From: Robert Henderson <rhend@triumf.ca>
Date: Tue, 8 Feb 2000 17:05:29 -0700
To: e614chambers@relay.phys.ualberta.ca
Subject: E614 Joblist for module production
To: E614 collaboration DRAFT ONLY
From: Robert Henderson ==========
Date: 07 Feb 2000
Re: Joblist for E614 spectrometer modules
Hello Guys,
Below you will find a draft of the `joblist'. This is only that
associated with the making of the modules and there are some things I've
missed (like the straws). Please read ASAP we will be discussing this at
tomorrows meeting.
Regards Robert Henderson
------------------------------------------------------------------
The E164 spectrometer modules consists of:
8 (UV)A (short) modules
6 (UV)B (long) modules
2 4(UV) modules
2 2(UmVm) modules
1 Target=2(UmVm) module
--
19 Modules
Excluding the machined G10 gas-box pieces, these 19 modules contain:
44 drift wire-planes
12 MWPC wire-planes
18 cathode-only planes
19 gas-boxes in the 5 types
840 ribbon assemblies (10 position) - kflex, PCB0 and Mill-max
112 ribbon assemblies (7 position) - kflex, PCB0 and Mill-max
148 cop-tape assemblies (7 position) - copper-tape, PCB0 and Mill-max
56 Service bd #1
56 machined delrin HV input blocks (1 type)
38 Service bd #2
38 machined delrin pulser/LV blocks (4 types)
304 machined aluminium preamp guides (6 types)
448 machined preamp mount blocks (1 types)
74 cathode foil assemblies
112 Finished preamps 16-channel
112 Finished preamps 24-channel
24 Finished 24-channel daughter preamps (for ADCs)
76 Machined G10 dowels (5 lengths)
152 Aluminium cital `plugs'
----------------------------------------------------------------------------
In addition, I believe we should make the following extra modules:
1 (UV)A (short) modules
1 (UV)B (long) modules
1 4(UV) modules
1 2(UmVm) modules
1 Target=2(UmVm) module
--
5 Modules
Excluding the machined G10 gas-box pieces, these 5 extra modules contain:
12 drift wire-planes
8 MWPC wire-planes
4 cathode-only planes
5 gas-boxes in the 5 types
300 ribbon assemblies (10 position) - kflex, PCB0 and Mill-max
40 ribbon assemblies (7 position) - kflex, PCB0 and Mill-max
48 cop-tape assemblies (7 position) - copper-tape, PCB0 and Mill-max
20 Service bd #1
20 machined delrin HV input blocks (1 type)
10 Service bd #2
10 machined delrin pulser/LV blocks (4 types)
80 machined aluminium preamp guides (6 types)
160 machined preamp mount blocks (1 types)
24 cathode foil assemblies
40 Finished preamps 16-channel
40 Finished preamps 24-channel
16 Finished 24-channel daughter preamps (for ADCs)
20 Machined G10 dowels (5 lengths)
40 Aluminium cital `plugs'
----------------------------------------------------------------------------
Adding these and a little a few spares we get something like:
Item
----
56 drift wire-planes + spares => 58
20 MWPC wire-planes + spares => 22
22 cathode-only planes + spares => 24
24 gas-boxes in the 5 types + spares => 24 #11
1140 ribbon assemblies (10 position) + spares => 1250 #1
152 ribbon assemblies (7 position) + spares => 168 #2
196 cop-tape assemblies (7 position) + spares => 208 #3
76 Service bd #1 + spares => 80 #5
76 machined delrin HV input blocks (1 type) + spares => 80 #4
48 Service bd #2 + spares => 52 #8
48 machined delrin pulser/LV blocks (4 types) + spares => 52 #6
384 machined aluminium preamp guides (6 types) + spares => 408 #7
608 machined preamp mount blocks (1 types) + spares => 680 #9
98 cathode foil assemblies + spares => 120
152 Finished preamps 16-channel + spares => 170 #10
152 Finished preamps 24-channel + spares => 170 #10
40 Finished 24-channel daughter preamps + spares => 46 #10
96 Machined G10 dowels (5 lengths) + spares => 110
192 Aluminium cital `plugs' + spares => 220
The cables for the spectrometer are:
112 16-channel alpha cables (2 types) + spares => 124 + test => 132 #12
136 24-channel alpha cables (2 types) + spares => 150 + test => 166 #13
56 SHV cables + spares => 60 + test => 68 #14
38 Preamp power cables (+4V) + spares => 42 + test => 44 #15
19 Pulser cables + spares => 21 + test => 23 #16
Now I will give a brief description of some of these items, the work
required and whether it can/should be done off the TRIUMF site.
