Strobe delay scans in STcontrol Jrn Grosse-Knetter Intro: strobe - - PowerPoint PPT Presentation

Jörn Grosse-Knetter

Strobe delay scans in STcontrol

27 June 2011 strobe delay scans 2

Intro: strobe delay scans (1)

 See talk from Joseph Hu last week

− Timing and time walk behaviour of FE

 A lot of code from ROD was there, but based

• n MCC (similar to TurboDAQ scans):

− T0_SCAN:

 loop 0 scans strobe delay, loop 1 scans trigger delay,

• nly one LVL1 is used, PlsrDAC=1000 (fixed)

 End-of-loop-0 action: fit strobe delay → time threshold  End-of-loop-1 action:

− Find trigger delay with most working pixels → store for later use − Neighbouring trg. delay: use for strobe delay calibration → stored

in FE config

− Shift strobe delay for best trg. delay by 5ns → store for later use

27 June 2011 strobe delay scans 3

Intro: strobe delay scans (2)

 Existing code cont'd:

− TIMEWALK_SCAN:

 Run with trigger delay from T0 scan, read only one LVL1  loop 0 scans strobe delay, loop 1 scans PlsrDAC  End-of-loop-0 action: fit strobe delay → time threshold  End-of-loop-1 action: none (to be time walk analysis?)

− INTIME_THRESH_SCAN:

 Run with trigger delay and strobe delay from T0 scan,

read only one LVL1

 Otherwise normal threshold scan with S-curve fit

− Re-activated all 3 scans and adapted to FE-I4

27 June 2011 strobe delay scans 4

Strobe delay scan & fit (1)

 Scan occupancy vs

pulser delay DAC

 Online S-curve fitting

(almost) fine:

− Ignoring points below

100% eff. after plateau

− Needed adaption of

code to not use PlsrDAC calibration → new end-of-loop action DELAY_FIT

 Result in TIMEWALK

histo.: calibrated to ns

27 June 2011 strobe delay scans 5

Strobe delay scan & fit (2)

 Only available offline

(ModuleAnalysis) since slow:

− Box fit with Minuit − Recommended

settings, seems to fit reliably

− Added new

FIT_PAR0...5 histograms for general fit output

− Fit status displayed

and abort possible

adjust to your scan results / settings

27 June 2011 strobe delay scans 6

Strobe delay scan & fit (3)

 Box-fit parameters (in

delay DAC's not ns!):

− Centre = central pos. − Width = lower 50% to

upper 50% point

− Sigmas: erf.-smearing

25.6 ns Width of box fit

27 June 2011 strobe delay scans 7

DELAY_FIT vs box fit

DELAY_FIT: trise in ns Mean: trise = 15.37 ns Box fit tcentre in delay-DAC units Mean = 28.1ns → trise=28.1ns-25.6/2ns=15.3 ns ✓

27 June 2011 strobe delay scans 8

Result T0 scan

Trigger delay = 56 Trigger delay = 57 Default: 1 of 6 mask steps

27 June 2011 strobe delay scans 9

T0 scan end-of-loop action

 After complete T0_SCAN:

new end-of-loop action T0_SET:

 Set trigger delay in group  Set PlsrDelay and delay

• calib. (in ns/DAC-step)

27 June 2011 strobe delay scans 10

Timewalk scan

 Can plot (calibrated) strobe

delay vs PlsrDAC

 Fit doesn't seem to work for

every scan point → check

 Check PlsrDAC problems at

high charge (→ ToT mode?)

 So far no analysis

module avg. single pixel single pixel

27 June 2011 strobe delay scans 11

In-time threshold

• verdrive (e)

(in-time-thresh. – discrim. thresh.) in-time threshold (e) (tuned to 3000e discr. threshold)

27 June 2011 strobe delay scans 12

Summary

 Adapted all strobe delay scans from MCC to

FE-I4, re-activated all timing scans

 Added more fitting functionality in MA  All seem to work reasonably well, but:

− Problems in reliability of online fitting, may have to

be looked at

− Need analysis for time walk scan

 Code for TurboDAQ data exists in MA, adapt?

 Code is in svn (not in release 1.3, trunk only)