there would require a completely new 
mechanical concept for the mounting, it probably isn't the right thing to do the change that con-
cept right now. So I think we are all right by taking a different approach, in this case. #Alright, 
now let me raise my broader, more general question of a collimation plan. Is it too soon to think 
about it. Would it be good to have a task group of people working this question? #Well, that job is 
part of the mechanical designer's job to do that and I would not want to interfere with the mechan-
ical design, when a guy designs something he is going to put a lens here, he has got to figure out  
how to put it there, how to measure it, how to know it is there, and if he needs a little reference 
service or some register or something to bolt on he just automatically says, oh, I need that I'll put 
that there and puts some holes and more. The mechanical design.. #It's true but the collimation is 
an interesting combination of mechanical issues and optical issues and there are collimation 
tricks, since this is my first instrument, that I think exist that I know nothing about. So, I have 
always felt at a disadvantage with this subject. It seems to me as though we need some melding of 
skills, #and I guess I have always worked the other way around with the optical designer has 
given us a pretty good idea. #Or at least an outline of how we do it, and then the mechanical peo-
ple worry about where the put the fixtures and stuff like that to make it work. #The optical design 
shows you what the parts are and where they need to be, now they just exist kind of in air. Now 
the mechanical designers job is to put the mechanical structure around that that holds those parts 
in those places. #Yeah, but to line things up is a combination of mechanical tests, as I am not com-
ing to understand and optical tests, and there are standard techniques for optical collimation that I 
feel I need to know.. #What you are saying is that those standard techniques and technologies for 
optical testing are not commonly known, but then I can help with that, I know a fair amount about 
that subject. How about Brian, how is he on this subject. #He doesn't know anything at all about, 
he never got into anything like that before. Jack knows a lot about it, #Well, we're collimating, it 
means putting everything in line, this is an instrument that has no center. #Well, it takes special 
tricks. #Generalize the idea of collimating to putting the things were they need to be. #The only 
thing that needs to get collimated is the collimator mirror, and you use your standard tricks there. 
After that then you have to redefine where the center is because the tent mirror has no center and 
the next thing you run into is the grating and it has no center except the pivoting axis a mechanical 
construction. And so those two elements are going to be hard. Then the camera itself is a unit that 
is collimated by somebody else or by standard techniques, many people can collimate the camera. 
#You just have to put it in the right place. #make it aim at your center line of one half of the instru-
ment, which you have to define mechanically, there is no optical center. It your flat mirrors. It is 
not too early to start on that, but you have to break somebody loose to follow that through, from 
some other job. #Usually if you go into industry where they are building an optomechanical 
device, part of the drawings will include the assembly drawings, that show you from first part that 



you start with until you have got the whole thing put together. Show you all the steps that you do. 
And part of those steps, are step ninety four, you bolt on this fitting here, and step ninety five, you 
put the Davidson in there, and step ninety six is you look down and focus on that mirror and so on 
an so forth. #Yeah, that is the plan we need. That is the kind of plan we need. #And you develop a 
plan like that, you have to say, oh, I need a register at this point and I need to be able to see from 
here to there and so on. #I guess my worry is that we can certainly come up with a plan, but in the 
end the optical designer has to tell you what you designed in there was right. I have always 
approached it the other way. #I am happy to come over and look at what you got or..#to where the 
optical designer says you that you got to mount the collimator and you know, the misalignment 
can be within this range and then we are fine. #Do you think that a collimation plan is legitimately 
part of the CDR? #We should have a pretty good idea of how we are going to do it. #I think so. 
Ok, why don't we put that as a task item for the CDR, and I would suggest that we nominate three 
people to start thinking about this, or four, and Harlan, I think you should be one of them. #I guess 
I should be one of them, #I think I should be one of them and Jack and Eric, we could choose 
either one of you, but it seems to me that everyone has to sign off on this. #They are intimately 
involved in mechanical design.#Right. #And maybe after we get going, Harlan, you can bow out 
and Brian can take your place, so we might bring Brian along in the beginning to get familiar with 
these ideas. #He needs to come along to get initiated, as he has no experience with these instru-
ments and techniques at all. So I think we need about a two day retreat. #A retreat. #Which we 
sort of go away and thrash this out. When is a good time. #I don't have any time blocked out for 
anything in the near future. #Jack is pretty involved in the slitmask handler right now. #I would 
like to put this as far off as we could possibly do it without jeapordizing what we need it for. #I 
think september would be a good time frame in my mind. #Ok, lets think about mid september, if 
that is possible. #How about you Harlan? #I think I have got, let's see here. September, I am out 
of commission for the ninth to the seventeenth, #So slightly after that. #Right. #After that would 
be easy. #September 20tb, that's good for me. #Do you want to put a tentative date down? #I said 
September 20th. #So you think a two day effort. #I would guess two days. #What day of the week 
is that? #It's a wednesday. #Wednesday and thursday. #I sort of dread this. #That would be Har-
lan, Brian, Sandy, Eric and Jack. #Do you need a lot of space or just a little space? #Oh, a little 
space, preferably a tiny space. #Small space, no chairs. #I think I know why it is so dreadful, it is 
because it is because once you have got a collimator plan, it means you really understand the error 
budget and the tolerance. Your whole flexure compensation system. You understand what the 
whole instrument is going to do. #I would be very happy to get the collimation plan sans flexure 
compensation system, figured out. #That's the  ??? how do you put it together, and didn't have to 
do anything but sit there? #If things move around by a few microns, it is not going to kill us. #We 
may end up with only the framework in certain parts, same thing with our error budget, but at 
least we will know what we are missing. # But this certainly means that the error budget has to be 
done before we do this. That last twenty percent is killing me. #So, we can go back to software. 