----------------------------------------------------------------------------
Item #1 Ribbon assemblies (10 position) QUANTITY: 1250
-------
This sub-assembly consists of three components:
(a) kflex - small kapton ribbon (21x20 mm) - 1 off -
(b) PCB0 - small 1/16" PCB (4.7x23.4 mm) - 1 off -
(c) Mill-max female sockets - 10 off -
The 10 mill-max sockets are pushed throught holes in the PCB. The kflex is
fitted ontop of the 3/16" of the mill-max protrusion past the PCB. Making
sure everything is laying `flat', the Kfex is soldered to the 10 mill-max
sockets.
After soldering the assembly is cleaned in an ultrasonic bath. Then the
assembly is held, and glue added at a second line of holes in the kflex. This
gluing is to strain relieve the kflex to the PCB. Continuity checks and no
contacts between adjacents traces.
It is essential that the ribbon assembly is done with good care and QC. The
ribbons are NOT rugged. If we get bad contacts in the ribbons, wires will
float and wire-planes will trip. Since they alternate signal and ground,
shorts between adjacent traces will lose that channels signals.
This is a job that should be done off-site.
----------------------------------------------------------------------------
Item #2 Ribbon assemblies (7 position) QUANTITY: 168
-------
This sub-assembly consists of three components:
(a) kflex - small kapton ribbon (21x14 mm) - 1 off -
(cut from wider 10 channel klex)
(b) PCB0 - small 1/16" PCB (4.7x17.4 mm) - 1 off -
(c) Mill-max female sockets - 7 off -
Same exact process as 10 postion ribbon assembly (Item #1)
This is a job that should be done off-site, obviously by whoever is doing
item #1.
----------------------------------------------------------------------------
Item #3 copper-tape assemblies (7 position) QUANTITY: 208
-------
This sub-assembly consists of three components:
(a) Smale piece of thin copper sheet (16x14 mm) - 1 off -
(b) PCB0 - small 1/16" PCB (4.7x17.4 mm) - 1 off -
(c) Mill-max female sockets - 7 off -
The 7 mill-max sockets are pushed throught holes in the PCB and soldered.
The piece of thin copper is aligned and soldered to the PCB. Because it uses
a piece of thin copper sheet, this assmebly is more rugged than the Ribbon
assemblies (items #1 and #2).
This job should be done off-site, probably by the same group that does
items #1 and 2.
----------------------------------------------------------------------------
Item #4 Machined delrin HV input blocks QUANTITY: 80
-------
These delrin blocks are 16x8.2x6 mm. They have two #4-40 tapped holes thru
the 6 mm thickness and one #19 (0.166") hole thru the 16x6 face.
This block hold a banana plug socket for SHV input for a wire-plane.
This job should be done off-site.
----------------------------------------------------------------------------
Item #5 Service board #1 QUANTITY: 80
-------
These PCBs will be laminated into the gas-box walls. Each service boad
services one wire-plane and it's cathode foil, they have an extra connection
for the cathode-foil-only detector layer. The ribbon cable assemblies (items
#1, #2 and #3 soldered to the detector-layer lamels) plug onto male mill-max
pins of this service board making all signal and ground connections.
These PCBs need to be stuffed:
(a) There are 4 output sockets that need to be soldered onto this service
board. Two of these are Samtec MMS-133-01-L-SH which has 33 outputs, 66
solder spots. The other two connectors need 49 outputs, since Samtec only
make it up to 40 contacts, two MMS-133-01-L-SH will be used, one cut to
16 outputs, the other left at 33 outputs. Each of the two resultant 49
output connectors have 96 solder spots.
(b) There are a total of 185 Mill-max male pins to be soldered onto each
service board. Before installation, these male pins need to be bent at
30 degrees. I will specify this bend and if neccessary design the bending
jig. These bent pins are soldered onto the service board in groups of 10
and 7. Since they are loose individual pins, an alignment jig will be
required to position them as they are solderd. This step needs to be done
properly or the ribbon cables will not be able to connect as required.
(c) There are two 1 Mohm chip resistors to be soldered on the board. Also a
a large HV resistor. Two of the 4-40 clearance holes need to be
countersunk to mount the delrin HV block (item #4).
----------------------------------------------------------------------------
Item #6 Machined delrin pulser/LV blocks (4 types) QUANTITY: 52
-------
These delrin blocks hold banana plug sockets for module services,
i.e. preamp power (+4V), pulser input and ground connections.
