1. 1 UA9 telescope first ideas Rome – 12/3/2010 Mark Raymond – m.raymond@imperial.ac.uk

2. 2 APV inner barrel sensor 12 96 CMS FED (9U VME) CMS LHC Si strip readout system APV25 0.25  m CMOS FE chip APV outputs analog samples @ 20 Ms/s APVMUX multiplexes 2 APVs onto 1 line @ 40 MHz Laser Driver modulates laser current to drive optical link @ 40 Ms/s / fibre O/E conversion on FED and digitization @ ~ 9 bits (effective) APVMUX laser driver lasers analog optical receivers ~100m analog opto-hybrid analogue readout x15,000

3. 3 8.1 mm 7.1mm pipeline 128x192 128 x preamp/shaper APSP + 128:1 MUX pipe logic bias gen. CAL FIFO control logic APV25 Peak Decon. 128 channel chip for AC coupled sensors slow 50 nsec. CR-RC front end amplifier 192 cell deep pipeline (allows up to 4  sec latency + locations to buffer data awaiting readout) peak/deconvolution pipeline readout modes peak mode -> 1 sample -> normal CR-RC pulse shape deconvolution -> 3 consecutive samples combined to give single bunch crossing resolution noise 270 + 38 e/pF (peak) 430 + 61 e/pF (deconvolution) note: only discrete 25nsec samples of above shapes are available in asynch. test beam choose timing to get close to top of peak mode pulse shape

4. 4 APV readout FED readout analog opto-link digital header 128 analogue samples APV O/P Frame 20 Ms/s readout -> 7  s no zero-suppression (sparsification) on detector pedestal, CM subtraction and zero suppression on FED raw data also available for setup, performance monitoring and fault diagnosis can read out raw data at low rate – VME - < 1 kHz can read out sparsified data faster – VME ~ 10 kHz (to be verified – some uncertainty here) Slink faster – 100 kHz – but needs incorporation (and customized use) of other CMS components (probably not possible this year) VME ~ 10 MB/s Slink to CMS DAQ trigger APV provides a timeslice of information from all 128 input channels following external trigger (trigger must be timed-in correctly)

5. 5 opto-electric conversion 10 bit 40 MHz digitization pedestal and CM subtraction hit finding (sparsification) formatting and transmission of data up to higher DAQ level check of APV synchronization all tracker synchronous, so all pipeline addresses of all APVs should be the same FED checks received APV pipe address matches with expected value (APV logic emulated at trigger level) off-detector FED functionality 9U VME

6. 6 PA APV inner barrel sensor 12 96 CMS FED (9U VME) UA9 telescope readout system APVMUX laser driver lasers ~100m analog opto-hybrid make use of most components but different sensors – no PA readout fibre ribbons plug straight into FED

7. 7 peltier heatsink fan HV, LV I 2 C, RST Ck/T1 AOH Al support plate with cutout beneath sensor D0 sensor 60 um pitch (+ intermediate strip) ~ 8 um resolution ceramic hybrid ceramic piece (same thickness as hybrid) telescope sensor module

8. 8 HV, LV I 2 C, RST Ck/T1 AOH sensor HV, LV I 2 C, RST Ck/T1 AOH sensor interface circuitry optical fibre adaptors power supply conditioning peltier cooling control ….. power slow control fast control (40 MHz ck, trigger) fibre ribbon readout XY plane crossover area ~ 4 x 4 cm 2

9. 9 250 mm ~50 mm baseplate (dimensions not critical) adjustable feet for levelling XY plane box (light tight)

10. 10 XY plane XY plane XY plane XY plane few 10’s m ~ m FED TTCexTTCvi VI 2 C SeqSi crate controller 9U/6U VME I2C: 1 bus per plane actively split inside plane module also opto-isolated Ck/T1: 1 shielded pair per plane CK/T1 combination at VME end (separate module) 1 fibre ribbon (50% utilised) per plane note: will need trigger to initiate APV readout (who will provide?) LV/HV power supplies not included here

11. 11 software first thoughts - not my area of expertise will need: setup lots of programmable parameters in CMS readout system bias levels, modes of operation, timing offsets (synchronize to beam trigger), … run control well behaved start/stop look after data storage, format? prompt data analysis “online” (provide feedback to setup) beam profile, signal amplitude histos, …. offline what is required?

12. 12 I 2 C link buffer opto- isolate I2C de-mux level shift VI2C buffer level shift 1st APV/opto hybrid 2nd APV/opto hybrid ancilliary I 2 C circuits ~ 10’s m within front end XY plane enclosure VME (1 channel) separate VME buffer module (4 chan – can also incorporate Ck/T1 opto-buffering) Ck/T1 link SeqSi 5V 2.5V Ck/T1 combine Ck T1 opto-buffer ~ 10’s m opto-receiver 1st APV/opto hybrid 2nd APV/opto hybrid opto-bufferopto-receiver 1st APV/opto hybrid 2nd APV/opto hybrid opto-bufferopto-receiver 1st APV/opto hybrid 2nd APV/opto hybrid opto-bufferopto-receiver 1st APV/opto hybrid 2nd APV/opto hybrid fibre-optic level shift resets

13. 13 up to 2 m 80 mm 50 mm Optical rail system X48 system from www.newport.comwww.newport.com assume this will sit on stable table (provided by someone else) feet allow some adjustment for levelling will still need some other mechanism for overall height