1. 1 (Gerard Visser – STAR Integration Meeting 5/16/2008) STAR Forward GEM Tracker Readout/DAQ Integration G. Visser Indiana University Cyclotron Facility 5/16/2008

2. 2 (Gerard Visser – STAR Integration Meeting 5/16/2008) Modular readout crate architecture backplane ARCARM Optical fiber to D-RORC 7 pair cable or optical fiber from STAR TCD ALICE SIU (mezzanine board) Signal & Power cables, ≈ 8.3 m, to APV Connector/Motherboards Remote regulated DC power Two crates, each handles 12 cables, 10 APV’s per cable, a total of 15360 detector channels per crate ARM (APV Readout Module): 20 ADC channels and data processing FPGA’s (zero suppression, pileup rejection); power for APV on ARM or separate new board (4/crate) ARC (APV Readout Controller): control FPGA’s, STAR clock/trigger interface and ALICE SIU (data/control link) Connected by passive backplane, 30 MHz synchronous 24 bit datapath Uses commercial hardware (6U crate, VME P1 backplane) Crates mounted on west end ring of STAR magnet, e.g., in former location of SVT crate; magnetic field tolerant design

3. 3 (Gerard Visser – STAR Integration Meeting 5/16/2008) FGT DC Power and Grounding readout crate backplane Isolated dual DC supply (Wiener PL-508) 4x 4AWG sense lines 4C 18AWG +6 V @ 36 A, -6 V @ 22 A Quasi-isolated output +/-1.8V regulator Low-dropout low noise regulator +/-1.25V 3x 20AWGAPV25 ASIC’s 20AWG FEE GND OPTION JUMPER +1.8 V @ 0.90 A, -1.8 V @ 1.56 A STAR GND (magnet steel) GEM bias divider / bypass caps STAR GND (TPC structure) Non-isolated HV DC supply PREFERRED FEE GND OPTION ARM (only 1 shown here) APV Motherboard assy. (1 of 24 shown here) 8.3 m 24 m Local loads, not detailed here HV coax, shown here as two lines for clarity south platform 2 nd levelreadout crate (only 1 shown here) POWER SUMMARY (ENTIRE FGT LV) 240 APV CHIPS74 W FEE (MOTHERBOARDS) TOTAL106 W READOUT CABLES25 W READOUT CRATES577 W DC INPUT CABLES137 W DC SUPPLY LOSSES173 W OVERALL TOTAL1019 W Inside, at detector

4. 4 (Gerard Visser – STAR Integration Meeting 5/16/2008) (Stock) Cables from ARM to ACB/AMB Power & ground, 7C #20 AWG – Belden #5405FE, 5.1mm dia., 56.7 g/m +1.8 V force +1.8 V sense power return “ground” force ground sense -1.8 V force -1.8 V sense detector ground connection foil shield w/ #22 AWG drain wire Control & signal, 18Pair (3 unused) #28 AWG – Alpha #6398 or Belden #9819, 9.8mm dia., estimated 112 g/m CLK to ACB/AMB CLK loopback from ACB (arrives at ADC automatically in time w/ signals) TRG APV signals to RDO (10 pairs) I 2 C SDA/GND I 2 C SCLK/GND foil shield w/ #28 AWG drain wire, and tinned copper braid (rather not have it) Total cross-section passing last FGT disk: 19.2 cm 2 (20 power & 20 signal cables), total mass about 5.5 kg

5. 5 (Gerard Visser – STAR Integration Meeting 5/16/2008) APV analog output driving a long cable Noise level 11.5 bit dynamic range 1 1 1 0 1 0 1 0 1 0 0 1 With equalization, full swing is restored, sample-sample crosstalk almost completely cancelled, within 1 sample time (56 ns) Remaining <1% sample-sample crosstalk may be robustly removed with FIR digital filter APV ONLY – NO CABLE – 110 Ω LOAD19 m PVC CABLE IN/OUT (DOUBLE TERMINATED, AD8129 RECV.) In contrast to CMS, we will send the APV analog signals a considerable distance (8.3 m) from the detector, with no buffering or optical conversion at the detector. CABLE OUT (DOUBLE TERMINATED, AD8129 RECV., EQ. ) The APV “digital” header provides a convenient test pattern Works fine, even in this test w/ more than double the planned length (ok we could even consider to allow longer cable if we must… I’d like to know this soon!)