Transcript 01.Calabrese.proposal_LST_nov14

Proposal for LST-based
IFR barrel upgrade
Roberto Calabrese
Ferrara University
Workshop on IFR replacement, SLAC, 11/14/2002
Outline








General Overview
Layout geometry
Performance
Readout methodology
Electronics
Gas, HV, DAQ
Costs
Other presentations
General overview

The entire project is driven by the allowed space

The intrinsic efficiency of a standard LST tube is about 90%.
This is due to dead spaces in the LST tubes.
Efficiency is too low for our purposes. Not enough space to
put 2 standard layers.
Possibilities to improve efficiency
(given the allowed space)

Option 1: single-layer with a large cell (19x17 mm)
Readout of x and y coordinates from outside strips
Possibilities to improve efficiency

Option 2: double-layer with a small cell (9x8mm)
Readout of x coordinate from wire and y coordinate from outside strips
Possibilities to improve efficiency

Option 3: modified double-layer with a small cell (9x8mm)
Readout of x and y coordinates from outside strips
Detector layout: segmentation of a detector layer
PCB for cable connectors
Z strip signal
collection PCB
Z strip signal
collection PCB
Z strip
readout
Layer of
LST
Ф strip
readout
Ф strip signal collection PCB
(a similar one in the opposite corner)
PCB for cable connectors
Servizio Elettronico INFN Ferrara
Detector layout: details of strip signal collection PCB
Servizio Meccanico INFN Ferrara
Performance


Expected efficiency about 96%
Position resolution
better than 3 mm (z)
for standard LST
better than 9 mm ()

We do not need such a resolution and we can
increase the strip width, thus decreasing the
number of channels (  MC simulation)
Detector layout: questions

How many strips?
resolution,

cost, space
How many chambers/layer ?
installation,
cost
Detector layout: answer to the questions

We are considering 4 cm z-strips and 4.3 cm -strips
(2 -strips (along the wire)/LST tube)

96 z-strips for each layer, total 6912 z-strips

74 -strips for the outer layer, total 4572 -strips
About

11500 channels of electronics
2 chambers/layer (remove only one corner block at the time),
but the number of z-strips doubles (more cables, but same number
of electronics channels) or we need to decouple z-strips from the
chambers
Readout methodology

Only digital readout of strips

Time measurements could be implemented:
 OR
of 16 discriminated pulses
 Time resolution about 16 ns ( using BaBar reference clock)
 Implemented with FPGA
Electronics

FEC gain depends on the shape of the signal
 FEC ampli cannot be used

Existing FEC: to be modified if we want to use
them (when?)
 the baseline is to use new electronics
Front end module design : block diagram of
the NEW 96 channel (1 view of 1 layer) FEC
Threshold
96x Amplifier-Discriminator
Cost per channel inclusive of:
Shift/Load 6 x
Ck_Chain
Shift/Load
SHIFT REGISTER
6x
-PCB
12us Digital OneShot
96 x
11us Digital OneShot
12us Digital OneShot
11us Digital OneShot
-components
-crate&power supply
10 € / channel
Data Out
6x
To the Active Patch Panel
ahead of the FIFO Board
Implemented in a single high performance FPGA
(Field Programmable Gate Array)
Servizio Elettronico INFN Ferrara
Front end module design : schematic of the
front end based on Off-The-Shelf components
Power dissipation: 250mW
Servizio Elettronico INFN Ferrara
Front end module design : analog simulations,
effect of strip capacitance and impedance
b)
a)
c)
Simulation of the amplifier/discriminator
output from a 4pC input signal
(0.1mA * 40ns)
Comparator threshold = 50mV
dielectric thickness 0.75mm
a) dielectric FOAM (εr=1)
b) dielectric PTE (εr=3.3)
c) dielectric FR4 (εr=4.8)
Servizio Elettronico INFN Ferrara
Gas system



mass flow control system
main gas transport pipe system (existing)
final gas distribution and bubbling system.
We assume all the tubes in a layer with a single in/out

Safe gas mixture, like Ar/Iso/CO2 (2.5/9.5/88) (SLD)
HV, DAQ

Each tube has a separate HV connection
( 2 for a double layer tube)
 Possibility to
use a commercial HV system for LST available
from CAEN (SY546 mainframe + 12 channel boards A548),
150 $/channel

DAQ
 No
change, all signals are FEC-like
Costs: assumptions
double layer LST with 8mm cells; each layer with a
separate HV; each ground returning separately through a
measuring resistor
 96 strips (40 mm strips) in the z direction; 2 strips per
LST in the  direction
 2 chambers/layer (double the number of z strips signals)
 12 active layers

Costs (I)

Tubes:
 30
K$ (setup)
 450 K$ ( about 200 $/tube x 2280 double layer tubes)
Total cost tubes 480K$
(using single layer tube this cost would be about 300K$)

Strip readout planes
240 m2/layer x 6 layers x 50 $/m2 = 72 K$
Signal collection (PCB’s) inside iron 18 K$
Total cost readout planes 90 K$

Grand total chambers
570K$ (double layers);
390K$ (single layers)
Costs (II)
Total cost of signal flat cables 146 K$ (outside iron)
15 K$ (inside iron)
 Total cost of signal flat cable connectors 28 K$
 Total cost of electronics 115 K$
 Service panels 4 K$
 Total cost of LV ground wire 5 K$
 Total cost of LV plugs 4K$
 TDC system 20K$ (optional)


Total electronics + cables
322 K$ (maximum); 171 K$ (minimum)
Costs (III)

Total cost of HV wire 23 K$
Total cost of HV banana plugs 9 K$
Total cost of HV power supply+distributors 170 K$
Total cost of mass flow control system 11 K$
Total cost bubbling system 4 K$
DAQ, cooling no expected cost

Total this page 217 K$

Grand total detector 780 K$ 1100 K$





Addressing the various issues

R&D issues and status
Changguo Lu

LST experience and
production issues
Mario Posocco
Installation issues
Livio Piemontese
