Messing with the Innards

General Advice
The Dewar Strap Buckle
Opening the Dewar
Removing the Radiation Shields
Removing the Cold Stage
Preparing the Cold Stage for Shipment
Unpacking and Testing the Cold Stage after Shipment
Wavelength Change: Changing the Hornplate and Backshort
Wavelength Change: Changing the Optical Filters and Lens
Installing the Cold Stage
Installing the Radiation Shields
Closing the Dewar
Revision History

Back to BolocamWebPage
Back to ExpertManual

THIS WHOLE PAGE NEEDS SOME PICTURES.

General Advice

Use the right tool for the job and the right screw for the application. Most everything on the dewar is difficult or impossible to replace, so stripped threads or other damage is to be avoided at all costs.
Be very careful when handling any wiring, especially unpotted connectors. They're easy to break and can cost a lot of time or money to fix or replace.
Use gloves when handling radiation shields or other cold surfaces-- we don't want fingerprints on them.
Be careful of electrostatic discharge (ESD). Whenever the array or JFET gates are exposed, use grounding straps, especially in the low humidity environment at the CSO.
Aluminum tape can be made to stick well by rubbing with a blunt object, like the back end of a ball driver.
Aluminum tape is best peeled up by prying up a corner (use an exacto knife or tweezers) and then pulling with pliers.
Keep drinking water while working. One tends to focus intensely when working on the focal plane, forgetting to drink water. It's even harder to work on the focal plane with a pounding headache and nausea.

The Dewar Strap Buckle

These things are tricky; see the picture. There is a shaft with a slot in it; the free end of the strap is inserted into this slot and, by rotating the shaft so that the strap winds around it, the free end of the strap is held tight. The mating strap has a metal clasp on it that hooks onto a smaller shaft at the end of the bottom leaf of the buckle (the leaf that doesn't ratchet).

To tighten, you pull the upper leaf of the buckle away from the bottom leaf. There's a ratchet on the shaft, so you can close the leaves together and then open again to tighten further, or close the leaves together when done.

To loosen: note the sliding bar on the upper leaf. Pull the bar away from the shaft. This deactivates the ratchet, so then pulling on the straps will unwind the strap from the shaft. Be sure the weight of the dewar is being supported otherwise when you do this! When the buckle has been loosened, you should have enough slack to release the clasp on the mating strap from its shaft on the bottom leaf of the buckle.

How much to tighten? The dewar is held on the holder by the combination of the middle flange sitting on a ledge on the cart and the straps pulling the dewar against the holder. The straps should therefore be tight enough that the dewar doesn't bounce around, but don't overtighten because it will distort the vacuum shell (you will notice the shell compressing if you overtighten).

Opening the Dewar

Presumably you have already vented the dewar. The dewar should be still be sitting on its cart, cryogen ports pointing up, with the e-box already removed. See the ExpertManual for instructions on venting the dewar and removing the e-box.

Find the large plastic storage container for the dewar and radiation shield screws. It is likely in the cabinet in the 3rd floor lab (diametrically opposite the sidecab).
Choose a place to set the dewar down where you will have enough overhead room to raise the vacuum shell above the dewar and get it out of the way. It's just barely doable in the 3rd floor labs or outside on the 3rd floor under the 4th floor overhang.
Remove the black right-angle dewar vacuum port bellows valve. Put a KF blank or a plastic cap on the port to prevent dirt from entering.
Remove any cryogen exhaust lines that may still be connected.
You will see 4 threaded holes in the top of the dewar, on the same circle as the bolts that hold the dewar together. Remove the 4 screws nearest each insert. There should be 8 screws left after you've done this; be sure you leave at least 8 screws, 2 screws every 90 degrees! Use a 3/16" right angle or T-driver to minimize damage to the socket heads (they should be pretty tight!). Put the screws in the dewar screw container in the section labeled vacuum shell.
Screw the dewar support legs into the threaded holes. These legs should be found in the vacuum shell slot in the dewar screw container.
Pull the cotter pin out of the back of the dewar cart so the dewar holder can rotate.
Rotate the dewar holder 180 degrees so the bath ports point straight down. Be careful, go slow and easy: you don't want to unnecessarily stress the focal plane!
Put the cotter pin back in. It will require a bit of tweaking of the rotating joint to get the holes to line up.
Lower the dewar until it rests on the support legs. If coming down on a non-flat surface, it is useful to put a large metal or plywood sheet (1/4" or thicker) down to provide a flat surface for the dewar. Make sure the cart and dewar are neutral; the dewar weight should not be resting on the cart, and the cart weight should not be resting on the dewar (which can happen if you go too low).
Once the weight is safely on the support legs, remove the remaining 8 screws. You'll need a long 3/16" right-angle driver to loosen the screws, then a ratcheting allen driver or a short right-angle driver to get them out. The screws will be tight, don't try to use a ball driver to start with! Again, be sure to put them in the vacuum shell section of the dewar screw container.
Raise the vacuum shell up using the dewar cart. You have to go slow and check that the shell is clearing the superinsulation on the way up -- the cart tends to tilt the shell in an inconvenient way. Try your best to keep the dewar o-ring in its groove; otherwise, you will leave it on the sealing surface and will have to pull it up over the superinsulation, inevitably leaving vacuum grease on the superinsulation. Use gloves if you might touch the superinsulation.
Pull the cart away and lower the shell down. Leave it open-end-down so that dirt doesn't get inside.
Put the dewar o-ring in a plastic bag or such. Do not ball up the o-ring, do not throw it on the ground, do not put it in your pocket -- you want to maintain the o-ring's shape and integrity. Worst case, hang it on a hook in the 3rd floor electronics lab.
When working on the open dewar, be careful not to drop hard or sharp objects on the o-ring sealing surface!

Removing the Radiation Shields

You should have the dewar open at this point, with the vacuum shell out of the way. Use gloves to prevent fingerprints on the shields.