#Ok, primary activity during this last period has been by Steve, working on various design docu-
ments in preparation for the upcoming PDR and meeting with ?? and others. I think I will let 
Steve talk about that and then I'd like to come back and address a couple of other issues ?? the 
software stat?? #I have a bunch of documents which are growing as I make discussions with vari-
ous people and check out other things and because some of those documents turn out to be rele-
vant for comparison with the Keck books in Hawaii and also with ?? I have also put all the web, 
so that distant people can see them as well as local people and get a look at the specifications as 
they grow. It turns out that our image storage scheme as I had already laid it out looked an awful 
lot like Archon storage scheme, and that is just natural because the only way that you can commu-
nicate the information in a sensible way is the way that we are both doing it., as to which parts of 
a readout image are overscan information which parts are image which came from which of the 
amplifiers and so on. #Archon? #Archon is a array controller readout device that they use at 
CTIO, it basically lets them readout two amplifiers off the same  ??? at the same time and later on 
more than two amplifiers on more than one ccd. But it is nothing that we haven't already been 
doing, its just that it fully documents all of the pieces in such a way that they are ?? script and 
immediately going baseline correct and otherwise chop the data and arrange it into what the 
astronomer wants to see without having manual intervention required and I am setting up a 
scheme that is pretty much the same thing where once the image is in, all the image processing 
can be done automatically without the astronomer having to worry about it. In a meeting with 
Dave Koo and Drew it became evident that with any display technique we have for the Deimos 
images we cannot possibly look at pixel for pixel images on a screen. There is no way you'll get 
an 8,000 by 8,000 display. Dave and Drew are quite concerned that that would make the job that 
they typically find themselves doing in the telescope impossibly because what they do is to manu-
ally pan over the image to inspect it for quality problems and that if the image is bend by four or 
eight on the screen and if it is 8,000 by 8,000 pixels there is no way they will have the time to go 
through and actually do that, so it became evident that on the way in, when the image is coming 
off the controllers and being assembled in memory before it is written to disk, we should be doing 
some statistical looks at the image. To count of saturated pixels, to look for 6o hz contamination, 
just make sure that the baseline is within a given range, to be looking for shifts and anything else 
we can think of that.. image quality problems that can be detected as the image is being built up so 
that the astronomers can just look at at little box that says, there are n saturated pixels, there are m 
bad columns, and there looks like or doesn't look like there is 60 hz contamination in amplifier 
number 22. And that should be doable before the image is flushed out to the archival procedure so 
that all of these little statistics can be in the pix? headers archivally. Even though some of them 
will be correctable and not in the data that is taken home. I have looked at the way in which we 
will be able to ... for quick look and in the long run have a mouse run over an image and say this 
is at Ran deck something or other and there are no schemes in existence really implementing this 
except in radio astronomy which have as an integral part of the image information, really precise 
coordinate information. In radio astronomy they have been using them for quite a while, and if we 
turn them on here we can do the same thing with the Deimos instrument and we should do it for 
all the rest of the images around Lick for that matter. When we do that, we break vista?? #Ah, this 
is something I didn't understand, I got my email and didn't understand. #Because we have been 
using a couple of things in fifth? standard for purposes other than which that is now commonly 
accepted that they should be used and in making use of those correctly so that our system can 
understand the coordinates correctly and anybody elses data reduction system would know what 
the coordinates mean, vista breaks. So I have mailed around to the people who are in the know 
about Vista with an alternate scheme which would keep vista going into the future with the old 
way and still allow us to start using ?? coordinate systems the way that everybody is now agreeing 
that ?? ought to be used. I am waiting for the feedback on that but I don't see any problem with 
keeping vista going in the short run and the long run and still ?? everything else to be right. 
#Steve, when your statistical shooter manipulates the data and finds what it thinks are bad pixels 
and 60 cycle hum and whatever else it finds, how can it alert the scientist so that the scientist can 
say, oh, I'd better look at that? #I am imagining that it will have a little box on the screen that basi-
cally says, image quality monitor and it will have those things we know we tend to go wrong, and 
we will turn red flags on if they seem to be abnormal. #But then the red flag goes on, and the 
astronomer says, gee I think I ought to have a look at that, then what does he do? #Then we go 
back to whoever is in the know about the engineering of the ccd controllers and make sure that 
some voltage doesn't need adjusting or some wire hasn't come loose and needs to be soldered 
back down, and that is the sort of problems that they are afraid of, even though the solution isn't .. 
#That is a different kind of thing, that is a hardware malfunction, but things like saturated pixels 
could be just that the 5007 line was too bright or that a cosmic ray came through and so on, it has 
nothing to do with the wires. # Rather than having to pan through sixty four million pixels, look-
ing for the ones that are saturated and trying to figure out whether or not they need to take the 
exposure again they want something that flags it immediately and obviously. #But then can they 
isolate on that piece of the frame and look at that on a screen #Sure, basically to be able to turn 
those pixels red, if they have gone abnormal in some fashion, so the astronomer can quick run 
over there and say, oh that is on this particular galaxy which I really needed to get good data, we 
had better take another one. #I see that the astronomer does have the option in almost real time to 
get in and inspect the suspected pixels, is that ... #Right, the astronomer needs help to find those.. 
#Right, just finding them, but once found he can jump it? #Sure. #That's what I wanted to know. 
#That implies then that your statistics mechanism not only gives you a count but has a button you 
can push that says, give me a list of coordinates where these things are. #Or else, it has a mode 
where the color map  what you do to the display automatically switches to a grey scale where sat-
urated pixels are colored red. You can then say, oh, that's red, its not. #The assumption then is that 
you would have a clump of these sufficiently large that even binned at a fairly large factor you can 
still see them. #I think it is easy enough to color them in a color that will ensure that, even if they 
are binned, they will turn out red. #Ok. #And even if the verse? is 21? They are going to have a 
helper. It's just my helper on the last run was ??? #And those are the issues that have gone over 
and come up big in the last few weeks. #Ok, let me talk about what the rest of us have been doing. 
We had a Keck II software coordinating meeting that was hosted here, June 23rd and several 
items came up for discussion there that relate to Deimos. One, we mentioned our current plan of 
using terminal servers to multiplex the serial lines from the Galil controllers and perhaps ?? up to 
twenty-two devices back onto a single serial coax that would come out through a Deimos rem. 
