Title: n-XYTER Front-End Boards
1n-XYTERFront-End Boards
Christian J. Schmidt, GSI Darmstadt
CBM STS Workshop, Karelia, June 1 4, 2009
2Data Driven Front-End Asynchronous Channel
Trigger
detection of statistical, poisson distributed
signals
triggertimestamp reg.
comparator
Time WalkCompensationcircuit
FASTshaper 18.5 ns peaking
PDH reset
chargepreamp
dig. FIFO
chargeinput
SLOW shaper(2 stages) 140 ns peaking time
Peakdetector hold, free running
pulse height output
analogue FIFO
Asynchronous registry and storage in 4-level fifo
guarantees data loss lt 4 when read-out through
token ring
Read-out through balancing token ring ? 32 MHz
data registry and read-out
The DETNI ASIC 1.0, a front-end evaluation chip
in AMS 0.35µ
3n-XYTER Front-End Topology
4"Simple" FEB for the n-XYTER Starter Kit
- A simple hybrid PCB with signal fan-in, ADC and
interconnect to SysCore DAQ chain - Allow development of the DAQ chain
- Allow the readout of various detector prototypes,
using the n-XYTER front-end chip
5n-XYTER FEB At the limits of PCB-technology
Interference point of many technologies, each
imposing limiting boundary conditions
- A simple hybrid PCB with signal fan-in, ADC and
interconnect to SysCore DAQ chain
- Chip-In-Board solution avoids space eating vias
- allows pitch adaptation
- 50,7 µm on chip to
- PCB side 101,4 µm on two levels
6n-XYTER FEB Rev. C
- In spite of bonding difficulties, managed to
realize - 10 Rev. C boards, fully functional
- 4 rev. B boards with 64 channels functional
(the other 64 may be employed for direct
silicon det. tests)
We are striving to get the entire system (FEB,
SysCore, software) packaged to allow shipment.
Starter kits in distribution
Teams from Colcatta, Kiew, Dubna, Dubna (MPD),
GSI, Heidelberg were introduced into running the
starter kit with FEB and data chain at the GSI
detector lab
7Complex Mulitlayer Data Chain
- Detector
- n-XYTER-FEB
- n-XYTER
- FEB and Bonding Technology
- Documentation (Manual, newest version
this May with update on parameter description) - ADC
- Interconnect
- SysCore
- Firmware
- Embedded software
- Soft configuration
- Ethernet-Interconnect
- PC and DAQ
- KNUT
- GUI
- DAQ System DABC
Software design freeze 1.7, firmware, software,
knut Under heavy development (Anton Lymanets)
low level diagnostic toolbox for system analysis
to make successful deployment in other labs
feasible
8Additional Mimicing Triggered Mode of Operation
- Discussions with Petr Nomokonov revealed a nice
and simple mode of forcing n-XYTER to operate
classically in triggered mode - Switch chip to test trigger mode
- Switch off test pulse enable (important as
would otherwise imply a special test mode) - ? Test trigger signal will sample all signals
during monostable activity, window determined by
"iota" parameter 120ns to several µs. - Alow to study noise all the way to the pedestal,
where self triggered normal mode of operation
would drown in triggers!
9Test Trigger Mode alows globally triggered signal
sampling
test trigger input in test trigger mode
triggertimestamp reg.
comparator
Time WalkCompensationcircuit
FASTshaper 18.5 ns peaking
PDH reset
chargepreamp
dig. FIFO
chargeinput
SLOW shaper(2 stages) 140 ns peaking time
Peakdetector hold, free running
pulse height output
analogue FIFO
The DETNI ASIC 1.0, a front-end evaluation chip
in AMS 0.35µ
10n-XYTER FEBs
- FEB-Rev D next version to implement all
learnings - Relieve PCB specs by tightening bonding
necessities and challenge - Chip-In-Board, two layer bonding and metal in-lay
cooling contact - GEM-TPC FEB ongoing dual n-XYTER for PANDA
GEM-TPC - n-XYTER Quattro STS-Baby-Sensor double sided
readout ? towards a beam telescope - Mid-term Hybrid development for demonstrators
(see Volker Kleipas talk tomorrow) - several n-XYTERs on Silicon-circuitboard
11n-XYTER Quatro Double Sided Silicon Readout
- Read-Out of double sided Si, 256 x 256 channels
- First setup to occupy complete capacities of one
SysCore - Several Moduls may be set-up to make a beam
telescope - FEE of both sides live on one potential!
