Development of Calibration system for AHCAL 1
Requirements to calibration system Generate uniform near-visible UV flashes controllable in amplitude 0 to max = twice SiPM saturation adjustable pulse width (a few ns ) enabling each LED individually optical feedback from LED to PIN-PD signal channel LED triggering from DAQ Readout temperature from 5 sensors placed in the scintilator plane (12bits minimum) CANbus interface to Slow-control 2
Principal schema of LED-driver My concept Old concept as used in H1 Spacal calorimeter 3
Short history of the last 2 years CMB = Calibration and Monitoring Board 2004 Sept/Oct first ideas where to and how to go 2004 Dec 1CHannel LED-driver + PIN PD preamp 2005 Feb 2CH LED-driver 2005 Aug CANbus + Temperature readout /Slowcontrol interface 2005 July/Aug first version of 12CH CMB 2005 Dec prepared version of 12CH CMB 2006 Jun 1 st production of 20pcs CMB 2006 Sept 2 nd production of 20pcs CMB 4
Picture of 1CH LED-driver Nov 2004 5
Version of 2CH LED-driver Common input for signal LED-trigger, lemo, 50Ω One 12V (60mA) supply, protection Rise time 2ns, tested with P5050 scope probe (8pF, 500MHz) C=8pF, L=30nH f res =290MHz 24feb2005 6
Output, one LED leg, 500MHz passive probe this is a true RF technique :) 24feb2005 Zoom 2ns/div 7
3 different scope probes, what we see... JULY 05 LED driver Oscilloscope TDS5104, 1.5GHz Probes TEK GRN 16pF 200MHz Cyan 11pF 500MHz VIO 1.5pF 3GHz Same circuit at once P6185 passive 1kOhm 1.5pF 20x (3GHz) P5050 passive 10MOhm 11pF 500MHz 10x P2200 passive 10MOhm 16pF 200MHz 10x 8
UV-LED parameters We have to made own datasheet, due to lack of the information. To sort 500 pcs best out of 1000 pcs 9
UV LED forward V-A char Apr2005 YEL UV LED 10
Emission spectra of UV-LED normalized emission intensity [arb. units] 1.2 1 0.8 0.6 0.4 0.2 0 UV LED diode no. 1 emission spectra 380 390 400 410 420 430 440 450 460 wavelength [nm] 10 μ A 100 μ A 1 ma 3 ma 10 ma normalized emission intensity [arb. units] 1.2 1 0.8 0.6 0.4 0.2 0 UV LED diode no. 2 emission spectra 380 390 400 410 420 430 440 450 460 wavelength [nm] 20 μ A 100 μ A 1 ma 3 ma 10 ma 11
UV LED homogenity measurement 872 LEDs Temp corrected Calibrated Area Area [nvs] 45.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 300 400 500 600 700 800 900 1000 1100 LED # fiber 11 fiber 3 fiber 6 fiber 8 Tcorr Calib SUM 12
Selection criteria LED# 640-1179 (532pcs) 100 90 80 σ = 12.7% 70 60 50 40 For details see J. Zalesak, CALICE 2006, Montreal 30 20 10 0 under 0.50 0.60 0.70 0.80 0.90 1.00 1.10 1.20 1.30 1.40 1.50 LED # koef 100% ± 20 % ± 15 % ± 10 % ± 5 % RMS 1.5 1.35 1.22 1.11 all 1.000 872 773 650 498 260 13.1% 100.0% 88.6% 74.5% 57.1% 29.8% 640--1179 0.955 532 498 434 309 176 12.7% 100.0% 93.6% 81.6% 58.1% 33.1% 13
CMB = LEDdrv + PINdiode preamp + CANbus + temp readout CALICE oct05 All TESTED LVDS 14
15
LED driver temperature test Oct05 An ice on plastic house of LEDdrv chip Cooled by KALTE spray upto -40deg.C Warmed by solder tip to different components GND pins Uncalibrated, but we know a scale ~200deg.C 16
LED driver jitter JULY 05 TDS 5104 2.5Gs oscilloscope Fast probes Jitter = 200ps 1hr at 10kHz, same jitter as running 15min on picture Zoom 500ps / div 17
LED-driver test at Prague Oct 2005 Meas. by Jara Zalesak Single UV-LED response to LEDdrv linearity test 9 of 12 fibres measured by very low gain APD @190V M~2 to 3 Preamp + DAQ 10bit ADC Camac LV[V] is a recent controlling voltage 18
