HCal Electronics and pulser B Quinn Dec 20
HCal Electronics (and pulser) B. Quinn Dec 20, 2017 1
Cabling scheme for HCal One signal, one HV cable to each tube. (Plus one HV & 6 signal/section for pulser) Cables secured, strain-relieved on gantry behind HCal. Cables held in place by gantry so they can be quickly re-connected to correct tubes. (… and gantry moves with HCal ? ? ) On HCal electronics/cabling platform (new) patch panel BNC/Lemo to short Lemo jumpers to amplifiers X 10 amp Patch panel for output signal to electronics hut (BNC RG 58 100 m. long) Lemo cable outputs to UVa summing module Discriminator, 16 -fold fanout, TTL converter for pulser Third patch panel in electronics hut …. Assuming Nino is used… Asymmetric passive splitter (board with connection/housing/power for Nino) Needs to be designed (JLab? ) noise shielding important Signal cables run to one of eighteen 16 -input BNC sub-panels (S 01 -S 18) Sequential inputs go to sequential channels of Nino, f. ADC Nino outputs go to VETROC-based TDC or F 1 TDC
patch panel PMT patch panel Custom passive asymmetric Splitter and Nino interface Shielded DAQ hut Lemo Nino Disc. 100 m LVDS f. ADC VETROC (or TDC) X 10 amp Sum of 4 Sum of 16 fan out Disc To Ecal trigger (10 regions) (18 groups) 4 quad linear FO (e. g. Phillips 740 2. 5 V) 18 UVa Summing mod. 36 in -> 18 in (16 in used) 3 quad 4 -fold linear FI (UVa 133) 12 chan L. E. Disc. (3 X 4 or 1 X 16)
{ { { 1 -1 Cabling pattern. S 01 S 02 S 03 Three 16 -channel slots for each four rows. S 04 S 05 S 06 1 13 1 9 b 3 5 9 1 1 S 10 S 11 S 12 S 13 S 14 S 15 S 16 S 17 S 18 24 -1 9 2 6 10 3 7 11 4 8 12 5 14 2 6 15 3 7 16 4 8 1 10 14 2 11 15 3 12 16 4 11 15 8 12 16 13 6 10 14 7 1 -12 1 1 S 07 S 08 S 09 Adjacent electronics channels don’t Fire on same LED pulse and unlikely to fire in same cluster 5 1 1 1 24 -12
Map from electronics channels to HCal module -Amplifiers -Patch panels -Ninos -TDCs (or VETROCs) -f. ADCs 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 S 01 1 - 4 1 - 7 1 -10 1 - 2 1 - 5 1 - 8 1 -11 1 - 3 1 - 6 1 - 9 1 -12 2 - 1 2 - 4 2 - 7 2 -10 S 02 2 - 5 2 - 8 2 -11 2 - 3 2 - 6 2 - 9 2 -12 3 - 1 3 - 4 3 - 7 3 -10 3 - 2 3 - 5 3 - 8 3 -11 S 03 3 - 6 3 - 9 3 -12 4 - 1 4 - 4 4 - 7 4 -10 4 - 2 4 - 5 4 - 8 4 -11 4 - 3 4 - 6 4 - 9 4 -12 S 04 5 - 1 5 - 4 5 - 7 5 -10 5 - 2 5 - 5 5 - 8 5 -11 5 - 3 5 - 6 5 - 9 5 -12 6 - 1 6 - 4 6 - 7 6 -10 S 05 6 - 2 6 - 5 6 - 8 6 -11 6 - 3 6 - 6 6 - 9 6 -12 7 - 1 7 - 4 7 - 7 7 -10 7 - 2 7 - 5 7 - 8 7 -11 S 06 7 - 3 7 - 6 7 - 9 7 -12 8 - 1 8 - 4 8 - 7 8 -10 8 - 2 8 - 5 8 - 8 8 -11 8 - 3 8 - 6 8 - 9 8 -12 S 07 9 - 1 9 - 4 9 - 7 9 -10 9 - 2 9 - 5 9 - 8 9 -11 9 - 3 9 - 6 9 - 9 9 -12 10 - 1 10 - 4 10 - 7 10 -10 S 08 10 - 2 10 - 5 10 - 8 10 -11 10 - 3 10 - 6 10 - 9 10 -12 11 - 1 11 - 4 11 - 7 11 -10 11 - 2 11 - 5 11 - 8 11 -11 S 09 11 - 3 11 - 6 11 - 9 11 -12 12 - 1 12 - 4 12 - 7 12 -10 12 - 2 12 - 5 12 - 8 12 -11 12 - 3 12 - 6 12 - 9 12 -12 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 S 10 13 - 1 13 - 4 13 - 7 13 -10 13 - 2 13 - 5 13 - 8 13 -11 13 - 3 13 - 6 13 - 9 13 -12 14 - 1 14 - 4 14 - 7 14 -10 S 11 14 - 2 14 - 5 14 - 8 14 -11 14 - 3 14 - 6 14 - 9 14 -12 15 - 1 15 - 4 15 - 7 15 -10 15 - 2 15 - 5 15 - 8 15 -11 S 12 15 - 3 15 - 6 15 - 9 15 -12 16 - 1 16 - 4 16 - 7 16 -10 16 - 2 16 - 5 16 - 8 16 -11 16 - 3 16 - 6 16 - 9 16 -12 S 13 17 - 1 17 - 4 17 - 7 17 -10 17 - 2 17 - 5 17 - 8 17 -11 17 - 3 17 - 6 17 - 9 17 -12 18 - 1 18 - 4 18 - 7 18 -10 S 14 18 - 2 18 - 5 18 - 8 18 -11 18 - 3 18 - 6 18 - 9 18 -12 19 - 1 19 - 4 19 - 7 19 -10 19 - 2 19 - 5 19 - 8 19 -11 S 15 19 - 3 19 - 6 19 - 9 19 -12 20 - 1 20 - 4 20 - 7 20 -10 20 - 2 20 - 5 20 - 8 20 -11 20 - 3 20 - 6 20 - 9 20 -12 S 16 21 - 1 21 - 4 21 - 7 21 -10 21 - 2 21 - 5 21 - 8 21 -11 21 - 3 21 - 6 21 - 9 21 -12 22 - 1 22 - 4 22 - 7 22 -10 S 17 22 - 2 22 - 5 22 - 8 22 -11 22 - 3 22 - 6 22 - 9 22 -12 23 - 1 23 - 4 23 - 7 23 -10 23 - 2 23 - 5 23 - 8 23 -11 S 18 23 - 3 23 - 6 23 - 9 23 -12 24 - 1 24 - 4 24 - 7 24 -10 24 - 2 24 - 5 24 - 8 24 -11 24 - 3 24 - 6 24 - 9 24 -12 5
Electronics racks on HCal gantry (2 relay racks) (HV patches and cables not shown) S 01 S 04 S 07 S 02 S 05 S 08 S 03 S 06 S 09 } } } 9 lemo cables 144 BNC cables from top half of HCal (~5 m long) 6 9 X 16 -chan BNC/BNC patch panel (To electronics hut) S 16 144 BNC cables from bottom half of HCal (~5 m long) 144 lemo cables S 16 9 16 -chan PMT amps S 10 9 X 16 -chan BNC/BNC patch panel (To electronics hut) S 10 S 09 2 x 144 Lemo S 09 S 01 144 lemo cables 9 16 -chan PMT amps 9 lemo cables G 9 16 -chan G r summing r UVa 120 18 10 G 9 16 -chan G r summing r 1 UVa 120 9 S 01 144 lemo cables 3 X 44 lin. chan 10 Fan- sum ch out UVa- LEdisc 133 144 lemo cables
S 16 9 X 16 -chan patch panel/ Nino interface (BNC from gantry) VETROC or F 1 TDC Lemo S 16 S 10 F 250 f. ADCs S 16 VETROC or F 1 TDC S 10 9 X 16 -chan patch panel/ Nino interface (BNC from gantry) S 09 F 250 f. ADCs S 09 S 01 Electronics racks in shielded hut (2 relay racks) LVDS (converted to ECL for F 1? ) 7
~ 0. 2 V signal amplitude => ~3. 0 n. Vs => 60 p. C into X 10 amp after amplifier (3 V max) ~2 V amplitude 30 n. Vs Time (ns) (from impact with front face) 8
After X 10 PMT amp ~2. 1 V ~3000 p. e !!! wrong! 9
After X 10 PMT amp and 100 m RG 58 cable ~720 m. V 10
Nino testing at CMU (Status report) Nino puts ~150 m. V bias on input signal (for me, board didn’t operate with back termination) Used 370 p. F blocking capacitor on input line (wrapped in grounded foil to shield) LVDS to ECL converter didn’t seem to work with 100 m. V offsets which Nino puts on output. Used two 330 p. F caps to block DC at Nino board. Nino can sustain very high freq. oscillations, possibly from Feedback picked up by input lines? Unplugging/plugging power connector reproducibly initiated oscillation. Cycling 11 power supply reproducibly stopped it.
