Fast Timing Workshop June 8 10 th 2015

  • Slides: 21
Download presentation
Fast Timing Workshop June 8 -10 th 2015 FZU Prague Timing Methods with Fast

Fast Timing Workshop June 8 -10 th 2015 FZU Prague Timing Methods with Fast Integrated Technologies Jean-François Genat 1

Fast Timing/Imaging Photo-detectors Multi-anodes PMTs Dynodes QE CE Rise-time TTS (1 PE) Pixel size

Fast Timing/Imaging Photo-detectors Multi-anodes PMTs Dynodes QE CE Rise-time TTS (1 PE) Pixel size Dark counts Dead time Magnetic field Radiation hardness Total Charge Silicon-PMTs Quenched Geiger 30% 90% 0. 5 -1 ns 150 ps 2 x 2 mm 2 1 -10 Hz 5 ns no 1 k. Rad (PC) 90% 70% 250 ps 100 ps 50 x 50 mm 2 1 -10 MHz/cm 2 100 -500 ns yes noise x 10 Micro Channel Plates Micro-Pores 20 -30% 60% 50 -200 ps 30 -50 ps 1. 5 x 1. 5 mm 2 1 -10 k. Hz/cm 2 1 ms 15 k. G 1 k. Rad (PC) 0. 5 -2 C/year Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 2

Example: GHz Bandwidth Micro-Channel Plate Signals - Detector + electronics noise >> quantization noise

Example: GHz Bandwidth Micro-Channel Plate Signals - Detector + electronics noise >> quantization noise (LSB/√ 12) - Sampling frequency > 2 x Shannon-Nyquist=4 GHz Micro-Channel Plate Signal Spectrum Noise as small as possible 2 2 GHz 4 GHz 3 Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 3

Contributions to timing spreads: Timing spreads Detector: Transit Time Noisedetector Rise time Gain (Signal/noise)

Contributions to timing spreads: Timing spreads Detector: Transit Time Noisedetector Rise time Gain (Signal/noise) Electronics: Noiseelec Sample rate Analog bandwidth Stefan Ritt: sn = sn det + sn elec S = G Npe Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 4

Electronics: Fast Timing Techniques • Threshold techniques • • Single Threshold Double Threshold Multiple

Electronics: Fast Timing Techniques • Threshold techniques • • Single Threshold Double Threshold Multiple Thresholds Constant Fraction • Waveform sampling techniques • Waveform sampling using Flash ADCs (up to 8 -9 bit) or • Analog memories fast sampling + AD conversions with slow ADCs -bit) (12 and Digital Signal Processing Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 5

Fast Timing techniques Constant-fraction Multi-threshold Constant fraction Leading edge errors Extrapolated time Waveform sampling

Fast Timing techniques Constant-fraction Multi-threshold Constant fraction Leading edge errors Extrapolated time Waveform sampling Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 6

Electronics: Timing Methods: Single threshold Multiple thresholds Constant fraction Waveform sampling Time resolution vs

Electronics: Timing Methods: Single threshold Multiple thresholds Constant fraction Waveform sampling Time resolution vs Number of photo-electrons zoom Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 7

Time Stretching Différential: Ramp goes up at rate Time is stretched by , down

Time Stretching Différential: Ramp goes up at rate Time is stretched by , down at rate , measured using a “slow” TDC d d Resolution: ps ? To Counter Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 8

Single Threshold Effects of Amplitude and Rise-time spreads Amplitude and/or Rise-time spectra translate into

Single Threshold Effects of Amplitude and Rise-time spreads Amplitude and/or Rise-time spectra translate into time spreads in case of single threshold Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 9

Effect of Noise using Single Threshold Methods Time spread proportional to rise-time Jean-François Genat,

Effect of Noise using Single Threshold Methods Time spread proportional to rise-time Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 10

Discriminator walk effects Walk: Discriminator delay depends on the pulse slope across threshold Pulse