CMB VFE CMB T T T T Goettlicher, Reinecke T HBAB_TOP Polak, Janata, (Karstensen) T Gadow HBAB_BOT T T T CAN (ADC, Enable LED, DAC) T T 19
CMB Installation to AHCAL at CERN 20
Saturation curve of SiPM (linearity test of one plane) LED -> SiPM Made by Nanda Wattimena 21
SiPM response to low LED light 1 LED illuminate 18 SiPM See Sebastian Schaetzel talk 22
Conclusions LEDdriving system is temperature pretty stable, well bellow 1ns @ normal range (40degC) Cooled each one adds -300ps delay, heated some +300ps each Total delay is in range of -1ns to +1ns for estimated temp. range 200deg.C Spread of the Delay in one plane (12LED) around 1ns 1hr Jitter is in range 300ps 23 CMB installed into HCAL prototype CMB fullfil requirements details see in S. Schaetzel talk 23
Acknowledgment Inst. Of Physics Prague Jaroslav Cvach Milan Janata Jan Šťastný Jaroslav Zalešák DESY Hamburg Karsten Gadow Sven Karstensen Hendrik Meyer Sebastian Schaetzel Nanda Wattimena 24
BACKUP 25
UV-LED tests Apr 2005 Volt-Amp characteristics DC-mode Forward and reversed characteristics Tested with Characteroscope, 300μs pulses at 1kHz Compared with precise 6.5 digits DMM at lower current Optical spectrum, linearity test DC-mode Measured tnx to Martin Nikl, Inst. of Physics Prague Test equipment spectrometer ORIEL 50540 Small statistics 4 pcs of UV LED only 26
Wavelength emission of UV-LEDs Peak and spread derived from Gaussian fit on LED emission frequency spectrum (-2,+1)*σ Peaks @ (400±1) nm for all 4 UV-LEDs and all light intensities Spread of emission (1 sigma from fit) in average: (4.8), 4.9, 5.1, 5.4, 6.6 nm for (0.02), 0.1, 1.0, 3.0 10.0 ma Jaroslav Zalesak, apr05 27
ns technique is not trivial 28
UV-LED sorting 4 5 12 1 6 fibre profile 11 3 2 7 10 9 8 made by Jara Zalesak 29
LED pulse: blue & red in coincidence UV led LED pulse UV led 30
LED intensity vs current (coarse measurement) UV-LEDs Green LED Intensity determined as height of gauss. fit on emission freq. spectrum light intensity at spectrum peak position) Assumption of same spectrum shape (peak position and width as well) in all currents, else needed integration in some freq. range (~20% variation 1-10 ma) Not fully comparable absolute light intensity (adjustment of measurement uncertainty, small variation in spectrum shape, few points) Jaroslav Zalešák, apr05 31
UV LED forward V-A char 300 µs pulses at 1kHz rate i.e. 33% duty factor 300mA UV LED is more sensitive to the overheating than common LED damaging Ivo Polak, apr05 32
UV LED reversed V-A char Ivo Polak, apr05 33
Spectra of UV-LED normalized emission intensity [arb. units] 1.2 1 0.8 0.6 0.4 0.2 0 380 390 400 410 420 430 440 450 460 Ivo Polak, apr05 UV LED diode no. 1 emission spectra wavelength [nm] 10 μa 100 μa 1 ma 3 ma 10 ma normalized emission intensity [arb. units] 1.2 0.8 0.6 0.4 0.2 1 0 UV LED diode no. 2 emission spectra 20 μa 100 μa 1 ma 3 ma 10 ma 380 390 400 410 420 430 440 450 460 wavelength [nm] 34
Spectra of UV-LED normalized emission intensity [arb. units] 1.2 1 0.8 0.6 0.4 0.2 0 UV LED diode no. 3 emission spectra 380 390 400 410 420 430 440 450 460 wavelength [nm] 20 μ A 100 μ A 1 ma 3 ma 10 ma normalized emission intensity [arb. units] 1.2 1 0.8 0.6 0.4 0.2 0 UV LED diode no. 4 emission spectra 380 390 400 410 420 430 440 450 460 wavelength [nm] 20 μ A 100 μ A normalized normalized normalized Ivo Polak, apr05 35