Cara apparently has been doing this at some level and in fact has been using terminal servers from 
the same ?? electronics that we ended up selecting. they indicated on occasion that they have had 
some problems where the terminal servers seemed to lock up. They are not sure whether this is a 
power problem or a network overload problem. So what I am proposing to do with the unit that 
we now have in hand is to put back the various loop connectors on it and to simply set some long 
term tests going pumping as much data through it hour and hour over a period of some weeks to 
see if we get any anomolous problems. #And presumably you can send it up to Mt. Hamilton to 
give it a more realistic test in an environment where things fail constantly. #It is guaranteed to 
work down here perfectly. #I don't see any problem once we get the test going down here to move 
the test to Mt. Hamilton. William Lupton also announced that the long awaited epix portable 
channel access server has been released in Beta test form and he is now trying to bring that up at 
Cara, this is actually sooner than I had expected it might appear and I am still awaiting William's 
evaluation of the Beta test version. #The beta test version of what, I missed it. #The epix portable 
channel access server. This is, up until this point, if you wanted to run epix, it dictated the use of 
the Xworks and the VME crates and the VME crate controllers and without buying in to that 
whole hardware and software architecture, there was no way you could use epix. The portable 
channel access server actually runs under a generic unix environment and is now I believe they 
have versions running under Sun Os, Solaris, also HP and they are working on the DEC Alpha for 
it as well. What this means is that the epix channel access messenging system now becomes avail-
able to talk to process running on the unix host. #Sort of painless. #Well, that remains to be seen, 
but if they done a good job on this, there is a potential we might want to consider using the epix 
channel access messenger system as an alternative to the sort of music messenging system that we 
used in HIRES and were planning to use in Deimos. I am hoping, perhaps by September, we will 
have this far enough that we will get a version of it here for evaluation because we definitely need 
to make that determination before we get to the software PDR in January. There is also consider-
able discussion on the question of image display standards for Keck II instruments. A number of 
possibilities were identified, I don't think any consensus was reached. George Brims, of the NIR-
SPEC has been tasked with taking a survey of Keck observers as to their feelings regarding the 
image display requirements, likes and dislikes with existing displays. We hope to have the results 
of that in by the next Keck II SCC meeting which is in September 23rd or something. #Uh, SSC. 
#No, SCC its confusing, the Software Coordinating Committee is SCC as opposed to SSC. #So 
has Deimos generated a response? #We have sent some comments, but I don't think we have col-
lected an official Deimos response to that yet. #This was a topic of the meetings between you 
David Koo and Drew Phillips in June, at least I thought it was. #Yeah. #Did you make progress? 
#Our impression was that there is no system that we have seen in recent history that does quite 
what we are going to need for Deimos, which is basically one, if possible, one full monitor dedi-
cated with a window that is equal 8,000 by 8,000 and which has a little pan box on it that you can 
drag around, and the output of that pan box is on another monitor. #Right. #So that you can pan 
around the whole image and see what you are zooming on. And that there hasn't been images ?? 
to it? quite like that since an experimental one SAO had about eight years ago. #So what are we 
going to do? #We have to take apart something that already exists and works relatively well so 
that we can split its little zoomer into a separate window, or build our own. #So whatever we say 
to George Brims doesn't matter because we are not in that loop. #No, I think there is an effort here 
to try and identify something that could support the needs of all the instruments, if you can come 
up with a proposal that meets that perhaps we have some of the convergence. #But surely no other 
instrument has as extraordinary a problem as we have. #They have other extraordinary problems, 
they are different extraordinary problems, in terms of bandwidth and readout rates of the ir 
devices and extraordinarily high frame rates compared to what we do. #I am just getting the feel-
ing that we are repeating history here. Remember the incredible efforts to get homogeneous 
approach to user interface, image display and so on so that somehow every instrument would look 
like every other instrument, it just was impossible because the instruments are different. #Well yes 
and no. In fact the one area we can consider some level of success as much as it has lots of warts 
and many people hate it, of the four existing instruments in use, three out of four of them display 
their images with figdisk?? If we could even achieve that level of commonality among the Keck II 
instruments, I think that is an advantage to observers going between instruments. I think it is a 
very big advantage and I think Cara is extremely anxious to get as much commonalities as possi-
ble because otherwise it is a support headache for them, its a difficulty for someone like Tom 
Bede to be totally up to speed on a half a dozen different image display systems and so on. We 
will give it our best shot and we will see where it goes. Ok, other things that bear on Deimos that 
are not actually activity related to Deimos is this month in particular we have been pushing fairly 
hard to tie off committments to other Mt. Hamilton related instrumentation, primarily LERT? II 
and MOS to be observing runs. LERT II this week and MOS the end of this month. We have made 
fairly good progress on this and Dean Tucker, who has been almost full time on LERT II for a 
couple of weeks is now got that almost done. It is an up and running telescope. The other signifi-
cant announcement is the position of a programmer dedicated to Mt. Hamilton instrument soft-
ware support is finally moving it is in recruitment. The advertisements for that will appear in the 
paper next week and the week after, the closing date on that is August 21st, which means we will 
be interviewing for that position in early september. We will have that person on board probably 
October one. #What kind of person do you envision there. What kind of a degree does that person 
have. #We are looking for someone with a background both.. [End of Tape]

Tape five

#..so we will be posting this also to the web, we will be sending it out to observatories and hope it 
will generate a lot of interest. Once this person is on board a lot of the activities that are currently 
are occupying a good deal of my time, Dean's time and even at some level, Steve's time will 
migrate to that person, meaning that we can... #And hopefully my time. #And Richard's time as 
well. #I am buried under work. #Really I didn't know that. #.. focus on our related tech projects. 
