Simple "Package Design" Ion Chamber
Monitors for TRIUMF's Proton Beamlines
Daniel Gray and Brian Minato
TRIUMF, 4004 Wesbrook Mall, Vancouver, B.C. Canada, V6T2A3
Abstract. In the beam line designed to supply 100 uA of 500 MeV protons to the two IS AC
production targets at TRIUMF, 13 profile monitor stations were required. The design allows
each station to be fitted with either an air driven wire scanner module for high currents or an
ionization chamber for low currents. Ring shaped multilayer G10 circuit boards were designed
for the latter to enable a simple modular "gas package" that is easily serviced and aligned. These
gas packages have only five basic parts, two outer window frames with 0.010 in. thick E-beam
welded Al windows, two ring shaped circuit boards with 2 mm wire spacing and edge card
connectors (X and Y use the same design of board) and one center frame for mounting to the
inserting mechanism and holding a .001 in. Al foil. The circuit boards are critical components
due to the necessity to hold vacuum along their edges. Signal traces pass from the inner part of
the ring that is gas filled to the outside of the ring that is in vacuum. The windows and center foil
frame are at -300 V bias. This gas package design led to a similar design used to upgrade the
existing (1970's vintage) proton beamline ion chamber monitors.
PROFILE MONITORS FOR THE ISAC BEAMLINE
2A Beamline Standard Profile Monitors
From the TRIUMF 500 MeV cyclotron, the 2 A beamline [1] supplies up to 100 |iA
of protons to the ISAC targets. Nine profile monitors are required for the main 2A
beamline and 2 additional monitors in each leg supplying the 2 target stations.
Two types of profile monitor are used in the 2A beamline. Both use the same drive
mechanism, but they can be assembled as either an ion chamber or a wire scanner
Fig. 1. The drive mechanism utilizes an air cylinder with a 6 in. stroke, end cushions
and air speed controls. Motion is guided by a linear slide fitted with two guide blocks
and a 6 in stroke edge welded stainless steel bellows. This monitor drive mechanism
is a modification of a prototype designed and tested in 1993 [2].
CP648, Beam Instrumentation Workshop 2002: Tenth Workshop, edited by G. A. Smith and T. Russo
© 2002 American Institute of Physics 0-7354-0103-9/02/$19.00
439
Monitor boxes are installed at a 45° angle and have been precision manufactured to
allow changing of monitor drives without realignment. Hand operated toggle clamps
and alignment dowels simplify monitor removal and replacement.
Monitor Station With Ion
Chamber Installed
Monitor Shown With
Wire Scanner Head
FIGURE 1. Standard 2A beamline profile monitor station.
Wire Scanner Head
Three signal blades (0.125 in. wide x 0.002 in. thick Be-Cu) pass through the beam
at a 45° angle. Signal blades are installed so that the beam sees the 0.002 in. edge.
Horizontal and vertical blades provide X and Y information. A third blade is
perpendicular to the scan direction. A tomograph program generates hexagonal
contours of the beam density. Each blade has two 0.005 in. diam Mo, Au plated, bias
wires installed 0.2 in. away from the 0.125 in. wide faces. The bias voltage is +100V
through a 10MQ current protection resistor. The signal blades and bias wires are
spring tensioned. During a scan, sensing of the head position is provided by a 7 in.
stroke, precision wire wound lOkQ linear potentiometer. The scan speed is set to
approximately 0.3 m/s using air cylinder speed control valves. Beam current limits are
set at 10 jiA due to heating of the blades or from beam trips caused by scattering.
440
GAS PACK ION CHAMBERS HEAD
The standard 2A ion chambers gas packs use only 5 basic parts, two outer
windows, two wire boards, and one center mounting frame Fig. 2. The chamber must
be gas tight with one atmosphere of 90% Ar/10%
inside and vacuum at 10~7 Torr
outside.
Rear Window Frame
"X" Wire Board-
Center Frame With Foil-
Gas Plugs
-"Y" Wire Board
-Front Window Frame
FIGURE 2. Exploded view of 2A gas pack ion chamber.
The main component of the chamber is the multi-layer wire boards made from G10
Fig. 3A. The same board layout is used for X and Y; one board is reversed and rotated
90°. The wire board is actually three ring shaped 0.062 in. thick circuit boards bonded
together to form a 3/16 in. sandwich. A cross section of the board is shown in Fig. 3B.
