ATLAS SCT Detector Irradiation QA Procedures
FIRST DRAFT 28/05/99
- Baby Detectors
- Full-size Detectors
- Pre-Irradiation Tests
- Post-Irradiation Tests
- IV Curve
- CV tests
- Strip Integrity
- Leakage Current Stability
- Biassing Resistance and Interstrip Resistance
- Interstrip Capacitance
- Charge Collection Efficiency
- Strip Noise vs Bias
- Equipment Required
During production it is nessary to periodically verify that detector behaviour
after irradiation remains consistent with that observed during
the qualification tests. Any changes from qualification data
may indicate variations in processing which may only be evident
A very small number of full-sized
detectors will be selected for irradiation during production,
and a thorough evaluation of these detectors will be performed
for detailed comparisons with qualification data.
It is anticipated that larger numbers
of baby detectors (identical to the full-size detector but
only 1cm2 in size with 98 8mm-long strips) will be used for
irradiation tests, as their small size means that they can be irradiated
more quickly and easily. It is anticipated that
the measurement of the post-irradiation IV
characteristics of baby detectors will provide a minimum check of
1.1 Baby Detectors
It is anticipated that many detectors will be delivered to ATLAS
institutes with baby detectors diced from the same wafer.
At this stage
we would propose to use the baby detectors mainly for a check on
post-irradiation IV characteristics, though many of
the measurement techniques described in this document can in principle
also be applied to the baby detectors.
The main convenience for using these devices is their small size,
and therefore scanning during the irradiation at the PS will not be necessary
and the full fluence can be obtained in several hours.
1.2 Full-Size Detectors
Although full size detectors will provide the most realistic
comparison with qualification data, their size requires scanning
in the proton beam at the PS, and several days are required to
reach the full fluence.
2.0 Pre-Irradiation Tests
On delivery the full set of detector measurements described in
document ATLAS SCT/Detector FDR/99-7 should be performed to
ensure the detector is fully characterised.
The detectors are then glued with araldite 2011 to ceramic support
cards and bonded to pitch adaptors for compatibility with
readout by both binary and analogue readout electronics.
During irradiation the strip metals are shorted together (via bonds on
the pitch adaptor to a common rail) to simulate the condition of
being bonded to readout electronics.
The detector bias rail and backplane must be connected to ~3cm long
leads (via bonds to
the pitch adaptor and/or flexible PCB)
terminating in 2-pin SIL connectors for biassing.
The IV characteristics should be remeasured after gluing to ensure no
deterioration has ocurred during assembly.
3.0 Post-Irradiation Tests
Unlesss where otherwise specified, all post-irradiation detector
tests are performed cold (-10C) in a freezer containing dry air,
and with the detector ceramics screwed to an aluminium support frame
protected by an aluminium cover lid. To ensure good thermal
contact, the aluminium support frame should itself be in direct
contact with a large thermal mass inside the freezer.
Annealing times (when the detector is brought to room temperature for
measurements, bonding/soldering work etc) should be recorded in units
of days at 25degC equivalent temperature.
After irradiation the detectors should undergo a controlled beneficial
anneal for 7 days at 25degC to minimise the depletion voltage
3.2 IV Curve
To measure the IV characteristic after irradiation.
This test requires a voltage source/picoammeter (SMU) to measure the
IV characteristic between the bias rail and the backplane.
The current is measured every 10V step up to 500V, with a 10 second
delay between steps. The temperature of the detector should be recorded
(either via a PT100 on the detector ceramic, or a PT100 in contact with the
large thermal mass inside the freezer).
On a subsample of detectors, the IV should be remeasured at -18C to
verify that thermal contact is good and current scales correctly
The detector displays a characteristic at -10degC which is below ?uA
at 350V, and without any significant increase in current up to 500V.
3.3 CV Measurement
To measure the CV characteristic after irradiation.
This measurement requires a CV meter equipped (if necessary) with an
external bias adaptor and a voltage source.
Connect the biasrail to the AC
output of the CV meter bias adaptor, and the backplane to the
voltage output of the CV meter bias adaptor.