One final note is that we have also had the beginning of the July run a somewhat reorganization of 
the observatory org chart relating to the software group. Previously we had sort of two groups, 
myself, Dean, Lee Rottler, and Steve Allen that have been directly primarily to instrument sup-
port, Keck projects, data reduction, science types of things as opposed to more computering facil-
ity management, network information systems and so on. The group of Dee, Ted Asox, Eric 
Carter, John Tomas, Jessica Elner and up until July one Dee reported to me, so all of the things 
pertaining to that group at some level would come back and I would have to deal with. As of July 
one, I no longer supervise Dee Clark and she is reporting now to the directors office. So I am hop-
ing that will also allow me to focus more of my energies to Keck related ... #So we have more sci-
ence computing and administrative computing. #Well, I am not sure administrative is the word to 
put on it. But computing support more focused on Santa Cruz operations, although Dee's group 
will still have some component of support for network connections at Mt. Hamilton, but that we 
have a little clearer division of science instrument related activity versus other activities. Dee still 
will have impact on science things in terms of archiving scientific data bases. It makes for a more 
level organization chart, it you started looking at the depth of how far down we went in the soft-
ware, it's a rather vertical looking thing I think we will function more efficiently with this scheme. 
So, basically I end up with a net change. I lose supervision of Dee, but I will be picking up super-
vision of the Mt. Hamilton program, so basically I will stay the same. #Could I say one sentence 
about software if you are finished? #Sure. #Cause it sounded like you were. Archiving comes 
under the heading of software and my job this month has been to write a proposal to the National 
Science Foundation to support the phase one of the deep science program which is being done 
with Low Res. So technically that doesn't have anything to do with Deimos. But in a sense it 
does, but we have decided to take the rather unconventional step in this proposal, mainly to make 
us look more attractive and give us a better chance of being funded, to put the data into a public 
archive after two years. #Now these are low res data so they are only 2k by 2k, and the thought of 
solving this problem for the Deimos 8k by 8k data is horrendous, I don't think we could do it at 
the moment. We can certainly contemplate doing this with the 2k by 2k data. So I have been talk-
ing with Steve and also with Dee Clark and Drew Phillips trying to flush out a plan in enough 
detail, say in a couple of paragraphs of a proposal what this might look like and I think this is sort 
of being the observatory's first toe in the water of public archive if we get funded. #Good. #I am 
going to ask one of the questions somewhat related to this but it touches on a number or areas and 
that is the garter which is being developed for Mt. Hamilton. Joe has talked about the ESI the pos-
sibility of using that for Keck for ESI. Finally I was also talking to him about, it is not clear what 
is happening now with what the company, Photometrics, Dick Akens is out and now it has been 
taken over by somebody else. #Oh, really I didn't know that. #Yeah, it was bought out. And also I 
am not sure that it is particularly great guide I am not particularly happy with them. What would 
be involved in making our guide of the Hamilton one operational with Keck. I mean is there 
major software interface changes that are required to do that? #Yes. But I presume you are talking 
about the new guide camera that Wei is. #The computer interface to that hasn't even been fleshed 
out, isn't that one of the projects we are going to put the new electronics engineer onto as a sort of 
a test. So there is at least an opportunity to address the issue of making it easier to incorporate into 
the Keck system. Since we don't have anything down now anyway. #That would be very...#On 
the ESI project, at least we waved our hands in the air and nominated a chunk of money to do this 
and it was just on order of #How much? #Twenty or forty thousand. It was on order of the cost of 
the camera. #This would be wonderful because, my axe to grind is I figure that if we are doing all 
of this in the interface we will de facto have a mucher bigger role to play and greater leverarge 
over exactly how the information is displayed out there and how the telescope operation and 
astronomer interact with it. Maybe I am wrong in that. #That unfortunately that involves a lot of 
the Cara side which is already in place, but I hope that we could fake that at least. Ok, we 
shouldn't take too much time with this I just wanted to clarify what status things are that. #Hope-
fully the rest of this will roll along pretty quick, do you want to talk about the structure. #Well 
there is not really a whole lot to talk about the structure, not much has changed since the last time 
we met. I am starting to make some minor changes ?? will know more about the base comple-
ments that are going in there but the basic scheme hasn't changes. #And we sent off yet another 
package to Frank for him to review. #Right, the subject of that package was, I have it in my hand 
here, but .. #Just progress on the various mechanical compounds. #Mostly, slitmask handler grat-
ing and tent mirror. How is the interaction with Frank going, what did you get back from him on 
the previous package? #A telephone call. Not too much, we are not expecting a lot. Mostly we are 
expecting him to go through and pick out anything he sees as a red flag and concentrate on that. 
#Do you get the feeling he is thinking about it. #Yeah, he reads it all, it's kind of a no news is good 
news deal with Frank. #What are you thinking about if he thinks that needs to be said? #It is a 
question relating to whether he was actually reading. When he called, he called you or Dave or 
what? #He left a message on my machine and talked to me. #So it was clear he had read it and 
thought about it and just... Ok #Where are we on a prototype for the slitmask changer? Were we 
going to build a prototype and run it? #Yeah we are building a prototype full size that is mounted 
to the turning prototype drive disk, the new eight foot diameter drive disk. This is not the changer 
part yet. #No but the caterpiller part. #Yes the caterpiller tent position handler and we have got a 
counterweight on the other side. Where we are fabricating it and making sure we don't have any 
surprises. the chain works sideways and that's because the grating handler runs into that device if 
we don't do it this way. So that is where we are and why we are there. We have taken a little of a 
hit because of the line freeze and other projects that have been in the shop like the Keck I second-
ary. But now we are up and running. So hopefully by next meeting we will have photographs of 
that. It will be manually operated but putting a motor on it is not important. We did talk about 
Barry mentioned how to position, that's not the right word, how to encode such a device, it's 
unusual. I think we have a plan for that. But actually extracting a mask out into the field of view 
doesn't impact the grating handler so much, so I have left that as a separate project. Remember 
this caterpiller is the thing that is moving the drive disk around. It's either we back up from the 
telescope to fit in the cell. #The new function was to move the drive disk.  #What we want to do is 
freeze the design, and the drive disk is something you want to locate and not move it later. We are 
working on something in the way of structure by working on these other components. #When do 
you think a prototype will be done so that you can start manipulating it. Or are you already done? 
