Status of n-XYTER tests for GEM

Report
Status of n-XYTER read-out chain
at GSI
Rahul Arora
[email protected]
GSI, Darmstadt
GEM-TPC Meeting,18th November 2009
n-XYTER : Neutron- X, Y, Time, Energy Readout ASIC












Front-End:
128 channels
Front end for either polarity input signals
Charge sensitive pre-amp and peak detector
Time stamping with 1 ns LSB, 2ns resolution
Peak detection and analogue storage (10 bit electronic resolution)
Purely data driven, autonomous hit detection (self triggered)
Average per channel hit rate 160kHz with 10% dead time
(determined by pile-up on slow channel)
Readout:
Per channel analogue energy and digital time stamp FIFO (1ns
resolution)
De-randomizing, sparcifying Token Ring readout at 32 MHz
2
Data Driven Front-End: Asynchronous
Channel Trigger
Detection of Statistical, Poisson distributed signals
charge
input
FAST
shaper 18.5 ns
peaking
Time Walk
comparator Compensation
circuit
charge
preamp
SLOW shaper
(2 stages)
140 ns peaking time
trigger
timestamp reg.
PDH
reset
Peak
detector &
hold,
free running
dig. FIFO
pulse height
output
analogue FIFO
Asynchronous registry and storage in 4-level fifo guarantees data loss < 4%
when read-out through token ring
3
n-XYTER Front End Topology
4
Power Consumption







preamplifier
fast shaper
slow shaper stage 1
slow shaper stage 2
discriminator
peak detector and hold
analogue FIFO
7.4 mW
2.5 mW
1.7 mW
2.5 mW
2.1 mW
2.7 mW
2.3 mW
Overall we find about
21 mW/channel
 air-cooling needed and installed for the tests
5
Experimental Built-in Safety-Belts with
First Submission
Testability and Diagnostics






Individual analogue test
channel
Built in test pulse generator
Programmable mask for
every channel
Programmable, forced trigger
of PDH for every channel
(check signal pedestal)
Programmable dead time
Diagnostic counters for pileup and token statistics
Backup Previsions /
Safety





Every individual channel may
be shut off
Clock signals derived on chip
may be fed in separately
Time stamp clock may be
reduced without change of
readout clock
Various parts may be shut off
In addition to global threshold,
5 bit programmable local
threshold
6
N-XYTER FEB
•Dimensions: 91 x 95 mm
•8 layers
•Revision C
7
Readout chain








The readout chain starts with the Revision C board having 50 micron silicon
strip detector on it having detector capacity around pF.
The source for the detector is Am140
The next step is the ROC board connected to the FEB by flat cable.
The DC voltage supplied to the ROC and FEB is as follows
 For ROC : Voltage - 5 V
Current - 3 Amps
 For FEB : Voltage - 5 V
Current - 1.5 – 2.0 Amps
The versions of software and hardware used are
 ROC hardware : 1.7.2.21
 ROC software : 1.7.2.0
The ROC is communicated using ROCsoft software in a PC running linux.
The analysis is done using go4 sofware 4.3.1
The FEB is maintained at a constant temperature of 20 °C using air cooler.
8
Testing of one chip board
The pulse shapes are studied from the slow and fast
shaper using the oscilloscope.
The one chip board with silicon strip detector with pitch
of 50 micron is tested using Am140 .
The results are analyzed using go4 software.
9
Go4 plots for ADC count and channels
ADC count for all channels
10
Event signal with baseline subtraction
and test pulse signal
noise
noise
Test Pulse
Am140
11
Overall status & Known bugs

Operational tests:



Full chain is operational
Drift with temperature,
but if stabilized works for days
Dedicated tests (with developer)


Cross talk (?) may be present
(mixed layout?)
Temperature dependency of E-output
(work below 60 °C)
12
Outlook




2-chip board will be attacked as for
as possible
Input to be changed to gas-detector
Data for the engineering run nearly
complete
Submission foreseen before end of
2009
13

similar documents