Remove the aluminum tape at the base of the 77K shield. If the tape is attached to Cryolam, pull the tape down at an angle to prevent it from ripping the superinsulation. Remove all the screws, putting them in the 77K shield section of the dewar screw container.
Ease the 77K shield off its mating surface and lift it off the dewar. Set it down somewhere, being careful of the very thin mesh filter at the top.
Disconnect the protective Cryolam retaining the cables on the 4K layer. Only untape the top; the Cryolam can be folded down to get access to everything else.
You will see a 9-pin micro-D coming from the right-most cryocable connecting to another 9-pin micro-D. The whole assembly is taped down to the electrical feedthrough wall. Remove the tape and break the connection. There is a grounding lug to the 2nd 9-pin micro-D; unscrew this and let it hang free, putting the screw and washer somewhere safe. JS 2008/03/03: This 9-pin connector was used to read out the snout diodes. We no longer read these out, so the cable will probably already be disconnected
Remove the aluminum tape at the base of the 4K radiation shield. Remove the tape covering the inner set of screws connecting the 4K shield to the electrical feedthrough wall, as well as any tape at the bottom of the electrical feedthrough wall that is connected to the 4K radiation shield. Do not remove the tape on the outer set of screws; you will not need access to these screws, and you'll just have to redo the tape job if you remove it now.
Remove the screws attaching the 4K shield to the cold plate (socket heads) and to the electrical feedthrough wall (phillips head). JS 2008/03/03: These phillips head screws have been changed to socket head. Put these in the 4K shield section of the dewar screw container. Remove the 4K shield and put it somewhere safe.

Removing the Cold Stage

You should have the dewar open and all the radiation shields off at this point. Be careful about grounding yourself when necessary!

Clear off bench space on an ESD mat for the cold stage.
Disconnect the auxiliary thermometry: 37-pin connector to IC 37-pin connector, connections to IC diode and heater (and GRT if necessary). Let the cables hang down to the side of the dewar, though tape the 37-pin connector to the dewar so its weight doesn't stress the wires. Untape any wires that are taped to the cold stage and would prevent its removal.
Disconnect the fridge thermometry: disconnect the 50-pin micro-D at the RF filter connector and the JFET box input connector. JS 2008/03/03: This cable may not need to be disconnected. This best done using a flathead screwdriver to pry the connectors up; do no pull on the wires! Once disconnected, you will probably need to untape the cable bundle partway to allow removal of the cold stage. Remove as little tape as necessary. Let the fridge cable sit on the 4K plate or hang down off the dewar.
Unscrew the UC fridge, IC fridge, and IC cooldown heat switch heat straps. The fridge straps should be disconnected at the fridge end. The heat switch strap should be disconnected at the cold stage end. Be careful with the screws -- don't drop them into an inaccessible place! The screws should go into the heat straps section of the cold stage screw container.
Remove all the vespel post foot screws, they go into the wedding cake section of the cold stage screw container.
Put on a ground strap and ground yourself to the dewar.
Break the Cable A/JFET box connections. Use pliers or channel locks to grab the epoxy potting, then rock the connectors side to side and slowly ease them up. Be careful to avoid damaging the vespel posts while doing this.
Lift the cold stage carefully off the JFET box, tilting as you lift it to allow the IC heat strap to clear the UC head. Be careful, you can break fridge wiring if you lift the stage up too quickly.
Set the cold stage down on the ESD mat.
Reattach your ground strap to the dewar (presumably you had to remove it to carry the cold stage away). Attach grounding straps to the cold plate and to the JFET box input connectors. These grounding straps should have a 51-pin micro-D connector on one end, with all pins soldered together to a single wire, and a lug on the other end. The lug should be first attached to the cold plate using a 4-40 screw, and then the connector attached to the JFET box input connector.
If the dewar is going to be closed up for storage, tape down any loose cabling on the cold plate to the JFET box so that it does not get damaged when the dewar is flipped over.

Preparing the Cold Stage for Shipment

The cold stage should be sitting on a ESD mat. You presumably have found the special shipping containers, including the lucite box.

Put on your grounding strap.
Remove the UC heat strap, clean the grease off of it and the cold stage and store it in the dewar screw container. Clean the grease off the IC heat strap also, but leave it in place.
Remove 8 of the 12 screws holding the mesh filter holder to the hornplate. The screws may be difficult to take out due to the bow in the filter holder. Put the screws in the screws for filter holders container. Leave 4 screws at 90 degree intervals to keep the filter holder attached to the hornplate (you don't want to get dust down the horns!).
Once face of the lucite box has a bolt circle. Detach this face and set it on top of the array. The counterbores should be on the side away from the cold stage (the outside face of the box). Select screws of the appropriate length for screwing the lucite face to the hornplate from the 2-56 for hornplate section of the cold stage screws container. Note that the bow in the filter holders makes it hard to screw these guys in; be very careful not to strip the threads. Use as much thread in the hornplate as you can. It may be necessary to balance the torquing of the screws because, depending on what filter holder is being used, the lucite face may rest on the remaining 4 screws rather than flush against the filter holder.
Once you are happy that the screws are secure and can take the weight of the cold stage, flip the cold stage over and set it down on the lucite face. Tie the Cable A connectors to the vespel post feet: cut lengths of teflon plumbing tape, wind them up so they form ropes, thread them through the holes in the feet and the connectors, and tie them such that the connectors won't flop around during shipping.
Drop the cold stage into the lucite box. Screw the lucite face attached to the cold stage to the rest of the box. Now you can remove your grounding strap.
Set the lucite box in the first layer cardboard box. Set it in upside down so that the cold stage is resting upside down, not hanging from the lucite box. Put packing material in as necessary and close up the first box. Put in packing material to hold the first box immobile inside the second box, and close up.
Mark the box Bolocam focal plane if not already marked. Make arrangements for delivery of the focal plane to Diana Bisel for shipment to Caltech. Priority overnight (by 10 am) should be used to minimize the amount of time that the focal plane is in transit. Diana should be advised the shipment is totally non-hazardous.

Unpacking and Testing the Cold Stage after Shipment

After the cold stage is shipped, we do electrical checks and compare to previous ones to ensure there has been no damage during shipment. As of October, 2003, not a single working bolometer has been lost during shipment (after 4 round trips!).