#It's done to some level, we haven't mounted the holsters on the chain yet but we have the chain 
and you could go down and take a look at it. #Yeah, I'd like to do that. #One week, two weeks, 
Dick is back on that now. #It is probably seventy percent complete. So maybe another two weeks 
it will be... #It will be two weeks. #Move on to something else. #We will leave it in place, this is a 
model that we will keep playing and with getting ideas from.  #But you're welcome to come 
down and we will show you what we have. #Is this the place to ask about the rotation tests that 
were supposed to be done with the HIRES image rotator? Remember we were going to... #Yeah, 
you can ask, we haven't done them yet. #You still haven't done them! #No. #They were supposed 
to occurr is May. #Because the rotator doesn't have a huge priority right now because the glass is 
still at Corning we believe. #Ah, I see, Ok. #I am just letting it slip along a little bit. What we did 
is we took it apart and coated it all, so now it is coated and it is being assembled. It will be reas-
sembled probably by the end of next week, and move over to software. What we really need is 
this little terminal server, so we can ?? it in the vault, so that is pretty much done and ready to go. 
Once Dean knows how to talk to that and we can drive the rotator then we can put it in the vault. 
Ok. and we have fault?? with the money?? so it might cost?? you?? Gratings? Do you want to say 
something? #Nothing has happened there as I said because the grating is running into the slitmask 
handler so I wanted to do one first to make sure I really wasn't going to run into it, other than to 
make an inventory of fifteen future and present gratings, mirrors and filter mirror things, narrow 
band imaging things. Actually write down what they are, where their zero point is and what range 
of travel we expect. And the grating handler then has to handle those things. So we are defining 
the problem and that is the status. And I have asked the question of the astronomy community 
what do they do with large gratings, how do they handle them, how do they rotate them and sup-
port them. #You are not getting too much back on that, I gather. #Because there aren't many peo-
ple out there using 8 by 10 or 8 by 12 large gratings. The next step would be to go to Milton Roy 
and see who they sell them to and actually go to those people personally. #Did you see this mail 
message from Rem. He was visited on Mt. Hamilton by some people who make very large grat-
ings. #Who are these people? #I don't know, i have this mail message I will forward it to you, I 
assumed that he was forwarding it to .. oh maybe he didn't. #This is a US based company? #Yes. 
#Not Milton Roy? #No. #Is it Hyperfine? #I don't know. #Would you send me a copy of that also 
Sandy, please? #Yeah. #That's a Boulder Colorado firm? #Yeah. #We looked at them for HIRES 
once. #Is that about it on gratings? #What about the question of the inventory and the decisions as 
to which sizes and which numbers of lines are going to be bought up front and when they're going 
to be ordered and all that kind of stuff. Is that in the process or scheduled or.. my concern is that 
once all that is decided we have got to get the blanks made, we have got to ship them to Milton 
Roy we have got to get all the rulings cut and so on and so forth. #That does take a long time.  
#That can take, you can slip a year there easily. Milton Roy has a schedule. #There is a schedule. 
#I thought that Milton Roy supplied the blanks these days, you basically send in an order. #Well 
they will if you like, but for HIRES we supplied our own blanks. I don't know what we are going 
to do for Deimos, that is my question, has that been thought through and do we have a plan? #Yes 
there is a schedule for sending the gratings off, and of course it escapes me. But anyway. #I would 
think we don't want to douty on that. #I'll have to find it., but my recollection is that it is early 
next year. #Early in the coming year. #Ok. It's not too soon to start thinking about. #Sandy you 
have got to make the decision as to what sizes and what numbers of lines and so on are going to be 
in the initial complement. I presume you are not going to get all fifteen of them all at once. 
#Right, no I think we have a pretty good idea about what we want to buy up front, but we proba-
bly out to discuss it with you. If that would be a good idea. We have made some aluminum parts 
that are identical to the blanks that Dave wants, so we could put something into our rotator, #Oh, 
good,#That is the same weight and size without putting the grating. #They are currently holding 
my desk down. #They are doing what? #Holding my desk down. #They are heavy. Feel the urge I 
will lend them to you. #Maybe I ought to come over and lift them up and down a few times. 
#Does Hyperfine have a catalog, are they folks that now have a basic set up .. #I don't know the 
answer to that question, I haven't really had occasion to look into what they are doing the last sev-
eral years. But they are an alternative to Milton Roy, and for example they have been around a 
long time, they got the contract on the faint object spectrograph for the space telescope. I mean it 
is as long ago as that. The point is that they have a long track record. And there is also this com-
pany in France JNB, and they make all manner of gratings, both ruled gratings as well as holo-
graphic gratings. #Are you writing these down Dave. It sounds as though you are going to send 
off for catalogs. #And I've been hearing some really remarkable claims about the efficiencies of 
holographic gratings, recently. #Yeah so have I. #I don't know anything about holographic grat-
ings personally but I am thinking if the claims are even half true then we ought to be looking at 
them. #That is a good job to do. #They are sort of restrictive in dispersion is one of the problems 
with the holographic gratings. They tend to be higher dispersion than we would want for Deimos 
as far as what I can tell from what I know of the Milton Roy holographic gratings. #No we could 
use a higher dispersion grating, I wouldn't ?? it out of ??. #That might be an artifact of what they 
have been asked to do. #That may be true. #What their capabilities are may be quite different 
from the products we presently have on the shelf. #Can I ask a totally dumb question about holo-
graphic gratings? #Sure. #Do you shine light on them and the light goes in the same basis direc-
tions as a normal ruled grating. In other words can you take a grating out of Deimos and put in a 
holographic grating and have it work? #Yes, you can. They look for all intensive purposes just 
like regular gratings, just the physics of how the light gets diffracted is a little bit different. I have 
some in my collection of toys, I have got the second holographic grating that was ever made on 
the Earth, that Rudolf Kippinhan gave to me. The way he got it was he had a ne'er do well astron-
omy student who dragged on for years and didn't get his thesis done... the reason being the guy 
was really interested in holographic gratings and he finally figured out how to produce them. He 
gave a couple to Rudolf and Rudolf happened to be in UCLA, this is many many years ago and so 
he kept the number one and gave me the number two. I've got it, its a little guy about this big. 