Type a (UV) module - 46x17.43x8.2 mm (with 16x11.43 mm `notch' out of
Quantity: 33 the 46x17.43 face). Two blind #4-40 tapped holes and
five #19 (0.166") holes thru the 46x17.43 face.
Type b 2(UmVm) module - 30x25.43x8.2 mm. Two blind #4-40 tapped holes and
Quantity: 7 six #19 (0.166") holes thru the 30x25.43 face.
Type c 4(UV) module - 30x41.43x8.2 mm. Two blind #4-40 tapped holes and
Quantity: 7 ten #19 (0.166") holes thru the 30x41.43 face.
Type d Target module - 30x41.43x8.2 mm. Two blind #4-40 tapped holes and
Quantity: 5 six #19 (0.166") holes thru the 30x41.43 face.
This job should be done off-site.
----------------------------------------------------------------------------
Item #7 Machined aluminium preamp guides (6 types) QUANTITY: 408
-------
These aluminium blocks are attached to service board #2. The preamps are
solidly screwed to them, providing a good earthing. The blocks have slots to
guide the preamps during installation. These slots are about 0.085" wide,
36.6 mm long and 2.5 mm deep. They have four blind tapped #4-40 holes, two
for attached to service board #2 and two for attachment of cover PCBs. Each
preamp has two guide blocks which are mirrors. The `left' guide also has
another two #4-40 blind holes to hold the preamp-power-PCB. The four module
types will be accomodated by blocks of three different heights, making six
types in all. All blocks are 36.5 mm long and 10.5 mm wide.
Type 1 (UV) module - 21.5 mm height. One slot per block.
Quantities: 132 left, 132 right.
Type 2 2(UmVm) module - 37.5 mm height. Three slots per block.
and target module Quantities: 44 left, 44 right.
Type 3 4(UV) module - 45.5 mm height. Four slots per block.
Quantities: 28 left, 28 right.
This job should be done off-site.
----------------------------------------------------------------------------
Item #8 Service board #2 QUANTITY: 52
-------
Two of these PCBs are bolted to each detector module, one for the U-plane,
the other a `flipped' board for the V-plane. The following componets need to
be mounted on each board: 8 preamp guide blocks, 1 delrin pulser/LV block,
2 chip resistors, 2 filter caps and 1 cover PCB.
This is a relatively small job, so we'll probably do it on-site.
----------------------------------------------------------------------------
Item #9 Machined preamp mount blocks QUANTITY: 680
-------
Two of these blocks are screwed to each preamp board. They provide solid
metal `ears' to connect the preamp to the pair of preamp guide blocks. These
aluminium blocks are 20x10x6 mm. They have two tapped #2-56 holes through the
20x10 mm face and one #2-56 clearance hole through the 20x6 mm face.
This job should be done off-site.
----------------------------------------------------------------------------
Item #10 Finished 16-channel preamps QUANTITY: 170
-------- Finished 24-channel preamps QUANTITY: 170
Finished 24-channel daughter preamps QUANTITY: 46
Finishing the many preamp boards is a large job. It is done in two parts
which have quite different requirments.
Part A
------
Since the senses wire are at high voltage, the main preamps (not the
daughter boards) have HV capacitors at the input, either 16 or 24. QC is
ESSENTIAL for this job. A single tripping capacitor among the 80 of a
wire-plane means that the wire-plane is useless until the faulty preamp is
located and replaced. If the trip is hard enough, it may even result in
sparking inside the chamber and possibly even breaking of the wire.
The steps are:
- A shallow groove about 2 mm wide is machined on each side of the board and
deburred.
- The caps are carefully soldered into position.
- The board is cleaned in a small ultrasonic bath.
- A G10 dam is glued on one side with 5 minute expoxy.
- Epoxy is mixed and degassed under vacuum.
- The dam is partly filled with epoxy to well cover the caps.
- The preamp is put in a vacuum oven to cure.
- The other side is damed and glued the same way.
- The set epoxy is machined down to a specified thickness.
After the capacitors are installed, the board must be tested with HV. If
there are tics or trips, the offending capacitors must be `cut out' and
replaced, or the board rejected. The `finished' in the description means that
we want 170 16-channel preamps and 170 24-channel preamps that have PASSED
this HV test. QC is important, especially in handling the caps themselves.
Since the preamps will be in helium, the HV test may also have to be in
helium, since it has a far weaker breakdown potential.
This job could be off-site, but ONLY if QC is demonstrated.