You presumably have opened the cold stage's cardboard shipping box and pulled the lucite container out.
Put the lucite container on a ESD mat and ground yourself. Set the lucite container up so the cold stage is hanging from the top face.
Unscrew the top face from the rest of the box, lift it out, and set it down such that the cold stage is upside down, resting on the lucite face.
The Cable A connectors should still be tied to the vespel post feet with teflon plumbing tape. Leave them as-is, it will make it easier to do electrical checks.
Find a copy of the Array Checkout Sheet. There is probably already a PDF file sitting in the Bolocam user folder (log in as visitor) on pika, the PC in the main computer room. If not, you can download a copy from the Bolocam internal web site, from the ProceduresAndCheckoutSheets page.
Use the 51-pin micro-D at JFET box input column and measure the resistances of the indicated pin pairs. Note that the pattern is very simple: pin pairs are 1/2, 3/4, etc. Pins 17/18 are the bias lines, so they see all the 10M-Rbolo-10M resistor chains in parallel and should have resistances of 833 kohms to 1200 kohms, depending on the number of good bolometers in each hextant. Pin 35 is ground (it's not really grounded to anything). Compare the measured resistances to the last set of measurements to look for damage.
Cut the teflon plumbing tape holding the connectors to the vespel post feet. Check for damaged posts.
Flip the cold stage over so it rests on the vespel posts.
Unscrew the lucite face from the hornplate. Be careful not to pull the filter off during this process -- there should be 4 screws holding the filter to the hornplate, but sometimes they are forgotten. The screws may be difficult to take out due to the bow in the filter holder. Put the screws in the 2-56 for hornplate section of the cold stage screw container and replace the missing screws with screws from the screws for filter holders container. Pick the length appopriate to the current filter thickness. Be careful when screwing them in, the hornplate holes are not always cleanly threaded. Note that the bow in the filter holders makes it hard to screw these guys in; be very careful not to strip the threads.
Reattach the lucite face to the lucite box (you don't need to put on all the screws, throw the spares in the box), put it back in the cardboard shipping boxes, and store in a safe, dry place.
Use aluminum and/or aluminized mylar tape to seal up the gaps at the edge of the filter holder and between the filter holder and hornplate. This tape usually has to be removed to get access to the screw holes, so it usually needs to be redone after every unpacking in preparation for installation. You can save this task for later when the array is mounted on the cold plate, but don't forget to tape up around the 300 mK filter!

Wavelength Change: Changing the Hornplate and Backshort

When we change wavelengths, it is necessary to change the hornplate, backshort, some or all of the metal mesh filters, and the polyethylene lens. Changing the hornplate and mesh filters on the hornplate will always be handled by a specialist using a cleanroom in the UH office in Hilo. These are instructions for doing this work; but, you should never do this work unless you have been specifically trained by an expert.

Step 0: Preparatory Arrangements

To make sure you can do everything you need to do in a timely manner, you should make the following arrangements ahead of time:

If possible, have the day crew warm the dewar, open it, and strip it down to the 4K shield. Ed Bufil is well-acquainted with doing this.
Arrange with Ed Sousa at the IfA office in Hilo to make use of the clean room at IfA. Just say you are with the CSO and need to do some work on an instrument focal plane. You will need the room for 2-3 hours. In recent history, this clean room has not been in use by IfA and so this should be easy to arrange. Remember that you will need to do this on a weekday, and be aware of any holidays that may interfere.
Ensure that you have a full day to reinstall the focal plane and close up the dewar once you have finished with the work in Hilo. You don't want to leave the dewar open for an extended amount of time for both safety and to avoid a long pumpout if a lot of water vapor condenses in the dewar.
Remember that CSO rules preclude being at the summit alone, so again be aware of weekends and any holidays.

Step 1: At the summit

At the summit, you first need to remove the wafer-mount/backshort/hornplate assembly from the cold stage. You presumably have completed the Removing the Cold Stage step, so you have the wedding cake on an ESD mat.

In the Bolocam cabinet in the 3rd floor storage room, find the box labeled Bolocam Wavelength Change. This contains all the parts you will need, including a box of spare screws.
Ground yourself to the ESD mat!
Flip the wedding cake over so it sits on the hornplate.
Remove the 12 2-56 screws connecting the Cable AA's to the connectors on the bottom of the hornplate. Use a screwdriver to apply pressure to pry the connectors up -- avoid pulling on the cables themselves.
Flip the wedding cake back over so it sits on its legs.
You will see a number of 4-40 screws going through the hornplate. Some of these attach the hornplate to the wedding cake, some keep the hornplate/wafer-mount assembly together. Remove the long screws that attach the hornplate to the wedding cake; do not remove the screws that keep the assembly together. JS 2008/03/03: Some screws are marked with a red "X", these are probably the long screws.
Remove the hornplate/wafer-mount/backshort assembly and clean off any grease on the bottom side with isopropanol.
Pack up the assembly in bubble wrap and a ziploc bag, you should find one in the Wavelength Change box. It doesn't need to be ESD-safe. You can just put this assembly in the Wavelength Change box for transport to Hilo.
Check that you have in the Wavelength Change box the hornplate and backshort you will be changing to in Hilo. The hornplates have the 300 mK filters permanently attached to them, so there should be no issue about filters. Any filters in the Wavelength Change box are for 4K and 77K.
Collect the following items for your work in Hilo. You can just bring these in a cardboard box.
- most recent Bolocam logbook
- permanent black marker
- latex or other disposable gloves
- kimwipes
- isopropanol wash bottle
- canned air or some other thing to dust parts off with
- fine tweezers
- coarse tweezers
- exacto knife or razor blades
- scissors
- set of english ball drivers and allen wrenches
- nice handheld digital voltmeter (the kind that can read up to 1 GOhm resistance) and good leads (you need to probe at the micro-D connectors)
- aluminum tape (1" preferred)
- kapton tape
- aluminized mylar tape
- electrical checkout sheets
- three long 2-56 jack screws. See instruction 3 in "Step 2: In Hilo" directly below (JS 2008/03/03).

Step 2: In Hilo

You will change the hornplate and backshort in the IfA clean room in Hilo to avoid unnecessary accumulation of dust on the array. When you arrive, go to the front desk and let them know you contacted Ed Sousa to use the clean room. The person at the desk can lead you there. Do the following:

Setup

Ignore the signs about shoes and booties unless they are actively using the clean room for something else.
Enter the clean room through the nonfunctioning air-shower, turn on the room lights.
To turn on power to the filtered room, turn on the large power panel on the wall, then hit the leftmost On/Off button on the white power control panel just to the left of the door to the filtered room.
Turn on the light switch for the HEPA-filtered workbench above the bench.
Make sure the HEPA filter is running -- if you put a kimwipe on the workbench, you should see it wave in the air flow.
Clean off the workbench with isopropanol and kimwipes.
Get your parts out of their bags and put them on the bench. A key point about HEPA-filtered work stations is that they are cleaner than almost anything you bring in. The ziploc bags and bubble wrap you brought your parts in, kimwipes, etc. are worse in terms of dust and lint than just putting your parts on the table in the clean air flow. Minimize the number of unnecessary things you put on the table, and don't put anything paper like upstream of your work area.
Clean off the hornplate and backshort you will be installing. Be careful not to get isopropanol on the filters. Blow any kimwipe lint off of them. If they look clean when you take them out of the bags, don't bother cleaning them -- the kimwipes do more harm than good sometimes.