Record the capacitance in 10V steps up to 350V, with a 10 second delay
Use 100 Hz with CR modelled in SERIES, and with 500mV AC amplitude.
Plot the data as 1/C**2 (1/nF**2) vs bias (volts).
Interpretation in terms of depletion voltage is ambiguous after irradiation.
The data should be stored for information only.
3.4 Strip Integrity
To check for additional oxide punchthroughs caused by the irradiation.
This test requires a volt source/picoammeter (SMU) to check for oxide
punchthroughs, and a CV meter to measure capacitance anomolies
due to strip metal shorts/opens and pitch adaptor scratches/shorts.
The detector is warmed to room temperature and the lid of
the aluminium frame is removed. Any bonds still connecting strip
metals to the common ground rail of the pitch adaptor must be
removed. The frame is attached to a jig
on the chuck of an automatic probestation, and the planarity
adjusted such that the pitch adaptor is flat relative to the
All 512 metal pads of the pitch adaptor are then probed (in 4 groups
of 128) under computer control with the light on.
+10V is supplied continuously by the SMU to the backplane via the voltage output of
the CV meter (or the external bias adaptor of the CV meter if applicable),
with the needle connected to the AC output of the CV meter.
A series resistor may be necessary to limit current if the CV meter external
bias adaptor (which usually contains a large series resistance)
is not used.
The following technique for each strip measurement has been
demonstrated to work well without
any damage to the detector, and is therefore recommended, though
alternative techniques are acceptable:
The test (as demonstrated on a SUMMIT 10K probestation) takes about
- Before the first pitch adaptor pad is probed, source +10V from
the SMU and wait several seconds for the SMU current to
drop to <1nA (due to the capacitor charging in the CV meter external
- Raise the chuck to contact the pad with the probeneedle.
- Measure the current drawn from the SMU. Currents >> 1nA indicate
an oxide punchthrough.
- Measure the capacitance (1kHz, CR in SERIES, 100mV amplitude).
- In the case of an oxide punchthrough, drop the chuck and wait
several seconds for the SMU current to resettle to
<1nA. If there is not oxide punchthrough, no delay
- Move to the next pad, and repeast from Step 2 above.
The number of strip defects (due to oxide punchthroughs and strip
metal defects) is <2%. Care must be taken to exclude defects
arising from pitch adaptor scratches/track opens etc.
3.5 Leakage Current Stability
To check that any variation in leakage current over a 24
hour period is within specifications.
This test requires a voltage source / picoammeter (SMU), a meter for
temperature monitoring, an environment chamber, and, if
available, a switching matrix.
The backplane and bias rail
of the detector are connected to the high and grounded-low outputs of
the SMU respectively.
The assembly is installed
in freezer containing dry air (nitrogen) maintained
at -10degC. The bias is ramped to 350V, and after 60 seconds settling
time the current is monitored every 15 minutes over a 24 hour period.
Several detectors may be
measured in parallel by use of a switching matrix.
Maximum increase in leakage current
during 24hours is less than ?uA.
3.6 Polysilicon Bias Resistance and Interstrip Resistance
Determine the bias resistor value is within specificatons and that the
interstrip isolation is sufficient when under bias.
This test requires a voltage source and a volt source / picoammeter (SMU).
This measurement yields both the polysilicon bias resistance and the
interstrip resistance, and must be performed cold (-10C).
The backplane and bias rail should be connected to the high and
grounded-low outputs (respectively) of the voltage source.
The DC contact to a strip implant and bias rail should be connected to the high and
low outputs (respectively) of the SMU (it is nessary to bond from
the strip DC contact to a track on the pitch adaptor. Some bonds
from the strip metals to the pitch adaptor will need to be
removed to provide space for this).
Perform an IV (using the SMU) from -5V to +5V to
determine the resistance between the strip implant and bias rail as a
function of bias voltage (increase detector bias from 0 V to
~300V or until the measured resistance plateaus).
Interstrip resistance is sufficient if the measured resistance vs
detector bias plateaus. The plateau level resistance is equivalent to the
polysilicon bias resistance, and must be within 1.25 +/- 0.75
3.7 Interstrip Capacitance
Determine that the interstrip capacitance is within specifications.