Discriminator walk effects Walk: Discriminator delay depends on the pulse slope across threshold Pulse slope: detector rise-time + amplifier Use appropriate gain x bandwidth technology to match the detector speed Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 11

Multi-threshold (simulation) Multi-threshold: sampling times instead of amplitudes : - Number of thresholds -

Multi-threshold (simulation) Multi-threshold: sampling times instead of amplitudes : - Number of thresholds - Thresholds values - Order of the fit: 4 -8 equally spaced 2 d order optimum Extrapolated time Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 12

Multi-threshold performance Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague

Multi-threshold performance Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 13

Waveform Sampling and Analysis Fit to waveform and derivative templates B. Cleland E. Stern

Waveform Sampling and Analysis Fit to waveform and derivative templates B. Cleland E. Stern Fast Timing Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 14

Timing Resolution using Waveform Sampling abw= analog bandwidth = rise time (10 -90%) =

Timing Resolution using Waveform Sampling abw= analog bandwidth = rise time (10 -90%) = sampling period S Since = signal amplitude with and (Stefan Ritt, PSI) 15 Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 15

An Application of Waveform Sampling Two-dimension readout using delay lines Waveform sampling and analysis:

An Application of Waveform Sampling Two-dimension readout using delay lines Waveform sampling and analysis: • • • Picosecond timing with fast detectors Charge: centroids for 2 D readout Resolve double pulse Large area detectors read with delay lines in series Jean-François Jean-Francois Genat, March Fast 2009, Timing Argonne. Workshop, June 8 -10 th 2015, FZU Prague 19 th 16

Waveform Sampling and Analysis Two-dimension readout using delay lines B. Cleland E. Stern, BNL

Waveform Sampling and Analysis Two-dimension readout using delay lines B. Cleland E. Stern, BNL Signal Template Real MCP Laser data Extract precise time and amplitude from minimization of c 2 evaluated wrt a waveform template deduced from the measurements. Jean-François Jean-Francois Genat, March Fast 2009, Timing Argonne. Workshop, June 8 -10 th 2015, FZU Prague 19 th 17

Waveform Sampling Parameters - Sampling frequency: Set at twice the largest frequency in the

Waveform Sampling Parameters - Sampling frequency: Set at twice the largest frequency in the signal spectrum f 0 - Digitization accuracy: 2 f 0 Evaluate what is needed from signals properties: Example (simulation) Micro-Channel Plate signals Jean-François Jean-Francois Genat, March Fast 2009, Timing Argonne. Workshop, June 8 -10 th 2015, FZU Prague 19 th 18

Waveform sampling: Measurements vs Simulation 50 Photo-electrons 18 Photoelectrons Jean-François Jean-Francois Genat, March rms=3.

Waveform sampling: Measurements vs Simulation 50 Photo-electrons 18 Photoelectrons Jean-François Jean-Francois Genat, March rms=3. 82 ps vs 2. 5 ps (simulation) rms = 6. 05 ps vs 7 ps (simulation) Fast 2009, Timing Argonne. Workshop, June 8 -10 th 2015, FZU Prague 19 th 19

Waveform sampling: Switched Capacitors Arrays Sampling The input signal is sampled at the elementary

Waveform sampling: Switched Capacitors Arrays Sampling The input signal is sampled at the elementary delay period (ns-ps) Readout ADC can be very accurate (10 -14 bit), at the expense of the conversion time, external device Jean-François Genat, Fast Timing Workshop, June 8 -10 th 2015, FZU Prague 20

Today, 10 ps timing is available integrated 1 ps under work, looks promising from

Today, 10 ps timing is available integrated 1 ps under work, looks promising from the 10 -100 GHz range VLSI technologies Hear during the Workshop …? The End Jean-François Genat, Fast Timing Workshop, June 8 -10 th March 8 th 2007, 2015, DAPNIA, FZU Prague Saclay 21