Remove the aluminum tape on the bottom of the assembly (covering up light leaks near the connectors) and around the bottom edge. Do not remove the aluminum tape around the joint between the filters and the hornplate. Clean any tape residue off. If you erase any markings, be sure to remark them after you have cleaned.
Replacing the backshort

Flip the assembly over. Remove the 6 2-56 screws that attach the hexagonal backshort piece to the wafer mount.
Use 3 long 2-56 screws to jack the backshort off the wafer mount. You will see threaded 2-56 holes just to the outside of the 2-56 clearance holes that held the screws you just removed. Insert the long 2-56's into these threaded holes. As you turn the screws, they will lift the backshort off the wafer mount. Of course, incrementally ease them out in rotation -- don't screw one in all the way, then the next, etc. Once the backshort is free (it's about 0.2 inches deep into the wafer mount), just pull the backshort off. Be careful at this point -- nothing is preventing the backshort from moving transversely and the long 2-56 screws from hitting the array.
THE ARRAY IS EXPOSED. BE CAREFUL.
Remove the jacking screws from the backshort you just removed and insert them into the backshort you will be installing. Just put the old backshort to the side -- it's dangerous to be trying to put things in bags with the array exposed.
Screw the long 2-56 screws in far enough that they extend from the other side of the backshort, but not so far that they extend past the backshort itself.
Put the backshort in place on the wafer mount, being careful to respect the hextant labeling.
Unscrew the jackscrews, again incrementally, allowing the backshort to slip into the wafer mount. If it sticks or catches, jiggle it carefully to free it. Sudden motion is your enemy.
Once the backshort is all the way in, unscrew the jacking screws completely and insert and screw in the shorter 2-56's that attach the backshort to the wafer mount.

Replacing the hornplate

Flip the assembly so it sits on the connectors.
You should see 6 4-40 screws remaining in the outermost bolt circle (don't confuse this with the screws holding the filter holder together). Remove these.
Simply pull the hornplate off. If you feel any resistance, stop. The hornplate sometimes sticks to the wafer mount because of some kapton tape spacers in place. You can pry the two apart, carefully, by inserting a razor or exacto blade between the two at the side. You will see 6 tabs sticking from the hornplate into the wafer mount at this interface. There is an intentional 0.005 inch gap between the hornplate and the wafer mount at these tabs. You can insert a blade at these tabs and pry the two apart. Be gentle.
THE ARRAY IS EXPOSED. BE CAREFUL.
Set the new hornplate onto the wafer mount, again being careful to respect the hextant labeling. Also be sure to settle the tabs on the hornplate into their slots in the wafer mount carefully -- no sudden motions!
Insert the 6 4-40 screws you removed earlier to attach the hornplate to the wafer mount. Remember, only 6 of the holes are tapped on the wafer mount, the other 12 are clear, so be sure the screws engage in threads.

Do the electrical checkout using the checkout sheet.

You should find a mating solder-cup micro-D connector in the
screw box in the Wavelength Change box. Using it makes it easier to do the checks. Pin 1 on that connector is marked. The connector has 51 pins, 1-18 in the first row, 19-35 in the second row, and 36-51 in the bottom row.
Use the "51-pin micro-D at array" pin pairing column.
Note the date and that you are in Hilo doing a wavelength change. Compare as you go the previous checkout in the logbook to make sure you don't see any new problem detectors.
Save the completed checkout sheet in the logbook when you are done.

Retape the assembly -- tape over the gap between the hornplate and wafer mount around the outer edge, and also tape over the gaps around the connectors. Remember that the outer edge of the bottom of the wafer mount needs to sit flush against the wedding cake, so you need to make sure no tape is in that region. If you brought a wafer mount spacer, you can use it as a trimming template.
Pack up everything in the same manner as it came down to Hilo.
Clean up your mess, esp. any trash, and turn off all the power when you leave the clean room.

Step 3: At the summit

You will now remount the assembly on the wedding cake and check the electrical connections.

Find the right wafer mount spacer! There are spacer rings that go between the wafer mount and the wedding cake to put the horns at the right height. The spacers for 1.1 mm and 2.1 mm are different. WHAT ARE THE THICKNESSES? JS 2008/03/03: These spacer rings are labeled, but I "think" the 1 mm ring is 0.3 inches and the 2 mm ring is 0.225 inches. Make sure you have the right one!
Check the holes on the wedding cake that will accept the screws attaching the wafer mount to the wedding cake. Some have been cross-threaded over the years. Your goal is to avoid cross threading any more of them.
Clean the bottom of the wafer mount where it will mate to the wedding cake, and also clean the space ring.
Apply Apiezon N grease to the bottom of the wafer mount and to the spacer ring.
Put the spacer ring down first, then the wafer mount. The spacer ring is 18-fold symmetric, but make sure you have the wafer mount properly aligned as far as the hextant labels are concerned.
Carefully screw in the 6 long 4-40 screws that attach the wafer mount to the wedding cake. Be careful not to cross-thread any more holes.
Flip the wedding cake over.
Plug in the Cable AA's. You can use two ball drivers to push on the ends of the connectors. Once they are seated, use the same 12 2-56 screws you removed early to screw them in. Screw in the two screws for each connector alternately so that the connector is not canted one way or the other by the screw pressure.
Do the electrical checkout at the JFET end of Cable A (the only exposed connectors now).

When doing this check, it is convenient to tie the connectors to the feet of the vespel legs with dental floss or teflon tape so they don't move around while you are trying to probe them.
Remember that, because this connector is female, the pin order is mirrored from the checks you did at the array above.
Use the "51-pin microD at JFET box input" pairing column.
In addition to the entries on the sheet, you can test between pins 17 and 18 to get the full parallel resistance of each hextant through the bias resistor modules. A perfect hextant should give you 833 kohms. You can check past history for what these numbers should be. Doing this check first is a good idea because it immediately tells you whether anything major is wrong; but it is important to do the individual bolometer checks also, as the 17/18 check does not tell you whether the lines connecting from the bolometers to the JFET gates are connected.
Again, check against results in the logbook from the previous time this was done, and save your checkout sheet in the logbook

Now you will proceed with Installing the Cold Stage.

Wavelength Change: Changing the Optical Filters and Lens

When we change wavelengths, it is necessary to change the hornplate, backshort, some or all of the metal mesh filters, and the polyethylene lens. Changing the hornplate and mesh filters on the hornplate will always be handled by a specialist using a cleanroom in the UH office in Hilo. The other filters and the lens are very easy to change on site. Each filter now has its own dedicated holder, so changing filters is just a matter of unscrewing one filter holder and replacing it with a different one.

The filters that need to be changed are:

At the top of the 4K snout lid. There are actually two filters here. Only the wider, top one needs to be changed.
On the inside of the 77K snout lid.