Part B
------
All 386 preamps need to be `stuffed' and pulse tested. The components are:
- The input connector (2 mm pitch Samtec male)
- Input 1 Mohm chip resistors
- Diode or FET packages
- VTX
- Output chip capacitors
- Output connector (0.1" pitch double row)
- +4V input pin
- +4V filter components
- Output pulldown chip resistors (probably not required)
On the VTX board there are several 0.1 uf caps that need to be removed (or
relocated), since they cause significant cross-talk.
This job should be done off-site.
----------------------------------------------------------------------------
Item #11 Gas-boxes (5 types) QUANTITY: 24
--------
The gas-boxes are made for a variety of G10 parts, gas fittings and
finished #1 service boards. Two finished #2 service boards are bolted on,
complete with preamp guide blocks and cover PCB. The lid plate and finished
gas-box `bottom' have citals glued into them.
Lid Plate and Base plate
------------------------
Machined from 0.25" thick copper-laminated (single side) G10. Four plates
are rouch cut from each 4'x8' sheet. They are first tinned to protect the
copper suface, then NC machined and deburred.
A mylar sheet is glued to the inside (non-metal) surface of each plate.
This sheet covers the large central hole and then the central part of the
mylar is cut off, leaving a mylar `lip' extending inwards from the cutout in
the G10. The gluing step should be done on a granite surface.
G10 Arc pieces
--------------
The circumference gas-box wall is made in five arcs; two 86 degree readout
arcs, one 8 degree arc between the two readout arcs, and two 90 degree arcs
to complete total 360 degrees. The G10 arc pieces have Ri=330 mm and
Ro=340 mm, i.e. are 10 mm wide in the radial direction.
The three non-readout arc pieces are made from `full-thickness' G10. This
15 mm for the (UV) module, 23 mm for the 2(UmVm) module, 31 mm for the target
module, and 39 mm for the 4(UV) module. Radial slots and holes are made for
the brass gas fitting. There are tapped #4-40 holes thru the thickness. The
upper surface has an O-ring groove. These arc pieces are screwed and
laminated (glued) to the 0.25" base plate.
Each of the two readout arcs of the gas wall are made up of several layers
which are screwed and laminated (glued) to the 0.25" base plate. The layers
are:
A 0.125" thick G10 arc, #4-40 clearance holes.
Finished 1/16" thick #1 service boards, quantity 1, 2 or 4.
Spacer-pieces, about 6.4 mm thick, #4-40 clearance holes, quantity 0, 1 or 3.
Top piece with O-ring groove and tapped #4-40 holes.
The laminating (gluing) should be done on a granite surface to ensure
finished gas-box bottom it as flat as possible. After gluing, the O-ring
groove is hand finished to ensure a good seal around the entire
circumference. The straw tube sub-assemblies are glued to the inside holes
of the gas inlets.
Citals
------
The lid plate and finished gas-box `bottom' each have four citals glued
into them. They protrude 1.5 mm beyond the inside surface of the G10 plates
and the gluing must be done on the thick glass assembly table to ensure the
four citals sufaces are at the correct common surface.
This last gluing in of the citals is the only step that MUST be done at
TRIUMF. The rest could be done off-site if QC could be demonstrated. I
suspect all this job will be dobe at TRIUMF unless we are swamped.
----------------------------------------------------------------------------
Item #12 16-channel alpha cables (2 types) QUANTITY: 132
--------
At the spectrometor end the 16 mini-coax cables are soldered to two 1/32"
PCBs (about 15x30 mm size). These two PCBs are then then soldered to a Samtec
22 pin female connector (double row, 0.1" pitch). The cable bundles leave
along the direction of the connector row, i.e. it's at right-angles to the
preamp insertion direction.
The two cable-group types are for U and V planes. They have identical
ordering, the difference is which direction the cable bundle will leave the
preamps.
At the electronic rack end the 16 cables are split into two groups of
eight. Each group of eight are soldered to a PCB (size about 15x20 mm), this
PCB is soldered to a Samtec 16-pin female connector (double row, 0.1" pitch).
The finished assembly needs to be continuity and pulse checked, QC is
important.
This job should be done off-site.
----------------------------------------------------------------------------
Item #13 24-channel alpha cables (2 types) QUANTITY: 166
--------
At the spectrometor end the 24 mini-coax cables are soldered to two 1/32"
PCBs (about 15x40 mm size). These two PCBs are then then soldered to a Samtec
30 pin female connector (double row, 0.1" pitch). The cable bundles leave
along the direction of the connector row, i.e. it's at right-angles to the
preamp insertion direction.
The two cable-group types are for U and V planes. They have identical
ordering, the difference is which direction the cable bundle will leave the
preamps.