This test requires a CV meter and a voltage source, and requires
the detector aluminium support frame to be attached to a
jig to allow for bonding from the strip pads on the pitch
adaptor to solderable contacts (eg a piece of PCB
with appropriate gold tracking).
Remove the bonds (if not already removed) connecting the detector strips
to the common ground rail on the pitch adaptor.
From one of the rows of 128 pads on the pitch adaptor that corresponds to 6cm strips,
bond one strip pad out to a solderable contact on the PCB, and bond the neighbouring
strip pad on both sides to the common ground rail of the pitch adaptor. Bond out from the
common ground rail to a second solderable contact on the PCB.
The central strip should then be connected (via a cable soldered to
the PCB) to the AC output of the
CV meter, and the two neighbouring strips to the
voltage output of the CV meter (sourcing 0V).
The backplane and bias rail of the detector should be connected to the
high and grounded-low sides (respectively) of the voltage supply.
Measure the capacitance between the central strip and its neighbour on
both sides as a function of detector bias up to 500V, using 20V
steps. Use 100kHz test frequency with CR modelled in parallel.
Note: parasitic capacitance arising from the PCB and its cabling to the CV
meter needs to be subtracted from the measured capacitance
values. The best way to estimate the parasitic capacitance is to
remove the bond between the pitch adaptor and PCB that connects the
central strip, and remeasure capacitance.
Interstrip capacitance < 1.2 pF/cm at 500V bias.
3.8 Charge Collection Efficiency
To determine the onset of the plateau in charge collection efficiency
vs detector bias up to 500V.
Bond one group of 6-cm 128 channels to an analogue readout chip and
measure the signal collected vs bias triggered using a Ru106
The onset of the plateau matches the value observed during
3.9 Strip Quality vs Bias
To determine the number of strips with excess noise due to
Bond all 512 channels of the pitchadaptor to either binary or analogue
readout electronics, and measure the noise per channel vs
detector bias at 200V, 400V and 500V bias.
There is no evidence of microdischarge at 400V bias.
4.0 Equipment Required at ATLAS Institutes
A list of equipment required at ATLAS institutes for testing of
irradiated detectors is shown below. Recommendations are given for equipment,
but alternative equipment is acceptable if the measurement specs
4.1 Automatic (computer controlled) probestation
eg Wentworth AWP-1050, Summit 10K, Alessi 6100 etc
For all probestation requirements. A computer controlled high
precision probestation is an absolute requirement for strip integrity tests.
4.2 Voltage source
eg Keithley 487, Keithley 6517
For IV, CV, strip integrity, stability and general tests.
Require capability of 500V or more, both polarities, compliance
up to at least 1mA.
Many of these tests are anticipated to take place in parallel, so
at least two voltage sources are recommended.
eg Keithley 487, Keithley 6517
For IV, stability, strip integrity and general tests.
Require sensitivity of 1nA (minimum) to at least 1mA.
Many tests will be carried out in parallel, so at least two
picoammeters are recommended.
4.4 LCR meter
eg HP4263B, Wayne Kerr 6425
For strip integrity and CV tests.. Frequency requirement
from 100Hz to 100kHz or higher. The LCR meter must have an external
bias adaptor for applying bias voltages up to 400V.
4.6 Temperature/Humidity Monitor
eg dedicated monitors, or Keithley 2000/2001 equipped with
Pt100/thermocouple, humidity probe etc
Temperature monitoring during IV and stability
4.7 Switching matrix
eg Pickering 20-360, Keithley range etc
For stability measurements.
4.8 Bonding machine
eg K&S manual or automatic
For bonding work.
4.9 Freezers, containing large thermal mass
eg Commercial chest freezers, with sheets of >=1inch thick aluminium.
To provide a cold stable dark environment for all tests (except
for the probestation strip test). The freezers should be
supplied with dry air (eg nitrogen or argon).
4.10 Networked PC with GPIB card and appropriate LabView control
software and database software
For all data-taking and updates to the ATLAS