A filter holder consists of a two aluminum rings with a filter sandwiched between. The holder has two bolt circles. The inner bolt circle uses 2-56 screws and holds the filter holder together, sandwiching the filter. The screws are inserted from the bottom side of the holder. The filter is GE-varnished to the two rings. The outer bolt circle uses 4-40 screws and mounts the filter holder to the surface where it will sit. In the current design, the edges of the filter are open; a spacer between the two rings would prevent full compression of the filter, which ensures good optical sealing and thermal contact. Because of this gap, aluminum tape must be applied around the filter after mounting to the dewar to seal up the gap.

So, the procedure for changing filters is:

Remove the tape sealing up the edge of the filter holder.
Unscrew the filter holder from the lid to which it is mounted. JS 2008/03/03: The screws are on the bottom side of the holder.
Screw the new filter holder in place. JS 2008/03/03: The table below describes which filters go where. Note that B286 is swapped from 77K to 4K when switching from 2.1 mm to 1.1 mm, and vice versa.
Tape up the edge of the filter holder, allowing the tape to run both over the top of the filter holder (but not the filter!) and onto the surface that the filter holder sits on.

Some special things to deal with:

The 77K filter shares a mounting bolt circle with the radiation-blocker mesh on the other side of the 77K snout lid. The 77K filter uses 6 of the holes and the radiation blocker uses 6 of the holes. When replacing the 77K mesh filter, you probably have to remove the radiation baffle ring. You can use this ring as a stand to set the lid on so that the lid doesn't rest directly on the thin radiation blocker. When taping up the edges of the filter after mounting it, be sure to trim the tape so it doesn't interfere with the radiation baffle ring.

Storage: Filters not in use should be put into a plastic bag and stored such that no sharp objects can damage the mesh. Putting the plastic bag in bubble wrap or a jiffy-pack mailer is usually good.

Which filters go where (always listed in order out from the array) as of 2004/12/09:

location	1.1 mm	2.1 mm
300 mK	10 icm (B411) (PTFE ARC) 12 icm (B161) (PTFE ARC) 14 icm (B430) (PTFE ARC)	5 icm (B309) (no ARC) 7.75 icm (C195) (no ARC) 8.5 icm (B159) (no ARC)
4K	14 icm (B377) (PTFE ARC) 14 icm (B286) (no ARC)	14 icm (B377) (PTFE ARC) 8.5 icm (B131) (gore-tex ARC)
77K	18 icm (B445) (PTFE ARC)	14 icm (B286) (no ARC)

We always use the same thin radiation blocker (10 microns) at 77K. For a listing of all available filters, see the Bolocam internal OpticsAndEfficiency page.

The lens can be changed during the installation of the radiation shields -- you have to pull the lens off anyways to do a touch check between the focal plane and the lens mount, so at this point you should change to the new lens. The simplest way to identify which lens is which is by the bags they are stored in. If you have managed to lose this ordering, then you can attempt to identify the lenses by their grooving:

dimension	1.1 mm	2.1 mm
groove depth	0.0087 inches	0.017 inches
groove width (= groove pitch/2)	0.00435 inches	0.055 inches

This is of course not easy to do, so the best thing is to not lose track of which lens is which!

Installing the Cold Stage

You should have already done the electrical checks on the cold stage following shipment and unpacking or following a wavelength change. When handling the cold stage or working on the dewar, ground yourself to the dewar whenever possible until the cold stage wiring has been connected to the JFET input connectors. You can use a standard ESD grounding strap and plug the banana plug end into the screw holes on the outer edge of the cold plate.

Be careful to keep track of all screws! Lost screws can result in touches when the dewar is flipped over and force a warmup. Be especially carefully when working around the fridge, as screws can get lost in very inaccessible places!

Get all thermometry cabling out of the way; the thermometry cabling is usually taped to the top of the JFET box when the dewar is closed up for storage.
Check for any holes in the JFET box that need to be taped up.
Clean the IC and UC heat strap attachment points with alcohol.
Wipe down the JFET box with a kimwipe and some alcohol to get rid of any crud on the JFET box that may impede the mounting of the cold stage or end up on the cold stage when the dewar is flipped over.
Prepare the cold stage. Make sure there is nothing on the bottom of the vespel post feet. Make sure all the Cable AA's are tight and that all the Cable A's are properly tied down. Clean and regrease the IC heat strap. If the UC heat strap is already attached to the cold stage, clean and regrease its bottom surface. Use Apiezon conductive grease in both locations.
Connect your ground strap to the dewar. Disconnect the JFET input connector ground straps and remove.
Lift the cold stage off the bench, carry it over to the dewar, and slowly ease it down into place. Watch out for the micro-D connectors; they tend to hang up on the JFET box, and you do not want to set down the stage in such a way that there is a lot of stress or a tight bend in the cable coming out of the epoxy potting. When lowering the cold stage down, you will have to tilt it slightly to allow the IC heat strap to get around the UC head. Be careful to prevent the IC heat strap from catching on any of the fridge wiring, as it could break said wires.
Reattach your grounding strap to the dewar (you presumably had to remove it to pick the cold stage up off its bench).
Plug all the Cable A connectors into the JFET box input connectors. A large set of pliers or channel locks can be used to grab the connectors by the epoxy potting, which is the safest place to apply force. Be careful to avoid damaging the vespel posts with the pliers or channel locks. We do not screw down the Cable A connectors because of errors made during machining that offset the holes. Once all the Cable A connectors are attached, you can remove your grounding straps.
Screw in the cold stage feet. These are just short 4-40 socket heads, you should find them in the wedding cake section of the cold stage screw container. Get all the screws going in their threads first, then tighten down when all are in place. The holes in the feet and the JFET box don't line up perfectly, so you have to be careful of cross-threading and perhaps apply a little transverse force to the vespel posts to get the holes to line up. It sometimes helps to hold the screw from the side by its shaft with a pair of tweezers to align the screw with the hole.
Tape over any of the Cable A connector through holes that are not already taped over. This is necessary to block radiation from the JFET box from coming out through the misaligned screw holes. A small piece of aluminum or aluminized mylar tape usually does the trick.
Check for touches between the cold stage and the JFET box or anything else it should not touch. Do this now, before you make the remainder of the connections. Especially important to check are the Cable AA's, which come very close to touching the JFET box by design.
Screw in the IC and UC heat straps. The IC heat strap is most easily done with fillister head screws and a screw-starter; there should be one in the Bolocam equipment cabinet in the 3rd floor storage room. The screws should be found in the heat straps section of the cold stage screw container. You may need to use some pliers to hold the strap in the correct orientation while you put the first screw in. Torque these screws down with a good screwdriver, but don't apply too much force or you might damage the fridge or break the head off the screw. If the UC strap is not already connected to the cold stage, clean it off and grease it with Apiezon conductive grease. Use 2-56 round head brass screws and washers to attach it (also founds in the heat straps section of the cold stage screw container). Use short 4-40 pan head brass screws (from the same heat straps section of the cold stage screw container) to attach the heat strap to the UC fridge head, again after cleaning and greasing the surfaces that will mate. You will need 2 washers for each screw because the heat strap holes are oversized. Again, be careful about applying too much force, especially on the UC fridge head.
Attach the auxiliary side thermometry. There is a 51-pin micro-D connected to the inside-shield side of the left-most RF filter box. It has a number of wires coming off of it. These mate as follows. Be careful not to break any connections; some of the wires are not (and cannot) be potted.

