Production
Quality control and testing
carry HV, power, sensing and control levels from PP2 to detector
modules
each module served individually => 1 tape per module
barrel
according to SCT Bias Voltage Power Specification V2.03
(draft)
according to ATLAS SCT Low Voltage Power Specification
V2.0 (draft)
DMILL process limit for ABCD chip: V_max = 5.5 V (on chip)
analog (ABCD bipolar FE):
Control
SELECT for clock (opto redundancy scheme)
Minimize space
severe space constraints all way through ATLAS detector
identified hot spots (so far)
IPC standards for flexible printed
circuits taken as design guidelines
caveat: no IPC standards exist for aluminium
flexes
0.5 mm track and gap grid chosen to match 1 mm pitch, (25 mil = 0.635
mm felt too risky)
low current - minimal 0.5 mm tracks
high current - wide tracks
group track in two layers with matching supplies and returns for RF filtering and minimal loops
obey IPC-2223 specification for HV separation (500 V -> 2.5 mm)
take 50 um Al as conductor for inner part up to PP1 (4x smaller
X0 than Cu for same conductivity)
take 70 um Cu as conductor for
outer part (> 2x better conductivity than 50 um Al)
21 mm tape width compatible with on-barrel engineering constraints
9 mm remain for power lines, shared equally between digital and analog
keep same design also for outer part
view barrel
design and forward
design
basic material from GTS Ltd. (holder of ISO 9001 )
matched product line
inner part - "thin" low-mass tape
bottom to top
outer part - "thick" low-mass tape
bottom to top
of 4.5 mm power lines
of 0.5 mm lines
all values with nominal values for Al and Cu resistivity,
view voltage
drop table for baseline SCT power distribution
input:
4*4.5 + 13(12)*0.5 = 24.5(24) mm (in brackets forward cables without the TEMP2 line)
X-section of plastic
Radiation length seen at perpendicular impact for one cable (w
-
width over which the material is spread)
single thin tape
X1 / X0 = (S/X0 (Al})
+ S/X0 ({plastic})) / w
= (1.23(1.20) mm2
/
89 mm + 3.4 mm2 / 350 mm) / w = (0.014 + 0.010)
mm / w
= 0.024 mm / w
single thick tape
X1 / X0 = (S/X0 (Cu})
+ S/X0 ({plastic})) / w
= (1.72(1.68) mm2
/
14.3 mm + 3.4 mm2 / 350 mm) / w = (0.0120(0.118)
+ 0.010) mm / w
= 0.130(0.128) mm / w
thin cable
Beware of fast conclusions from these numbers ! In the worst case, tapes are traversed twice and at an angle of 35o (barrel) or 13o (forward).spread over cable (w = 21 mm) X = 0.114 % X0 spread over module (w = 60 mm) X = 0.040 % X0 end of stave over module (6 cables) X = 0.24 % X0 outer barrel radius (1056 cables, r = 520 mm) X = 1056 * X1 / (2.pi * r) = 0.78 % X0 (this should be the maximum) at cryostat wall (1056 cables, r = 1150 mm) X = 1056 * X1 / (2.pi * r) = 0.35 % X0 at end of forward cylinder (988 cables, r = 590 mm) X = 988 * X1 / (2.pi * r) = 0.64 % X0 (maximum in forward)
thick cable
Space and routingspread over cable (w = 21 mm) X = 0.619(0.610) % X0 spread over cryostat bore (1056 barrel cables, r = 1150 mm) X = 1056 * X1 / (2.pi * r) = 1.90 % X0 on cryostat face (1056 barrel + 988 forward cables, r = 1200 mm) X = (1056 * 0.130 + 988 * 0.128) / (2.pi * r) = 3.49 % X0 (maximum)
X-sections, nominal thickness (measured somewhat smaller)
minimal bending radius: 10 x tape thickness (IPC-2223 5.2.4.2.)
result:
eps (Kapton) = 3.5
eps (glue) = eps (Kapton)
capacitance: C / l = eps * eps0 * w / d
inductance: L / l = u0 d / w
for comparison: flat (TP) ribbon cables: C/l typically
50 pF/m, L/l 700 nH/m
Production is with a standard photolithographic process adapted
to the size of the circuits.
Mask production
Laminate polishing
Photo-resist lamination
Photo-resist exposure
Photo-resist development
Conductor etching
Photo-resist stripping
Polishing
Ni/Au deposition
Laminate/glue/laminate/coverlay pressing
Tape cutting
Mask production
Laminate polishing
Photo-resist lamination
Photo-resist exposure
Photo-resist development
Conductor etching
Photo-resist stripping
Polishing
Ni/Au deposition
Laminate/glue/laminate/coverlay pressing
Tape cutting