Embedded in this center board are Au plated signal traces that run from 0.015 in. diam
vias (plated through holes) located around the center opening in the board to the
perimeter of the board. The vias are spaced to permit the installation of 32 signal
wires at 2 mm spacing. Vias are used for ease of assembly and the ability to change a
single wire should it become damaged. Vias also retain the wire should the solder
joint soften with heat, although this is not a consideration in the 2A chambers. Signal
wires used are 0.005 in. diam Mo with 5% Au plating. Au plating permits easy
soldering. Traces at the perimeter of the board permit the use of a readily available
edge card connector (part # 345 034 500 202). Edac Inc. manufactures the connectors
from green diallyl phthalate plastic that has good radiation properties [3]. Two outer
circuit boards are bonded to either side of the center board by the manufacturer using
All-108 prepregs [4]. The wire boards must be bonded together with the ability to
441
hold vacuum along their edges. The outer boards have ground planes embedded
within them that reduce bias leakage to the signal traces. They have exposed tabs that
permit connecting the ground plane to an edge card connector trace on the center
board. Exposed faces of the outer boards act as an 'O'-ring sealing surface and must
be free of defects.
Signal wire
^Ground Planes
\ r.
/
U '
———*———
,062
Signal Traces
"D* ring location
Wire soldered into vias
Section A-A
not to scale for clarity
FIGURE 3B. Cross section A-A showing
multiple layers.
FIGURE 3 A. Plan view of wire board with
wires installed.
Two outer window frames are machined from 5086 aluminum and have a 0.010 in.
thick Al window electron beam welded into the center. It is important to match the
alloys correctly for welding. The welding is a critical operation as it must be vacuum
tight and mechanically strong. After calculating the maximum deflection of the
window, 0.060 in. [5], the maximum stress on the window at the center is calculated at
15660 psi [5]. The first time the chamber is pressurized, the window plastically
deforms to the 0.06 in. deflection; this increases the yield strength of the window.
An 'O'-ring groove is machined into the sealing face of the frame. Each window
frame has a screw-type gas purging plug set in epoxy. Care was taken when
machining the gas plug holes in the frame. The holes were machined with a flat
bottom that matches the end of the plug fitting. This minimizes the surface area of the
epoxy exposed to the gas, reducing the out-gassing from the epoxy into the chamber.
Gas plugs used are Cajon Ultra-Torr tube fittings (#SS-2-UT-A-4), with a small
machined plug inserted where the 1/8 in. tube would normally be used.
In the center of the chamber is a frame machined with two "O" ring grooves to
enable vacuum sealing with the circuit boards. The center frame has a tab for
mounting the chamber to the drive mechanism through insulating bushings. The
insulating bushings enables the chamber to be biased at -300 V. In the middle of the
center frame is a 0.001 in. thick 2 in. wide Al foil held in place by an Al wire ring.
This foil, and the two outer windows, act as high voltage bias planes.
442
Ion Chamber "Gas Pack" Assembly and Testing
Using fixtures the windows and wire boards are leak checked before assembly.
After leak checking, the signal wires are tensioned to approximately 5 N and soldered
into the vias. After assembly using good vacuum practices, a final leak check is
performed on the entire chamber. The assembled package is placed in a vacuum
chamber connected to a leak detector. Two tubes are installed into the gas plug
fittings. The tubes run from the gas pack to the outside of the vacuum chamber. The
vacuum chamber is evacuated and a flow of helium probe gas is passed through the
inside of the gas package. A leak rate Q of 10~6 Torr 1/s or better is required. The
volume Vc and the pressure PI inside the chamber is 0.136 liters and 760 Torr
respectively. If the leak rate is considered to be linear, then from equation (1), the
package would take t2 ~ 3.2 years to leak all its gas to the vacuum space [6]. After 6
months the pressure would be approximately 633 Torr. This meets the design
specification that gas loss is less than approximately 20% in 6 months. Since gas
quantity and signal gain are proportional, the monitor would still give profiles but with
a 20% siginal loss. Low gas loss also lessens the reverse stress on the window when
the beamline is vented to atmosphere. When in service, the chambers are routinely
refilled with gas at about a 6 month interval, corresponding to TRIUMF's shutdown
maintenance schedule. The gas used is 10% COz / 90% Ar; this ratio is not critical
therefore a standard welding gas, "Praxair, Mig Mix Gold" [7], is used.
PV
PV
Q=^^>t2=
(l)
Electronics and Operation
The signal electronics is beyond the scope of this article but a simple block diagram
is shown in Fig. 4.
l~~~l fMonitorl
I
FIGURE 4. Block diagram of ion chamber electronics
443
A scanning current integrator collects signals from one 32x32 wire monitor or two
16x16 wire monitors. A CAMAC module controls the integrators gain and enable. A
fiducial generator indicates wire position on an oscilloscope. A beam profile is shown
in Fig. 5. Current limits are set+ at 100 nA to avoid beam trips due to excessive
scattering, typical use is < 1 nA p .
