Science Run 2. (first half

1
Science Run 2. (first half)Detector Timing
Laser Interferometer Gravitational-Wave
Observatory (LIGO)

• LSC 2003 Spring Meeting
• LIGO Livingston Observatory
• Szabolcs Márka, Daniel Sigg

2
Real-time Timing

• Initialization
• GPS 1 PPS
• Diagnostics
• GPS Ramp gt absolute
• CPU meter overload gt lost IO cycles
• Polling ADC ready bitgt missed ADC samples
• Resync countergt missed computing cycles
• Cycle countgt inter-crate synchronization

• Processing Time _at_ 16kHz
• LSC 46ms, LVEA LOS 56ms,ETM 45ms, MC2 27ms

3
Real-time Timing (2)
4
How can we measure ?

• Special signal to measure DAQ timing
• GPS TRIG
• 1 ms long square wave
• GPS RAMP
• 1 ms long steep ramp
• The GPS second tic is aligned with
• the rising edge of the square wave
• the zero crossing of the ramp
• Fit the RAMP and compute the zero crossing
• Determine delay between GPS tic and DAQ time
  stamp
• Practical experience O(100ns) measurements are
  achievable!
• Similar procedure for the IRIG-B signal
• we extract the date
• DMT monitors do the job

5
Max Mean Min
Timing performance of LIGO LSC systems
LHO 4K LSC
2us
2us
!!!
LHO 2K LSC
1.7us
1us
!!!
!!!
6.5us
2us
5us
3.5us
LLO LSC
The first 31 days of S2
6
Max Mean Min
Timing performance of LHO 4K LSC systems
LHO 4K LSC
2us
2us
The first 31 days of S2
7
Max Mean Min
Timing performance of LHO 2K LSC systems
LHO 2K LSC
!!!
1.7us
1us
The first 31 days of S2
8
Max Mean Min
Timing performance of LLO 4K LSC systems
!!!
!!!
6.5us
5us
3.5us
2us
LLO LSC
The first 31 days of S2
9
Timing performance of LHO EX system
Max Mean Min
The first 31 days of S2
28ms
10
Timing performance of LHO EY system
Max Mean Min
The first 31 days of S2
51.7us
1us
11
Timing performance of LLO EY system
Max Mean Min
4us
19.3us
The first 31 days of S2
12
Summary

• We have not seen major problems with LSC timing
• Small jumps of 5us are observed
• There are problems with LHO EX
• Timing problems are also good indicators of
  channel hopping