IPM update A Jansson Recap At last IPM

IPM update A. Jansson

Recap • At last IPM status report, has just found and fixed some problems with the HV turn-on part of the Lab. View software. • Although this improved things, it did not fix the problem with the fluctuating signals. 12/3/2009 Tev meeting A. Jansson 2

Pbar jacking data 12/3/2009 Tev meeting A. Jansson 3

Code verification • Wrote Mathematica code to read binary raw data files and re -analyze independently. • Found no big differences in results • But, found something else… 12/3/2009 Data from ramp, flattop and squeeze Tev meeting A. Jansson 4

QIE encoding Encoded value [x 2. 6 f. C] 240 Histogram of raw data 190 140 90 40 -10 0 10 20 30 40 7 -bit QIE encoding 50 60 QIE 8 chip is a pseudo-logarithmic charge integrating ADC (small LSB, yet large dynamic range with few bits). Some values (e. g. 17, 19, 21…) should not appear in data from QIE, if the data has been converted into linear units. Nevertheless, they were observed… 12/3/2009 Tev meeting A. Jansson 5

QIE conversion fix • Turns out, the Lab. View initialization code did not set the control register bit that enables QIE number conversion in the buffer board. • This explains why the sigma varied with signal intensity (essentially a kind of saturation). • Fixed! 12/3/2009 Tev meeting A. Jansson 6

Data after QIE fix 12/3/2009 Tev meeting A. Jansson 7

Source of signal fluctuations? • Signal level still fluctuating! • Had excluded HV (long time ago) by adding delay in Lab. View code to ensure full voltage obtained – Observed no improvement with additional delay • Had excluded timing by observing clock stability w. r. t beam signal. – Also recently checked stability of timing distribution in the tunnel. • Vacuum pressure only remaining possibility, but – Saw no fluctuation in ion gauges – No similar effect observed elsewhere (e. g. MI signals very reproducible. 12/3/2009 Tev meeting A. Jansson 8

HV re-check • Decided to redo HV delay test, but in a different way. – Instead of adding delay in Lab. View, fix delay between pre-trigger and trigger event externally. – Using dedicated IPM trigger event (TVBS $DB), issued n seconds after the pre-trigger event (e. g. $2 A) 12/3/2009 Tev meeting A. Jansson 9

Signal vs delay ~750 ms 200 0, 7 180 0, 6 160 0, 4 Time-outs Signal [arb. units] 120 100 80 60 0, 3 Sgima [mm] 0, 5 140 Signal Sigma 0, 2 40 0, 1 20 0 0 12/3/2009 0, 5 1 1, 5 2 Delay [s] 2, 5 Tev meeting 3 3, 5 4 4, 5 A. Jansson 0 10

Result with 750 ms extra delay 12/3/2009 Tev meeting A. Jansson 11

Why did it take so long to figure out? • HV was the obvious candidate. • The previous test, adding extra delay to allow HV to come up, was affected by the bug in the Lab. View HV turn-on code which was recently found and fixed • As a result, the HV was exonerated incorrectly, moving the focus to other potential sources of the effect. 12/3/2009 Tev meeting A. Jansson 12

Data then and now 1 Month ago: After QIE and HV fix: Noisy pbar data on ramp due to bad timing (final cog after ramp). 12/3/2009 Tev meeting A. Jansson 13

Timing before ramp Set up a cyclic sequence of SPECs with different clock delay, to run before ramp Also ran during HEP Found “optimal” settings to within ~10 ns, further fine tuning needed. 12/3/2009 Tev meeting A. Jansson 14

Last shot 12/3/2009 Tev meeting A. Jansson 15

Conclusions • Understood the cause of fluctuating signals • Understood cause of sigma/signals correlation. • This also improved the horizontal signal level – Seems horizontal HV (even) slower to come up. – Still lower signal than vertical system. Consider changing MCP to improve gain. – Horizontal system has more hardware errors than vertical, may need to change some hardware. • Now that we have stable signals – Fine-tune SPECs for different Tev states, in particular timing settings – Redo studies (position scans, timing scans etc) to characterize detector – Fix horizontal system 12/3/2009 Tev meeting A. Jansson 16