a set of wires going to a 37-pin micro-D. This connector mates to the 37-pin micro-D that is screwed to the left side of the IC stage.
a set of wires going to IC stage thermometry. One pair of wires terminates in a white microdot connector, which should be connected to the IC diode (on the back part of the IC stage, also with a white microdot connector). Another pair of wires terminates in a black samtec with 2 pins; this should be connected to the IC heater (black samtec 2-pin connector, near the IC diode). A final set of 4 wires terminates in a 4-pin samtec, which for now connects to nothing (it would be connected to a IC GRT if one existed). This 4-pin samtec can just be taped to the JFET box where there is free space.
a set of 3 pairs of wires going to 4K plate thermometry. One pair (or maybe two pairs?) go to the 4K diode, which sits on the 4K plate very close to the RF filter box connector. The remainder are currently unused and can just be taped up.
Once all the thermometry is all attached, tape down the wires as necessary to prevent them from flopping around, creating a touch, or being damaged by the 4K shield when it is installed.

Attach the IC cooldown heat switch, which should be in the left back corner of the cold plate. There should be a strap, probably attached to the top of the heat switch, but possibly attached to the IC cold stage. Connect the free end to the appropriate place. To connect to the heat switch, it's necessary to use a nut. Make the connection tight but be careful not to damage the heat switch. To connect to the cold stage, just screw down to any easily accessible screw hole on the IC cold stage, being careful of the placement to ensure the strap does not accidentally thermally short to something (like the fridge pump shield or the 4K radiation shield).
The fridge side wiring is probably disconnected from its RF filter box connector and the JFET box. Make these connections, being very careful not to damage the pins as you plug in the connectors. One way to do it is to use two 3/32 ball drivers to apply pressure to the connector mounting through holes. Tape the wiring down to the JFET box; it needs to be taped to the same surface that the cold stage feet rest on to avoid being crunched by the 4K radiation shield (especially at the corner of the JFET box neares the fridge).
Inspect!

Are all the Cable A's connected to JFET box input connectors and their holes taped over?
Are the UC fridge, IC fridge, and IC cooldown heat straps connected (properly)?
Are all the fridge and thermometry cables connected properly and taped down as necessary?
Are there any holes on the JFET stage that have not been sealed up?
Is there anything that might impede the attachment of the 4K radiation shield or be crushed by it?

Electrical tests: You can test many of the electrical connection by doing measurements at the room temperature hermetic feedthroughs. Get a Cable G (one that mates and demates easily) and use it to do the following checkouts. You have to be VERY careful when mating/demating the Cable G to avoid damaging the pins; there is barely enough clearance between the connectors and the floor to do this.

Fridge thermometry: Find the fridge thermometry wiring table in the Bolocam black binder. You can also download a copy from the internal web site on the ElectronicsAndWiring page. JS 2008/03/03: You can also find this in the log book. Use the Cable G 2:50-s socket column for your checks. Diodes should show a voltage drop of about 0.65 to 0.75 volts when measured using the diode check scale (have to get the polarity right, of course). Diode polarity is unimportant; the warm electronics are set up to correct for the polarities. GRTs should show wiring resistance only (150 ohms or so). Pump heaters should be 200-500 ohms. Heat switch heaters should be about 10 kohms.
Auxiliary thermometry: Colocated with the fridge thermometry wiring table (or in the same electronic document) will be the auxiliary thermometry wiring table. The auxiliary hermetic feedthrough connector is the one maximally distant from the fridge hermetic feedthrough connector. Do the same checks as for the the fridge, noting that most lines are connected to nothing and that the fridge-base diode and IC GRT do not exist.
Bolometer bias/JFET bias: The 6 hermetic feedthroughs between the fridge and auxiliary connectors are the hextant feedthroughs. Hextant 1(A) is furthest to the left and they go in the sequence 1(A), 2(F), 3(E), 4(D), 5(C), 6(B) from left to right. The bolometer bias and JFET bias can be checked for each one. On the Cable G 9-pin D-sub connector, pins 7/8 are bolometer bias and pins 3/4 are JFET bias. Measure the resistances across these pairs; bolometer pairs should have resistances of 833 kohms to 1200 kohms. They vary depending on the number of good bolometers in each hextant. These should be the same resistances as measured during the cold stage checkout following shipment and unpacking. The JFET bias resistance should be 500-700 ohms in all cases, with some variation between hextants.

Installing the Radiation Shields

You should already have the array fully installed and cabled up as well as all thermometry. See elsewhere on this page for instructions for these tasks. Use gloves for all handling of the radiation shields and snouts to avoid fingerprints.

Make sure you have done all the checks at the end of the cold stage installation section! Have 2 people do visual checks independently to be sure.
Check for clearance problems to the radiation shield; especially the IC cooldown heat switch and its strap. The fridge will be tight but is known to clear (unless you've moved it!).
Make sure all wires are taped down so they won't be damaged by the 4K shield; note that it has a flange that sticks some distance into the cold plate surface.
Put the 4K shield in place. When dropping it down, try to minimize the scraping of of the inside of the shield against the cold stage and JFET box -- this will rip charcoal off and deposit it unknown places. If you here a lot scraping, take the shield off, blow away the charcoal, and try again.
Put in all the cold plate and electronics feedthrough face screws finger tight (the screws should be found in the dewar screw container in the 4K shield section). The shield does not fit perfectly, so it will require a bit of tweaking. Don't tighten any screws until they are all in. Screwing in the electronics feedthrough face screws will help a bit because it pulls the other holes into better alignment. Once all the screws are in, tighten them all down.
It is in general a good idea to check the clearance of the lens mount around the array. It unfortunately requires unscrewing and rescrewing a lot of screws. JS 2008/03/03: For the following instructions, note that the snout thermometry diodes are no longer being used.