Spectra of UV-LED emission intensity [arb. units] 10 6 10 5 10 4 1000 100 10 UV LED diode no. 2 emission spectra in the absolute scale 400 450 500 550 600 wavelength [nm] 20 μa 100 μa 1 ma 3 ma 10 ma normalized emission intensity [arb. units] 1.2 1 0.8 0.6 0.4 0.2 0 UV LED diode emission spectra, 1 ma 380 390 400 410 420 430 440 450 460 wavelength [nm] no. 1 no. 2 no. 3 no. 4 emission maxima within 399.5-401 nm Ivo Polak, apr05 36
Spectra of GRN-LED Agilent HLMP-CM15 normalized emission intensity [arb. units] 1.2 1 0.8 0.6 0.4 0.2 0 Green LED diode emission spectra 450 500 550 600 650 700 wavelength [nm] 20 μa 100 μ A 1 ma 3 ma 10 ma Pic from Agilent datasheet Ivo Polak, apr05 37
Spectra of RED-LED Ivo Polak, Prague Apr05 normalized emission intensity [arb. units] 1.2 1 0.8 0.6 0.4 0.2 0 Red LED diode emission spectra 1 ma 10 ma 3 ma 20 ma 550 600 650 700 750 wavelength [nm] Not a nice spectrum it was chosen no-name LED 38
Wavelength emission of UV-LEDs Jaroslav Zalesak, Apr05 Peak and spread derived from Gaussian fit on LED emission frequency spectrum (-2,+1)*σ Peaks @ (400±1) nm for all 4 UV-LEDs and all light intensities Spread of emission (1 sigma from fit) in average: (4.8), 4.9, 5.1, 5.4, 6.6 nm for (0.02), 0.1, 1.0, 3.0 10.0 ma 39
LED intensity vs current (coarse measurement) Jaroslav Zalešák, apr05 UV-LEDs Green LED Intensity determined as height of gauss. fit on emission freq. spectrum light intensity at spectrum peak position) Assumption of same spectrum shape (peak position and width as well) in all currents, else needed integration in some freq. range (~20% variation 1-10 ma) Not fully comparable absolute light intensity (adjustment of measurement uncertainty, small variation in spectrum shape, few points) 40
Large LED Quality test in Prague Setup: Ivo s LED driver, tested UV-LEDs Light adapter (~2-3 mm air gap), fiber bundle (glued one end) APD mask holding fibers on 9 APDs (small gain ~10) + 1 PIN Preamp + Camac / oscilloscope readout Measurements: 1. Qualitative -- 9 chosen fibers from bundle with 9 APDs for each UV-LED, sub-sample 30 (50?) pieces a) Light cone homogeneity turning led around axis, 9 fibers in cross average values and spread for each position (same current) b) Variation light intensity measure at low-medium-high LED currents 2. Qualification test selection 500 good LEDs of all Using 4 channels oscilloscope measurement of signal shape satisfying some criteria (amplitude, width,..) derived from 1. 3. Ageing test General problems: How to determined good value of LED current in MIP units How to figure out what amplitude corresponds to 1 or 10 MIPs Completed by Jaroslav Zalesak, apr05 41
TEMPERATURE sensitivity One CH LEDdrv, SiPM, linearity Calibration mode (high gain), ADC is saturated Small effect on temperature, will be estimated Cool zoom Room zoom Ivo Polak CALICE oct05 42
PIN photodiode response to LEDpulse with fiber feedback 43
Stability parameters of CMB Cooled each one add - 300ps delay, heated some +300ps each Total delay is in range of -1ns to +1ns for estimated temp. range 200deg.C LEDdrv system is temperature pretty stable for our purposes. Measured by oscilloscope TDS3054 Temperature stability is very good, well bellow 1ns @ normal range (40degC) 1hr Jitter is in range 300ps Spread of the Delay in one plane (12LED) around 1ns 44