March 22, 2005

1
String-21 Flasher Analysis

• Jim Braun
• University of Wisconsin

2
How can we use flasher data?

• String-21 geometry reasonably well understood
• System timing
• Provides test DOM time resolution
• Measure precision
• Detect systematic problems
• Properties of hole ice
• Differences in waveforms before and after
  freeze-in may yield insight

3
Summary of available flasher data
4
Summary of available flasher data

• Pre-1270 runs
• Suffer from cable bandwidth overload
• Large gaps in data
• Use cautiously
• Runs 1270-1271
• No monolith output
• DOMs missing from upper string
• 15 deadtime
• 1.1ms offset between flasher timestamps and
  events

Flashing DOM
Receiving DOMs suppressed
5
SPS runs 1270 1271

• DOM 39 (2095m) chosen randomly as flasher
• Perhaps not a good choice from standpoint of
  system timing!

6
SPS run 1270 Analysis

• Use raw TestDAQ data
• Must apply trigger
• Require 4 hits within 6ms
• Events observed first by DOM 38 are flasher
  events
• Flasher rate - 610 Hz
• Data Format
• 32 bins ATWD ch0 data for receiving DOMs
• 64 bins ATWD ch3 data for flashing DOM
• By disabling partner of flashing DOM, cable
  bandwidth of flashing DOM approximately
  equivalent to bandwidth of receiving DOMs
• Event rate is 515Hz

7
Run 1270 waveforms
8
SPS run 1270

• Flasher signal nearly factor of 100 stronger than
  background
• Most DOMs at top of string not present
• Isolated Q2 glitch
• DOM 40 disabled to maximize cable bandwidth for
  DOM 39
• Dust attenuation is apparent
• .1 SPE reached at DOMs 33 and 56
• Light directed upward is suppressed

DOM 40 disabled
9
SPS run 1270

• Large number of low multiplicity IceTop events
• Typical flasher event seen by 15 DOMs
• Vertical pulse propagation of 120m
• Less deadtime shifts distribution toward higher
  multiplicity

Flasher Events
IceTop noise
10
SPS run 1270

• Time between flash and observation increases
  somewhat linearly for DOMs away from flasher
• Subtle convex nonlinearity near flashing DOM

Nonlinearity near flasher
11
System Timing
A simple test
DOMi-2

• Enable flasher board on DOMi
• Calculate pulse arrival time differences between
  the two DOMs directly above with good statistics
  (105 events)
• Requires time-calibrated raw timestamp and
  discriminator crossing analysis of ATWD data for
  precision
• Width of distribution is characteristic of timing
  precision
• Mean of distribution is characteristic of optical
  distance between DOMs

DOMi-1
DOMi
12
System Timing

• Factors affecting width
• PMT response
• ATWD bin resolution (3.5ns/bin)
• RapCal precision
• Factors affecting mean
• DOM spacing
• RapCal systematic accuracy
• Ice properties

13
SPS run 1270
17m
17m

• RMS Dt 6.3ns
• 37 38
• Mean Dt 87ns
• equivalent to 20m path length difference in ice
• Why not 17m?
• Probably an effect of dust layer
• Need to understand this

14
Where do we go from here?

• More data!
• A run similar to 1270 with clear ice above
  flashing DOM should be taken
• Need ability to adjust flasher rate
• Need to understand flasher timestamp offset
• Pre/post freeze-in waveform analysis
• Understand hole ice
• Pre-1270 waveforms are still OK
• Photonics simulation?
• Need to understand propagation of flasher pulse
  in dust layer