Remove the lid at the top of the 4K snout. This will require removing the tape sealing up the crack in the mating between the lid and the snout and unscrewing the diode thermometer sitting on the 4K snout. The thermometer can probably stay attached to its wiring, but tape it down to prevent stress on the wires. When removing the tape, be careful of the wire going through a screw hole in the snout.
You should see a wire feeding through a screw hole in the 4K snout to a thermometer on the internal snout, with a connectorized joint close to the thermometery. Make sure there is a mark indicating the correct polarity of the connection, then break this connection.
There should be a thermometry wire running up the side of the 4K snout to a connector, which then breaks off into two pairs of wires, one for the thermometer on the 4K snout lid and one for the thermometer on the internal snout lid. Make sure the polarity of the first connection (before the split into two pairs) is marked, then break the connection and untape the wire running up the side of the 4K snout so you can remove the snout.
Unscrew the screws at the base of the 4K snout. You should be able to remove the 4K snout, being careful of the thermometry, especially wires that should have been disconnected or untaped and were not.
You should now see the lens mount/internal snout assembly. There are a set of copper wires running from the internal snout screws to the lens mount screws; these provide a better thermal link to the bath than can be provided by the screws through the lens to the lens mount. Unscrew the screws at the base of the lens mount. Pull the lens mount/internal snout assembly off. Check that there are no touches between the hornplate and the radiation shield. Use a flashlight to look for touches. In particular, look at an oblique angle so you can see below the radiation shield lid. By design, there should be no touches. Check that the hornplate is reasonably centered in the hole in the radiation shield.nbsp; JS 2008/03/03: If you are making a wavelength change then now is a good time to change the filters. See the instructions given here.
Reattach the lens mount/internal snout assembly, being sure to tighten down the copper straps going to the internal snout. JS 2008/03/03: If you are making a wavelength change, now is a good time to change the lens. See the instructions given here. If any of the copper straps are showing too much wear, replace them -- they are just large gauge copper wire.
Unscrew the screws holding the internal snout and lens to the lens mount. Remove the lens and internal snout (put them on kimwipes to avoid their accumulating any dirt). You should be able to see the top of the mesh filters on the horn plate. Use a flashlight to check for any clearance problems between the hornplate and the lens mount. In general there should be no problems, but, if the top of the mesh filters has changed height, you might have a touch to the Bock black. It may be necessary to remove the lens mount and trim down the Bock black on the lathe. Be sure to clean up any chips or kimwipe bits. You can degrease the part with soapy water, then blow it dry with dry nitrogen gas. Don't use kimwipes; you will leave lots of little bits on the sharp edges of the Bock black.

Reattach the lens and internal snout. The lens has no polarity, but you will notice 3 large slotted holes in its flange that line up with the 3 large holes in the lens mount flange. Be sure to reattach all the copper straps and to replace any damaged ones.
Reattach the 4K snout. Reconnect/retape the two thermometers that you disconnected and the line to the coming up from the 4K radiation shield and connecting to the line splitter, being careful to use the polarity markings. You may need to tape down the thermometer line that goes inside the snout to the internal snout; there is too much slack in that line.
Replace the 4K snout lid, again being careful with the thermometry. Before screwing down the lid, slide it around a bit to make sure there is no interference with the internal snout.
Retape the snout lid/snout interface to prevent light leaks. Tape down any excess length in the lines. Use aluminized mylar tape if taping to superinsulation; it causes less damage to the superinsulation when removed.

Tape up the cracks in the 4K shield. This will require 1" aluminum tape. Do the three following mating points: between the radiation shield and the cold plate, the overlap point between the radiation shield and the electronics feedthrough wall, and the divot at the bottom of the electronics feedthrough wall where it meets the radiation shield and the cold plate (this point is hard to do and requires a significant amount of tape to really cover it). Be careful not to damage the cryocables. The edge of the aluminum tape is sharp! Finally, check around the shield -- you shouldn't need to do anything else, but you may need to do some touch up work on the top flange.
Connect the 4K snout thermometry wiring to the spare 9-pin micro-D and test. JS 2008/03/03: Note that the snout thermometry is no longer being used.

The right-most 9-pin micro-D should just be hanging unconnected. There is a 9-pin micro-D taped to the outside of the 4K shield, as well as a grounding strap from that 9-pin micro-D. Connect the two 9-pin micro-D's together and screw the grounding lug down to one of the screws in the electronics feedthrough wall (any screw will do, we usually use one of the many 2-56 socket heads). Tape the assembly down so it isn't hanging free.
Electrical configuration: the 9-pin micro-D has 2 usable lines. Chassis ground provides a "third" line. The two diode thermometers on the 4K snout and the internal snout are read out using these lines. For historical reasons, both have one end connected to chassis ground. The free end of one thermometer is connected to one of the 9-pin micro-D lines and the free end of the other thermometer is connected to the remaining other 9-pin micro-D lines. This allows one to read out the 2 thermometers using only 3 lines. (One could get away with only 2 lines by wiring the diodes in anti-parallel, but then one has to be very careful about the polarities.)
To test: attach a Cable G to the hermetic feedthrough connector associated with the lines being used for the snout thermometry. This is the right-most one in the current orientation with the dewar open. Use a Cable G that mates and demates easily. You have to be VERY careful when doing this to avoid damage to the pins; there is barely enough clearance between the connectors and the floor to do this. The lines being used come out on pins 7 and 8 of the 9-pin D-sub connector on the Cable G. Connect a DMM to the two sockets of a Lakeshore diode thermometery supply. In general, to test, one line of the diode supply will be connected directly to the dewar chassis (use a banana cable plugged into an available screw hole) and the other line to pin 7 or 8 of the D-sub. You'll have to test each one of the 4 possible configurations to see how the diodes are wired up and which one is which. You can identify the thermometers by heating the 4K snout diode with your fingers or a heat gun and watch its voltage drop change. Obviously, the internal snout diode voltage should not change when you do this. It should be impossible to have miswired the diodes in such a way that they can't be read out at all, but this is your chance to double-check that.