At the electronic rack end the 16 cables are split into three groups of
eight. Each group of eight are soldered to a PCB (size about 15x20 mm), this
PCB is soldered to a Samtec 16-pin female connector (double row, 0.1" pitch).
The finished assembly needs to be continuity and pulse checked, QC is
important.
This job should be done off-site, probably by the same person that does
item #11.
----------------------------------------------------------------------------
Item #14 High Voltage cables QUANTITY: 68
--------
These High Voltage cables have regular SHV connectors at the electronic
rack end. At the spectrometer end each coax cable has two bannana plugs, for
HV and for ground. Small diameter cables are preferable to minimize area in
the cable trays. Non magnetic cable is required. Each cable needs to be HV
tested to say 3 kV, checking there are no trips or tics. The last 2 meters of
these cables and the bannana plugs will be in helium, so the HV test should
take that into account.
This job should be done off-site.
----------------------------------------------------------------------------
Item #15 Preamp power cables (+4V) QUANTITY: 44
--------
Two of these coax cables bring +4 volt power to each module, one for the
U-planes and one for the V-planes. These cables have regular BNC connectors
at the electronic rack end. At the spectrometer end each coax cable has two
bannana plugs, for power and for ground. Small diameter cables are
preferable to minimize area in the cable trays. Non magnetic cable is
required.
For the (UV) module, the power is lowest, each cable powers two 24-channel
and two 16-channel preamps. For the 4(UV) module, the power is highest, each
cable powers eight 24-channel and eight 16-channel preamps.
This job should be done off-site.
----------------------------------------------------------------------------
Item #16 Pulser cables QUANTITY: 23
--------
One of these coax cables bring a pulser pulse to each module. These cables
have regular BNC connectors at the electronic rack end. At the spectrometer
end each coax cable has two bannana plugs, for signal and for ground. Small
diameter cables are preferable to minimize area in the cable trays. Non
magnetic cable is required.
This job should be done off-site.
============================================================================
The other items on the list are all clearly on-site. Such jobs as:
Machining all G10 gas-box and cathode-foil components
-----------------------------------------------------
Making cathode foil sub-assemblies QUANTITY: 120
----------------------------------
- Stretching foil over the large granite.
- Gluing and `pressing' to first 1/16" G10 ring
- Installing carrier frames, cutting loose, flipping over
- Gluing and `pressing' to second 1/16" G10 ring
- Cutting of excess foil, hand finished, filling any voids.
- Inspeaction and QC
Making finished wire-planes QUANTITIES: 58 Drift, 22 MWPC
---------------------------
- Sand-blasting one side of glass plate
- Laminating Lamel onto glass plate
- Gluing 1/8" G10 part into central cutout
- Cluing in four 4 mm thick citals
- Installation of cathode-foil sub-assembly, check height
- Removal of cathode-foil sub-assembly
- Winding wire-plane
- Suveying finished wire-plane for tension, position and defects
- Replacing reject wires and re-surveying
- Solderng on Ribbon-cable and copper-tape sub-assemblies
- Cleaning and inspection
Making finished cathode-only planes QUANTITY: 24
-----------------------------------
- Sand-blasting one side of glass plate
- Laminating Lamel onto glass plate
- Gluing 1/8" G10 part into central cutout
- Cluing in four 4 mm thick citals
- Solderng on copper-tape sub-assemblies
- Cleaning and inspection
- Installation of cathode-foil sub-assembly, check height
- Removal of cathode-foil sub-assembly
Assembly of modules (in clean room) QUANTITY: 24
-----------------------------------
- Cleaning and inspection of gas-box
- Cleaning and inspection of wire-planes and foil sub-assemblies
- Installation of cathode-foil sub-assemblies
- Gluing mylar `seal' on top surface of cathode-only plane
- Gluing mylar `seal' on bottom surface of first wire-plane assembly
- Install detector layers in gas-box and close
- Glue temporary external-window-assemblies to outside faces of module
- Move to granite table and tighten screws
Bench Testing of Modules QUANTITY: 24
------------------------
- Flush with test gas, make sure module is gas tight
- Test pulser response and check noise levels
- Apply HV to each plane, check currents, look at sample outputs,
reverse bias if required, apply over voltage and irradiate with strong
source. Open and clean etc if problems found.
- When module looks `good', attach to scanner
- If scanner test passed, remove temporary external-window-assemblies
- Seal mylar gas box extension to first/last detector layers
- Test gas seals okay, test with DME
- Store finished module safely
E614 Joblist for module production / Robert Henderson
- Created for the The Center for Subatomic Research E614 Project Projects Page.
- Created by The CoCoBoard.