#Ok. #Well I guess I should be the one to say that we have to freeze this decision pretty quickly 
because I can't make a handler handle everything in the world and we are talking about quarters 
of an inch now so, if you go to other companies and their blanks are not 8 by 12, if they are 10 by 
15 or something. #We don't care if it's smaller, we can pad it out. But if you care if it is bigger. 
#That decision out to be frozen pretty quickly, it can't just keep forever. #And twelve inches is the 
maximum length you are designing to at the moment. #The thing called 8 by 12 is actually 8.66 
by 12.7 to 12.8, so we can look for a little while, but you probably want a date, like september 
first, #I was thinking more in terms of better ?? is higher efficiencies, whatever a bigger range of 
ruling numbers, all these parameters, not so much in terms of bigger. #No, but the point is that if 
these other companies have different standards we need. #Then you may want to start at a differ-
ent angle. #Then I will try to contact these guys and. #Yeah, if we got catalogs from them that 
would be good. #Oh, I think we talked about slitmask handler, #A note on slitmasks themselves, I 
was in Durmet? at the end of last week and talked to Jeremy Ellington Smith who is one of the 
PI's on GMOS, which is Gemini Multi-Object Spectrograph and also designed the LDSS II spec-
trograph that ?? and was shown an LDSS II slitmask which was milled and I didn't get a chance to 
look at it under a microscope, but it looked very nice. In fact it looked even better than the mills 
that we produced here for the LRIS. I was just going to suggest that somebody out to get in con-
tact with Jeremy and see. They use a commercial milling machine, produced by Rowland A.G. to 
make their mills for the initial LDSS II masks, but they have changed their milling machine since 
LDSS II came on line. I don't know what they are using now. Jeremy didn't know what they were 
using.  But they are probably using much the same process for GMOS. #This is interesting 
because since this is a smaller telescope their slits are smaller, so I would think that they have 
probably encountered the same size problem that we have. #We are probably going along exactly 
the same line. The way we milled the LRIS slits was with our existing NC machine, which has a 
relatively low velocity quill on it, so we are putting a higher velocity quill on it. #What is LRIS 
doing now, I understand they #They use a punch. #They punch these things but you actually 
milled some that they used. #We milled I think four. #Many more. #As kind of an experiment 
then? #Yes. #See whether .. #the greatest fear in my life is that we will mill them all. #That's why 
you are introducing a few errors into it. #When you compare them under the microscope, you 
know there is no comparison, the milled versions are much better than the stands or the EDM or 
the laser cut, what have you. #You don't need a microscope to see that. #Yes you do need a micro-
scope. #Not for punched versus milled. #No, its so obvious. #What about the amount of time to 
mill them. #We are looking into all of this. What's happened is that we looked at the slit quality, 
the edge quality in particular of three different sources, lets say CFH laser cut, laser cut at ART 
and laser cut by directive light and the best of any of those was by CFHT and it was about 89 
tenths of a thou edge quality. #Not very good. #Which is about four times worse then we would 
like. #And CFH is probably as good as anybody right now. It was also about three times worse 
than our milled slits that we were doing for LRIS at the time. So it has caused us to start looking 
at milling machines. #So it is possible that we conveivably might end up with milling as a stan-
dard technology for making the ... #It is possible, and it is possible that we will not have any 
machine that belongs to the observatory but that we will use an outside contracter. #I would think 
that's not a good idea because of time constant for doing that is too long. #It may be a mixture. 
#That means you have to get organized and think about what you are going to do. #There are 
times when you want to respond on short notice. #So anyway, we are just kind of started this. 
#Currently we have a spindle on order a high speed spindle at 26,000 rpm and it is only sixteen 
hundred dollars, so the cost is very low, and it can be adapted to our CNC mill. Also currently we 
have information on a router system from Keckham company in Florida is producing it, and the 
cost of it about thirty to thirty five thousand dollars complete. This is a machine that could be 
installed later on in Hawaii, so it has the coordinates of the slits can be plotted in there and an 
operator could be trained rather easily to go ahead and produce a slitmask on site and I think that 
is the ultimate goal that we have and looking at the data that we have on hand, they are supposed 
some more, but it seems to be a very good system and it has been proven, they are willing to make 
a sample for us and you know, send it us for our inspection, also they will give us names of com-
panies that are using their system currently that we can contact directly and see how they feel, 
because oftent the customers are more reliable in their comments and reports than the manufac-
turer. So we are looking into this here, but at the moment we are going to adopt our CNC with a 
high speed spindle, speeding up the process of cutting the slitmask and hope we have a very nice 
clean cut. We also have ordered the cutters themselves. 15 thousands of a diameter mask. We 
hope that we can cut one mask and throw the cutter away, rather than expecting to get two masks 
out of it. I think it is progressing pretty good. #Our going this route, does it have any effect on our 
slitmask material? #I had a comment about the LDSS II masks. The LDSS II masks, being for a 
much smaller telescope are much smaller and thus need to be cut thinner, but the material that 
they use is many times thicker than the material that is currently being used for the LRIS masks. 
#Many times thicker? #How thick? #Not many times.. I have no idea, I am not an expert on the 
thickness of materials, but I look at it and I say that this is much more rigid and much thicker than 
the LRIS masks. #How wide are the slits? #Well they are about half the width of a typical LRIS, 
they are quite small. #You don't have any numbers. #I wish I did, but we can find out. #We need 
to contact Jeremy or somebody who works on the LDSS II project directly. This was just Jeremy 
handed me this thing last week and said, here look at this.  #I beleive I have his email address if 
you.. #Oh, good. #I would have to ... #Sure, that would be great. #To answer your other question, 
I think it does impact the material, the laser cutters were the ones who declined to cut aluminum 
becauase as you heated it it has funny properties, kind of like welding, have you ever tried to weld 
aluminum. They recommended stainless. Although the CFH uses aluminum. But it means that it 
is pretty standard technology. Milling machine technology has been around longer, and so materi-
als are.. #So I assume that we would be looking at something slightly thicker. #?? having to make 
the material conform to the mask shape.  #???? . #Right. #In this case, is aluminum easier? Pre-
sumably it is, there is less tool wear and so on. #Yeah, we could be using a free machine stainless 
also. Aluminum is probably a better bet. #This sound real nice, I like the cost. #Don't we all? 