First stage: Fixed-size LED pulses to PMT Trying to determine effective threshold: Nino threshold set to max by on-board pot. (Confusing first test: 760 m. V through 34 d. B didn’t fire Nino -> threshold > 15 m. V ? ? ? HF noise? ) Reproducibly 600 m. V signal through 40 d. B just fires Nino-> threshold ~6 m. V Timing resolution (best case) LED pulse (600 m. V) to LE disc. with 120 m. V threshold and to Nino through 26 d. B (20 X lower amplitude, 20 X lower threshold) Nino -> LVDS/ECL -> ECL/NIM -> TDC RMS resolution: Nino 182 ps LE 132 ps 12
Second stage: Fixed-size LED pulses to PMT, 100 m cable to Nino Timing resolution (still fixed signal size & shape) LED pulse (600 m. V) to LE disc. with 180 m. V threshold and to Nino through 20 d. B after 3 X amplitude loss in cable (30 X lower amplitude, 30 X lower threshold) RMS resolution: Nino 151 ps (LE still ~132 ps) 13
Third stage: Cosmic (vertical)Hcal pulses 100 m cable to Nino Cosmics ~110 m. V to LE disc. with 30 m. V threshold and to X 10 amp (~1. 1 V) to Nino through 26 d. B after 3 X amplitude loss in cable (6 X lower amplitude, 5 X lower threshold) Trigger counters to TDC Hard to identify peak in Nino signal…. ~ns width Suspect change in delay of 100 m cable over days of cosmic run 14
Fourth stage: Cosmic (vertical) Hcal pulses all timing signals through 100 m cable Discriminate trigger counters (and make coincidence) before cable, re-discriminated after cable then TDC Cosmics ~110 m. V X 10 amp to cable (/3) to splitter (/2) to LE disc. with 30 m. V threshold (18 m. V effective) and X 10 amp to cable (/3) to 26 d. B (/20) after (36 m. V effective) 15
Analysis of 10 k cosmic events No walk correction (ADC cut, To. F cut, ignore tail) Nino: s=446 ps Leading edge: s=333 ps With walk correction (thanks Juan Carlos) from ADC Nino: s=334 ps Leading edge: s=276 ps With (black box) neural network (thanks again) correction based on: Nino start time, time above threshold and ADC Nino: s=285 ps 16
Next stage: Drive PMT with variablebrightness LED pulser Digitally-selected brightness 0 & 1 -63 180 -11, 500 pe added optical filter between LEDs and fibers 37 -2200 pe (c. f. ~400 for cosmics) Can simulate pulse height variation of cosmics to try to understand timing behavior. Take data at 100 Hz instead of 1 every 15 minutes! 17
ADC spectra for bit patterns 0, 7, 15, 23, 31, 39, 47, 55, 63 Pulser & fibers are very stable so number of p. e. can be found from ((Mean – pedestal)/width)2 18
Because of fiber-to-fiber variations (and rough LED calibration) number of p. e. is not a monotonic function of pattern But can be calibrated in a half hour pulser run (for all 288 PMTs) and has shown great stability. Allows setting of proper HV and gain matching in advance 19
Status Programmable pulser moved from test stand (light-tight box) to same HCal module used for cosmic tests. Fibers inserted. Control cables run, loading of patterns works, debugging pulse command to board Data with Nino card soon. 20
- Slides: 20