MULTIWIRE CHAMBER SCANS
Beamline
<> i
+ 2A O 4
PACE 1M CHELP)
Monitor group
<> A
O B
+A&B
Charging time: 10 mSec
2AVM3
Peaks:
2
4
16
6
8
10
17
12
FIGURE 5. Beam profile from ion chamber
ADAPTATIONS OF THE GAS PACK DESIGN
Entrance Module Profile Monitors
In addition to the 13 standard profile monitor stations, there is also a profile
monitor in each of the Entrance Modules at the ISAC target stations. The modules
have been designed with the ability to change the profile monitor sensing head to
either an ion chamber or a wire scanner, using a remote handling hot cell. Sensing
heads are similar to the ones used in the 2A beamline, however, they are installed onto
a frame which permits interchangeability on the same monitor. The gas pack is hard
wired to a plug attached to the frame. During the initial start up and low current
commissioning, the first target station was fitted with an ion chamber Fig 6B. The
chamber was subsequently changed to a wire scanner Fig 6A. Currently the entrance
modules to both target stations are fitted with wire scanner heads. Due to design
constraints the sensing heads move in and out of the beam in a vertical direction. The
vertical profile is measured directly but the horizontal profile is derived from a
tomograph program.
444
FIGURE 6A. Entrance module profile
monitor, fitted with wire scanner head.
FIGURE 6B.. Entrance module profile
monitor, fitted with gas pack ion chamber.
Upgrades to existing Monitors
It was recently decided to upgrade the original swing style ion chamber monitors in
other proton beamlines at TRIUMF. These old chambers used circuit boards with
signal traces on the surface. These boards then required additional G10 boards to be
epoxied to each face to provide an "O" ring sealing surface [8]. This assembly led to a
high failure rate, due to gas leakage into the vacuum space. The leaks usually
occurred from gaps between the epoxy and traces. The wire boards were hard wired
to multiple 9 pin feed-throughs (FT) making service difficult. The chambers swing
90° in and out of the beam via an in-house fabricated ferro-fluid FT. After years of
service these FT's became prone to vacuum leakage. They would have to be
periodically "topped up" with ferro-fluidic fluid.
The upgraded monitors have been fitted with a gas package style ion chamber of a
similar design to those used in the 2A beamline. The upgraded chamber is made with
circuit boards of either 16 wires at 3 mm spacing or 16 wires at 5 mm spacing.
Unlike the 2A boards, a unique board is required for the X and Y. However different
wire spacing can be used for X and Y in the same chamber.
The drive mechanism was also upgraded. A new off the shelf ferro-fluid FT was
installed and new radiation tolerant 41 pin signal FT's were also installed. The
original drive motor and electronics were maintained. A prototype of the ion chamber
is shown in Fig. 7A.
1AM8 an ion chamber with remote gas flow
A special "gas pack" ion chamber is used on the 1AM8 monitor Fig. 7B. This
monitor is under several layers of concrete shielding blocks, making regular service
445
difficult. For this reason the monitor was fitted with remote gas flow. On this
installation the gas plugs were replaced with tube fittings. Metal tubes were piped to
allow gas to flow through the chamber continuously. Gas flow is approximately 1
cc/min. Polyimide insulating connectors were spliced into the lines using epoxy. This
allows the gas package to be electrically isolated at -300 V bias.
FIGURE 7A. Prototype swing style ion
chamber
FIGURE 7B. 1AM8 special ion chamber with
remote gas flow
ACKNOWLEDGMENTS
The authors would like to thank the following, G. MacKenzie, for his many years of
knowledge relating to beam instrumentation, A. Hurst, for his encouragement and
support, W. Rawnsley, for the electronics, J. Yandon, for vacuum related topics.
G. Dennison, for the layout of circuit boards, D. Ross and the TRIUMF design office
for help in preparing drawings, R. Roper and TRIUMF's machine shop, for the welded
windows, T. Ries, for stress calculations.
REFERENCES
1. G. M. Stinson, TRIUMF report TRI-DNA-96-05, TRIUMF, 1996. (Internal report)
2. W.R. Rawnsley, "Beam Diagnostics at TRIUMF", Beam Instrumentation Workshop, AIP Conf. Proc. 333 1994
page 125
3. NASA, SP-8053 (June 1970), Nuclear and Space Radiation Effects on Materials, Page 11
4. B. Devonald, BH Devonald and Associates, West Vancouver B.C. Canada, Email communication, April 4 2002
5. Roark's Formulas For Stress & Strain, Warren C. Young, 6th edition, MacGraw Hill, page 457 and page 477.
6. John Yandon, verbal communication.
7. Praxair Inc., Email communication, April 16 2002
8. G. Mackenzie, IEEE Trans. On Nuclear Science, NS-26, 1979, page 2316.
446