#You have to understand this thing is going to be turning around thirty k in rpm, it is not some-
thing that is totally benign in its use. #No machine tool is. #But there is different safety issues. #I 
agree I think that there is almost a different ?? from the laser problems that we would be facing. # 
Are the cutters just straight edged or can you get things beviled. #They are just straight edged. 
They are fifteen thousandths in diameter, there is not much else you can do. #All you have to do is 
get a point to get through the top of the metal then they can be as wide as you want at the bottom. 
#We kind of had this discussion this morning. #How about cutting wavy lines, can it do that? That 
is another good thing over the punch. #And interestingly enough we will get a chance to cut some 
LRIS masks. So I actually have two definitely and probably three quotes to cut demo masks for 
laser cut for Deimos. They all turned out really expensive. #We never sent them out, we just got 
quotes for cutting. #Quotes for cutting. The cheapest was about a thousand per and the most 
expensive was about ten k. #A demo mask. #At this point I am thinking that I am going to slow 
down. #We won't do that, take those demos and send them to the router company. #The router 
company has offered to cut us one I think for free, but we haven't seen this. If they are not I doubt 
if I want to pay for them. The big problem is that the laser companies by and large don't have the 
beds to cut these things and the specifications have obviously.. we send them out with our specs, 
and they have taken a look at them and realized that they are pretty tight especially over a long 
stroke and it is really hard for them to hit that. #This company makes any table you like as a mat-
ter of fact this one is sixty one by a hundred and forty six inches, the one quoted is twenty four by 
thirty six inches x, y. #Do they meet our ten micron spec? #They also meet our positioning spec, 
so. #And there is factors of tens more of these sorts of companies around then the laser cutters. It 
is nice to hear that Gemini might be one of them #Is there sort of a cost per slitmask target assum-
ing that we get down the line we are in operation, # I can tell you, but don't laugh. It's around ten 
dollars. #Ten dollars? That's amortizing all costs, the operator and the machine. #That is the dollar 
value that comes to mind #Is that right? #Are you remembering a different one? #I don't think we 
ever had this. I would have said that you are optimistic by a factor of two. #Yeah probably. 
#Twenty maybe. #If you go with this technology and new tool ametization?? there is some opera-
tions cost for the Keck personnel and so on but it will be..[end of side]...in a high volume mass 
production mode then is what he is saying. #I would like to differ a little bit because each tool is 
fifteen dollars, right now we use one tool per slitmask, #All I was being asked was did I have an 
idea of what we were kind of going for. #It can go easily to fifty dollars. #Fifty, yeah, by the time 
you amortize the machine. #Well maybe, but it is still ok. #That would be on the high side, we 
should struggle to get a little bit lower than that. #The goal is.. it's not thousands, not hundreds, I 
am hoping tens. #It is cheap, compared to the cost of helium at the moment for the telescope, this 
is cheap. #I might offer a little plug for the micromachining approach, I am familiar with micro-
machining in another context where they do things that are infinitely more complicated then what 
we are talking about to higher precision than we are talking about, and they do it routinely and 
cheaply with computer controlled micromachining equipment that you can buy commercially, 
there is a laboratory, for example, in Los Alamos, a micromachining laboratory that does a lot of 
this kind of thing, and they make just the most enormously complicated parts that the characteris-
tic dimension is only a millimeter and within that millimeter they have got all kinds of little 
machining of holes and this and that and the other thing, and they just do that all routinely. #Yeah. 
we have one little problem over that, in that we have this thirty inch thing. #Well they have things 
that make thirty inches seem not so big. They have big things that need little details, they have got 
all kinds of stuff is what I am saying and they do it routinely and effectively, so I am putting in a 
plug for micromachining, I think that is a good idea. #Ok, we are sort of running out of time. 
#Yeah, tent mirror? #Well we have made our little flex bearing here and I test it, it is in a test fix-
ture right now, Erich was able to flex that about a quarter of a degree, without any problem which 
is about twenty times more than actually intend to flex it. And it is quite stiff in the other direction, 
we also tested that. #What does that little guy do, Eric? What's its job? #That is part of the flexure 
compensation system for the tent mirror, and that is a bearing about which we are going to rotate 
the mirror. #A bearing with no moving parts. #Oh. #The problem with putting a rolling bearing is 
that we are only moving just a tiny little bit and after a while you drill little holes where the ball 
sits. #Burn out the bearings. #So this one does it without that and no friction, either. For moving 
small amounts on the order or I think we said forty two arc seconds, ball bearings don't work very 
well. #So the idea is that the shaft there could be bolted down to the table and then the little part 
with the trenians?? on it can rotate a little bit relative to that shaft? #Yes. #I see. # Basically it ends 
on those veins. #Exactly. #The center axle there would be a integrated part of the entire unit, it 
would be one part. #This is one step we haven't quite decided whether we mount such a thing 
internally too the block of zerodur or glue it externally to it. We have to go over that. #And we've 
ordered some pieso electric actuators that we have not received yet which will allow us to do the 
actuation, actually moving the mirrors and when we get all this together we are going to build a 
prototype and mount it on that same model out there. #Tweak it around and rotate the model and 
see how it works, #Probably using an aluminum mirror, though. #Yes, an aluminum mirror 
though. #Great. #And that's where we are with that. #We are coming along real well with that. 
#Just moving along, quarterly report four its somewhere out there in graph land, #If you guys 
have drafts, if I could get them by tomorrow. #Ok, well you need the scientists to add their two 
cents worth, right? #Yes. #So Harlan that means you and me, I think, are you a part of this draft 
yet. #Well, I have got my rough draft, but I haven't handed it is yet, but I will do that tomorrow. 
