Transcript Cross-talk, past & present

LHCb Outer Tracker
Cross-talk measurements with Fe-55
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
1
Cross-talk, past & present
• A particle traversing a straw
tube causes not only current in
anode wire but also produces
electrical signal induction in
other electrodes, such as
neighbouring straws. This is
cross-talk.
• CT was investigated on a prototype module in 2000 by
Gromov & Sluijk. For a Kapton XC + aluminium shielding
the analogue CT was measured to ~0.5% (adjacent).
• This study concerns CT between adjacent wires in same
plane. Measurements done on module #2.
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
2
The choice of source
• A Fe-55 source was used to induce
ionization in the wires.
– Primarily emits 5.89 keV photons from EC.
– A photon induces a cloud of ~226 electrons
in one straw tube.
– If a signal is recorded simultaneously in an
adjacent wire it is highly probable that it is
cross-talk
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
3
Some words on the gain
• To read out signals from
module the latest VV-50
preamp board from
Heidelberg was used.
• Calibrated with pulse
generator
• Saturates just around 1800V
• Gain calculated using conversion factor 0.42 V/pC (from
calibration)
• Gain increases rapidly in region 1500-1800 V
• It is expected that cross-talk increases with higher
voltage → measurements done at 1800V.
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
4
Experimental setup
• Wires are connected to HD-preamp board which
also supplies HV.
• Output from preamp on two separate channels
to scope card in computer.
– Active channel is illuminated by source, used as a
trigger for the
– Channel of interest
• adjacent wire (to measure cross-talk)
• distant wire (for reference)
• LabView is used to set trigger levels, sampling
rate etc. Transfers digitized signals to txt-file.
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
5
Cross-talk measurements
• CT between straws , 1800V
– Adjacent wires (32/33)
– Distant wires (32/48)
• Electronic CT on preamp, disconnected from module,
using pulse generator
– Adjacent channels (32/33)
– Distant channels (32/48)
• To eliminate electronic noise the signals are averaged
+
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
+
→
6
Pulse characteristics
• Pulse generator was set to
give an output signal looking
like a Fe-55 signal in
amplitude and time duration.
1800 V
Pulse gen.
Fe-55(averaged)
1800 V
Flatness due to range
settings in read-out
x-axis in ns, y-axis in V
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
7
Electronics cross-talk
-
Adjacent channel 32/33
=
Distant channel 32/48
ch33-ch48
• Dotted line=active channel, solid line=ch. of interest*100.
• Electronics cross-talk is present. Ratio between
amplitudes gives a measure of the Elec-CT: ~ 2‰
• CT is bigger on adjacent channels, right plot shows
difference between ch33 and ch48. Differences are of the
order of 0.5 mV.
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
8
Cross-talk between wires
• Since elec. CT is present it’s expected to make up one
component in the total CT measurement.
-
Adjacent straw 32/33
=
Distant straw 32/48
str33-str48
• Similar patterns as in the electronics CT measurement
can be seen, the elec. CT component is present. The
question is of course…
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
9
…can we disentangle straw CT?
• Due to differences in time duration of pulses…
– Fe-55 pulse ~ 175 ns
– Generated pulse ~ 110 ns
… the two “difference plots” cannot easily be
subtracted from each other in order to remove
the electronics cross-talk component.
• There are still some observations that can be
done though at this stage:
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
10
“First peak” observations
str33-str48
0.70 mV
ch33-ch48
0.50 mV
“First peak”
Due to differences between
active signals the red line is
drawn to indicate
comparable regions in time
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
Pulse gen. signal
amplitude is 20% higher
than Fe-55 but the first
peak in Fe-55 difference
plot is 40% bigger than
pulse gen. difference
peak…
11
A straw cross-talk estimate
• If one assumes that the amplitude of electronics
CT grows linearly with the pulse amplitude on
the active channel then
– The Fe-55 first peak should have a elec. CTcomponent of about ~ 0.4 mV…
… but the Fe-55 first peak has an amplitude of 0.70
mV!
• So, from this (too?) simple reasoning one can
draw the conclusion that the straw CT
component is ~ 0.30 mV
• Analogue CT between adjacent straws ~
RCT= 0.30 mV / 500 mV = 0.6 ‰
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
12
Conclusions and outlook
• Hard to resolve inter-straw cross-talk due to electronics
cross-talk on preamp board.
– In any case, the analogue CT is less than 2‰
– Could be as small as ~ 0.6‰
• The straw tubes are well-isolated so a possible scenario
is that the main source of CT will come from the read-out
electronics.
• For more accurate measurement:
– Re-measure with separate preamps (not possible with current
VV-50 preamp board).
– The other preamps available now (old HD-preamp board and
Gras’ preamp) are not silent enough for this.
Aras Papadelis,
OTR meeting NIKHEF, June 17th 2004
13