Cross-Calibration of FDs Sean R. Stratton TA-All Meeting – June 2010.

Cross-Calibration of FDsSean R. StrattonTA-All Meeting – June 2010

MotivationWe’re still trying to understand

the relative energy scale differences between our FDs.

The UVLED was used to cross-calibrate all three FD sites.

Although RXF tests have been done in the past, we needed a longer, more thorough test to be done.

BRM/LR Calibration Method

Get absolute calibration for selected tubes.◦ CRAYS.

Gain-match remaining PMTs by adjusting individual HV.

Correct tube gains in each camera according to:◦ Camera temperature.◦ HV supply.◦ Relative gain from XF data.

Gains relevant for 1 hour periods.

BRM/LR Calibration Method

Get absolute calibration for selected tubes.◦ CRAYS.

Gain-match remaining PMTs by adjusting individual HV.

Correct tube gains in each camera according to:◦ Camera temperature.◦ HV supply.◦ Relative gain of center PMT (0x77).

Gains relevant for 1 hour periods.

Hardware Level

BRM/LR Calibration Method

Get absolute calibration for selected tubes.◦ CRAYS.

Gain-match remaining PMTs by adjusting individual HV.

Correct tube gains in each camera according to:◦ Camera temperature.◦ HV supply.◦ Relative gain of center PMT (0x77).

Gains relevant for 1 hour periods.

Software Level

The Calibration Constants

G0

Absolute Calibration Constant – Only applies to specific subset of tubes (channels 0x33 & 0x77 of every camera and 0xB3 in cameras 0&6). Measured with CRAYS, has units of FADC/photon.

G1

XF Non-Uniformity Correction – The HV levels in each camera are adjusted to get a flat response across all channels using the XF. This factor accounts for the non-uniformity of the Xenon source.

G2

Xenon Flasher Correction – Tube gains are corrected using run-time Xenon flasher data. Mirror splash-back effects are accounted for by using data taken during HV adjustment.

G3YAP Correction – YAP source is used to monitor PMT aging effect. This value is currently unity (?).

G4

Temperature Correction – Each camera’s temperature is monitored, and gains are corrected using a typical dependence of -0.72%/°C.

43210total GGGGGG

Case Study: BRM 20090225Clear CLF run nightLots of data (~10 hr.)

BRM G0 as of Feb. 08

BRM Gtotal for 20090225

G0 = 0.5087T0 = 24.55

dTG

dG0072.0

0045.0dV

dG

Camera 08

Camera 07

When/Why was HV Changed?

Camera 10

Camera 05

Camera 07

Camera 10

4 5

4 5

6 7

8 6 7

10 11

10 11

25 Feb 200922 Sep 2008

Camera<FADC>

08092205<FADC>

09022505Temp.

CorrectionCorrected <FADC>

% Change

0 37.5 41.9 1.058104 39.6 5.60%

1 47.5 48.2 1.053568 45.7 -3.69%

2 49.3 51.1 1.059616 48.2 -2.18%

3 43.4 45.1 1.057384 42.7 -1.72%

4 50.1 53.2 1.060912 50.1 0.09%

5 86.2 90.7 1.057024 85.8 -0.46%

6 47.0 59.7 1.062568 56.2 19.54%

7 71.2 91.1 1.058608 86.1 20.87%

8 54.3 57.3 1.06444 53.8 -0.86%

9 35.3 37.8 1.057888 35.7 1.22%

10 83.0 89.5 1.06444 84.1 1.30%

11 48.6 53.8 1.061848 50.7 4.25%

Camera 05

Question 1

Are the built-in XF used as an absolute light source or is it used to check uniformity and relative gains? In other words, do we know or care exactly how many photons are produced by each XF per flash?

Question 2

It is our understanding that the HV supplied to each tube is adjusted to match the output signal of the absolutely calibrated tubes (0x77, 0x33, & sometimes 0xB3), using covered mirrors and the built-in Xenon flasher apparatus. How is this done? In software? Low-level analysis?

Question 3

If each tube’s HV is adjusted to gain-match the absolutely calibrated PMT’s, then why would we need to adjust the HV of an absolutely calibrated PMT (from CRAYS)?