#My intention again in writing the rough draft was not to lock in any of the wording but to give 
you an idea of my thinking . #I will have something for you on monday, Deanne. #Ok. #Ok. Pro-
spectus, I guess that hasn't changed, I still haven't seen a copy, but..#No, the world is like this, 
there is a frictional decay time of about three days. If you stop bugging people.. so as a result of 
course, since we were all away for a month, nothing happened. #So the prospectus is not done 
yet? #No, its not done yet, I got a request for a figure caption earlier this week, which made me 
wonder, I talked to Sandy and it hadn't been done. #Where is the thing. #It is very close. #No, I 
mean physically, is it at the printer or where? #The message that I got from Lynn said, once this 
caption is in it will go to the copier next week and it should be available within a week or so after 
that. #Which was the situation a month ago. #But now do I understand correctly we are only 
going to make a few copies of this thing. #That's right. #Can I have one? #Yes. We are going to 
make some nice copies, and then we are going to make crummy copies which.. I think it is unclear 
whether they will have color in them or not. We can make tons of black and white copies, for our 
internal purposes. #What is the purpose of not making quite a few color copies, are they expen-
sive or something? #There are two purposes. First of all, they are expensive and second we have 
an agreement with Mr. whatsisname, Russmeyer who is the photographer that we are not sup-
posed to widely disseminate the good versions, because we paid him a hundred bucks per photo-
graph to make ten good copies. #I see. #And if we get very much beyond that they we have to 
start thinking about our ethical situation with regard to him. #Ok, Schedule. We are more or less 
on schedule.. a part of the schedule was circulated with the quarterly report four and I am just 
starting to go through it in detail the other day. #Is the draft, what is with the current draft, is there 
a schedule is there a budget with the current draft? #Just a schedule, and its the outline schedule. 
#Ok. Where is the budget? #Well, that is what we are about to talk about. #Maureen, as you know 
and I have been left in charge of the business office and a few days ago we introduced a new 
accounting system and a new computer system for which we have little answers and it is taking all 
of my time, literally. #And are entire cost code system teams date of July the first with the excep-
tion of Deimos. #Cost codes in the sense of the id's changed, the Deimos id's didn't change. 
#Only minorly. We kind of planned it around that. #But still Marleen has to deal with the change 
of everything else. So we are kind of stalled here. #I am hoping to get that when you do services 
and budget always gets sort of thrown away for a while, so I don't even know what to promise, 
this week has been terrible. #Do we have any alarms or any surprises or anything that we antici-
pate? #I can go in and look and I have and the piesoelectrics were the big cost item in the last 
quarter and there was a bunch of minor stuff. #So as long as we think our design work is basically 
on schedule, and our manpower is tracking along, just about the way we would expect it. #So you 
are not expecting any surprises or anything. #No, not at all, but still Marleen when she gets a 
chance has to go through and do this exercise, do it officially. #Have you made any massive 
changes to the budget of the sort that occasion footnotes and things like that? #No, with the one 
exception that on the tent mirrors, we are pulling down the design time because we realized we 
are not going to need as much design time and funnelling part of that.. #You are putting that 
somewhere else. #yeah, just over the side into materials for the tent mirror, so instead of spending 
design time we are going to buy material. #On schedule I want to mention that I understand that 
the observatory is going to far with the design and production of this new Eschelet spectrograph 
for the Keck II. It is a political matter as to how that is scheduled vis a vis Deimos, it is now a par-
allel project I guess, #Pretty much. #And so I guess.. #It actually starts after and ends sooner. 
#Well it started already as far as the optical part of it and at some point the decision has to be 
made as to whether the optical elements are going to be made in the shop here or whether they are 
going to be farmed out. #Farmed out is.. #Pardon me? # They were all to be farmed out. #Has that 
decision been made, I guess of course, as usual I am not informed about any of this stuff. #Well 
you were in Joe's office or conference room when we last talked about it. #Well I didn't hear any 
decision made, you know talk is cheap, it kind of fills the room until people stop talking. #A state-
ment was made that Deimos was essentially taking the optical shop all of it for eighteen months to 
two years. Well the statement was also made if this eschelet is finished in two years, nobody 
wants it and so on and there is a big hubbub about it has to be done instantly and but there again it 
is all talk, I haven't seen anything written down, I haven't seen any schedule I haven't seen any 
statement to the effect that those optics will be produced out of house or inhouse, so #Joe is the PI 
and.. #Well I understand that.. #So talking to him.. my understanding is all of the optics will be 
made outside. #Nice meeting. #So, if there are collisions they are not going to be in the optical 
shop. You are tracking this, obviously. You are the manager or both projects. #I am. #I am putting 
myself in your hands. #Yell if you see a problem. #The problem is I won't see a problem before it 
runs me over, you are the one who will see it. Ok. #Wait! #Item fifteen. #What? #Can we schedule 
a date for our next meeting, because I know that everybody schedule for the rest of the summer 
are kind of busy. My threw out two dates of august 17 or august 24th, but by looking at calendars 
I got, it doesn't look like the twenty fourth will be good for a lot of people, maybe the seven-
teenth. #Did you send out a request for scheduling? #No, I just got the calendars that everybody 
filled out once a month, I got them today. #So we will tentatively set the seventeenth? #What day 
of the week. #It's a thursday. We have been doing fine on thursday all of this time, why change. 
#You are proposing the seventeenth of august, I will be out of town, so. #I've got Garth out of 
town on the twenty fourth and I don't have a calendar for Sandy. #I am here either day. #So then 
twentyfourth or seventeenth? #I will be gone on the twentyfourth. #Lets do it the seventeenth. Ok. 
... #At some point we will probably pick up the frequency of these because this one was kind of 
longer than I thought, so after this next one I think we will go month to this next, and then we are 
probably going to pick it up and go two weeks, try to keep them shorter. #Sounds good, Here I'll 
give this back to you. #Was there a decision on the meeting? #Seventeenth. #Actually I put 
xeroxes and just put it in your box...[End of Tape.]