? Rely upon readout-caps!
12n-XYTER Quattro
- four chips bonded on one board (Carmen Simons at
GSI DetLab) - currently under tests
- then
- bond a CBM baby sensor on one side only and test
- bond both sides and test
- We will make 4 3 spare beam telescope elements
for September beamtime at GSI
13Feb Rev D, Layout Start a new with Eagle
Software
Enhance availability of starter kits
- Eagle allows script based layout
- Generate real footprint
- Much easier to follow nets
- Easily make bond-plan
- Input pitch 5.4 mils 137µ
- Max bonding angle 22
- Max bond length 3.4 mm
- Targeted but may not be ready for beam-time
14Preparation of Engineering Run (H.K. Soltveit)
- Test Pulser
- Identified very narrow line in test-pulse engine
that introduces non-linearity of pulser - Cal setting changes power consumption 0 to 10mW ?
implications to DC-level in current prototype
through Temp Coeff. of shapers - Discriminator was designed as "above pedestal
trigger" - ? not linear over full threshold span i.e.
S-curves are non-linear, explains beamtime gain
conundrum - Temperature coefficient in shapers addressed
- modification gives 0.2 mV/K, down from 20 mV/K,
simulated 50 to 90C - Observed Monostable crosstalk identified and
addressed
End of June 2009 Engineering run schematics
review meeting End of July 2009 Submission
readyness review meeting Then submission Expect
chips in Sept/Oct.
15Some In-Channel Discriminator Feedback Detected
...upon removal of discriminator-power decoupling
correlates with internal discriminator trigger
correlates with external test-pulse release
signal (blue)
These issues are particularly important with the
self triggered architecture! They will be
addressed even more in the next engineering run.
16Investigating Individual Channels,
Triggerefficiency
Trigger efficiency in Treshold Scan The
S-Curves used to be the technique to get analogue
signal information w/o ADC
Caveat information on discriminator operation
Derivative gives image of noise!
This scale is not linear with pulse height,
i.e. input charge!
- Input of test pulses at fixed rate, - scan
threshold while measuring detection rate
17Thermographic Analysis n-XYTER in Action
- Chip temperature around 70C on FEB with simple
passive cooler - Digital readout and Time Stamp Clk Distribution
surprisingly the hottest part
18Conclusions
- Starter Kits begin to become available
- Feb Rev C fully bonded with SysCore
- Feb Rev D ? high yield, fabricatable FEB
- Data chain went through extensive debugging
phase, currently software consolidating, tools
for beam-time - Engineering run really in preparation, chips
expected Sept./Oct. - many learnings addressed
- future n-XYTER will be cured from child deseases
and design-mishaps - black box operation much more feasible
- n-XYTER on Silicon circuit board technology on
the radar - parallel activity on CBM-XYTER development ongoing
19(No Transcript)
20New Detector Lab Facilities at GSI
Construction of the new GSI Detector Lab has
commenced! Located between GSI UNILAC and the CBM
Offices (C27)
- 600 m2 of climatized cleanroom space with
relative humidity down to 20 - cleanroom class 10000 in general and better
locally - dedicated bonding labs forseen
- Additional 600 m2 available for detector testing,
storage and other activities - Additional 100 m2 of high clearance (10m
headroom), dedicated device setup space (detector
integration and tests)
21(No Transcript)