There is a hanging piece of Cryolam (aluminized mylar with a weave on one side) that is used to retain all the cryocables. Raise this up and tape it to the top of the 4K shield. All the cables should now be protected from the 77K radiation shield.
Do a last check to make sure everything is in order -- no cracks in the shield, all thermometry wiring ok, all cabling connected and retained so it can't be damaged.
Attach the 77K radiation shield. Note that it has a specific orientation; there is a mark on the shield and on the flange it mates to. The shield never goes on particularly easily -- it tends to rest on the plate that it is supposed to mate to. You have to stick your fingers in to pull the shield out so it slips down over the mating surface. Be careful of the cryocables! They run out under copper protectors so they should be ok, but it's not a good idea to rest the shield on the protectors. You may have to rotate the shield a little bit to get the screw holes to line up (note that they are severely oversized). Put in 4 screws from the 77K shield section of the dewar screw container, with washers, finger-tight.
Remove the 77K snout lid, again removing the tape over the mating interface first. JS 2008/03/03: If you are changing wavelengths then now is a good time to change the 77K filters. See the instructions given here.
Tweak the 77K shield so it is centered on the 4K radiation shield lid. This is a "by-eye" adjustment, you can't be very precise, we just want to prevent an egregious misalignment. Use a flashlight to check that the 77K lid flange does not touch any of the 4K snout lid screws or the thermometer. Once the shield is aligned, tighten down the 4 screws at the base of the shield. Insert and tighten the remaining screws with washers at the shield base.
Put the 77K snout lid back on, making sure there are no clearance problems to 4K. The likely places where you might have a problem are the 77K mesh filter holder and the radiation baffle sticking down. Slide the lid around to make sure you feel no friction that might indicate a touch. Tape over the crack where the snout lid meets the shield.
Tape over the screw holes and gaps at the bottom of the 77K shield. 2" aluminum tape makes this easiest. When taping, you do not want the tape to make contact to the cables coming out under the copper cable protectors and going under the superinsulation. Tape to the copper cable protectors only. You will have a small gap under the cable protectors where light can leak in; it has been demonstrated that this is ok. Also, try to avoid taping to the superinsulation blanket; you should be able to tape directly to the metal of the shield or to a single layer of Cryolam that lies under the blanket. If you are running low on 2" tape, contact the Bolocam support person so he can order more; this is not something the CSO normally stocks.

Closing the Dewar

You should already have all the radiation shields on and taped up and have done electrical checks to make sure the bolometers and JFETs are connected. See elsewhere on this page for details on these tasks. Use gloves if you don't like getting vacuum grease on your hands, and also if you might tough the superinsulation while guiding the shell down onto the dewar.

Remove the optical window at the top of the dewar. You need to see down the snout in order to center the shell as it is dropped onto the dewar. Put the window somewhere safe, and store it in a way that the sealing surface and window are safe from damage.
Clean off the o-ring sealing surface at the base of the dewar with alcohol and kimwipes. Make sure there are no bits of stuff on the surface that may result in a vacuum leak. The surface already has many scratches, these don't seem to cause vacuum leaks (if you use enough vacuum grease!).
Clean the large dewar o-ring with alcohol and regrease it. Be liberal with the grease -- there's no harm. Insert the o-ring in the upside-down o-ring groove on the dewar shell. If you start with your fingers at one point and run them round separately until they meet on the other side, you will get the o-ring to stick to the upside-down shell. The idea is to push all the air between the o-ring and the bottom of the groove out so that the o-ring seals against the groove; this keeps it from falling off. You should in general not clean the groove itself -- it's likely that you will leave kimwipe bits in the groove and thereby do more damage than good.
Move the vacuum shell over the dewar, orienting it so the azimuthal origin arrow is 180 degrees away from thee cryogen exhaust ports. Lower the shell down slowly. The tricky part is when the bottom of the shell reaches the bottom layer of 77K superinsulation and when the bottom of the snout reaches the top of the 77K snout -- there is not much clearance in either case, and the tilt of the vacuum shell tends to make things harder. If you can't get the shell to go down, you can try loosening the straps holding the vacuum shell so that you can tilt the shell (see above for how to deal with the straps). Be sure to keep looking through the top of the snout to make sure you are not doing damage to the superinsulation. Watch out for the thin radiation-blocker metal mesh filter on the 77K snout also; it will rip easily if the shell hits it. You can use your fingers to check the clearance between the shell and the superinsulation at the bottom; you will need to adjust the position of the shell as it drops to prevent it from damaging the superinsulation. Use gloves so that you don't get fingerprints on the superinsulation. Finally, be careful about the shell ripping little bits of superinsulation off and leaving them on the sealing surface -- these will cause leaks. JS 2008/03/03: This entire process may be easier if you level both the dewar and the shell before lowering the shell.
Once the shell is completely down, release the straps and pull the dewar cart away -- the cart makes it difficult to orient the shell properly. Line the shell up with the dewar by eye (the azimuthal origin arrow should be 180 degrees away from the cryogen exhaust ports), then use a screw to line up the holes in the dewar and in the shell. The screws should be taken from the dewar screw container section labeled vacuum shell. Put in, finger-tight, 8 screws, 2 every 90 degrees, in the 2 holes midway between the nearby support legs (trying to screw a bolt into the holes near the support legs is difficult). Using long screws will make it easier to find the holes in the shell (except note that one of the shell holes requires a short screw due to a damaged thread). Be careful to not cross-thread the holes! Once all 8 screws are finger tight, tighten them down hard in a star pattern to compress the o-ring. As when opening, it's easier to use a 3/16" ratchet or short right-angle driver for getting the screws most of the way in, and then a long right-angle driver for tightening.
Restrap the dewar to the cart. Be careful to attach the dewar to the holder with the correct orientation; see the picture. Note that the bath ports are pointing up, not down, in the picture! Make sure the straps are tight, but not overly tight; see above for details.
Raise the dewar up so that the top of the dewar holder is 4-5 feet off the ground. You need to have enough room for the snout to clear the floor when the dewar is rotated. Pull the cotter pin from the back of the dewar holder and rotate the dewar 180 degrees. Be careful, go slow and easy: you don't want to unnecessarily stress the focal plane.
Put the cotter pin back in. It will require a bit of tweaking of the rotating joint to get the holes to line up.
Unscrew the support legs and put them in the vacuum shell section of the dewar screw container.
Put the remaining screws into the flange, again using screws from the vacuum shell section of the dewar screw container. If you are having trouble, try using some lubricant on them (spray graphite lubricant is good because it doesn't make a mess). Be careful to not cross-thread the holes! Tighten them down hard with a 3/16" right-angle or T-driver. It's good to do a star pattern here also.

Revision History

2004/01/11 SG
First version
2004/01/14 SG
Forgot to include lens change in optics changing instructions.
2004/01/29 SG
Add links to get back to main pages.
2004/12/09 SG
Update filter listing information.
2007/04/05 SG
Add instructions for change of hornplate and backshort during wavelength changes.

2008/03/03 JS
Made some minor changes that are not marked, and added several new instructions/comments that are all marked with "JS 2008/03/03". All of the changes I made were recommended by John Vaillancourt.

Questions or comments? Contact the Bolocam support person.