In-Flight ACIS Charge Transfer Inefficiency and Gain Corrections

Valid for -120 C data
Last updated June 19, 2000
Previous releases:
Ftools script (by L. Van Speybroeck) and coefficients for correcting CTI and converting PHA to PI
(16896 byte tar file)
Plots of peak centroids vs row number for corrected data:

I0 I1
I2 I3
S0 S1 S2 S3 S4 S5
CTI correction for centroids of peaks: [Explanation]

 cti = a1 + (b1*pha0) + (c1*(pha0^2.))

 eps = a2 + (b2*pha0) + (c2*(pha0^2.))

 pha = (1. + (cti*rows) + (eps*(rows^2)) ) * pha0
Energy scale correction for spectrum: 
 E = (x0 + pha) / x

List of OBSIDs used for determining CTI and gain corrections
ASCII version of this table
Coefficients:
Chip Node a1 b1c1 a2 b2c2 x0 x
0 0 0.0002808 -1.691e-07 3.815e-11 5.026e-08 -3.013e-10 1.686e-13 0.4 260.3
1 0.0003726 -3.046e-07 9.739e-11 -8.296e-08 7.976e-11 -2.610e-14 0.4 252.2
2 0.0003525 -2.728e-07 8.450e-11 -5.256e-08 2.796e-11 -4.326e-15 0.5 260.6
3 0.0003110 -2.496e-07 7.971e-11 -5.367e-08 5.567e-11 -1.940e-14 0.5 253.7
1 0 0.0003139 -2.374e-07 7.158e-11 -4.888e-08 3.134e-11 -5.587e-15 0.5 252.7
1 0.0003688 -2.987e-07 9.548e-11 -8.298e-08 7.619e-11 -2.415e-14 0.3 248.4
2 0.0003678 -2.940e-07 9.254e-11 -8.407e-08 7.327e-11 -2.213e-14 0.5 250.3
3 0.0003172 -2.539e-07 8.067e-11 -7.179e-08 6.953e-11 -2.293e-14 0.4 253.2
2 0 0.0003823 -2.938e-07 9.046e-11 -6.512e-08 5.802e-11 -1.846e-14 0.6 260.7
1 0.0004463 -3.570e-07 1.108e-10 -8.562e-08 8.596e-11 -2.835e-14 0.5 263.8
2 0.0004278 -3.203e-07 9.458e-11 -7.254e-08 5.916e-11 -1.656e-14 0.7 264.8
3 0.0003733 -2.946e-07 9.297e-11 -6.524e-08 6.451e-11 -2.229e-14 0.6 260.0
3 0 0.0003995 -3.263e-07 1.038e-10 -8.941e-08 9.170e-11 -3.069e-14 0.5 254.4
1 0.0004311 -3.487e-07 1.129e-10 -7.906e-08 7.985e-11 -2.913e-14 0.5 253.2
2 0.0004526 -3.825e-07 1.275e-10 -8.862e-08 9.817e-11 -3.803e-14 0.3 249.4
3 0.0003821 -2.967e-07 9.238e-11 -5.982e-08 4.970e-11 -1.495e-14 0.6 255.2
4 0 0.0005426 -3.750e-07 1.029e-10 -8.945e-08 4.056e-11 -1.943e-15 0.5 269.9
1 0.0006182 -4.565e-07 1.357e-10 -1.033e-07 7.803e-11 -2.327e-14 0.7 265.6
2 0.0006371 -4.732e-07 1.408e-10 -1.091e-07 7.721e-11 -1.991e-14 0.7 256.2
3 0.0006362 -4.708e-07 1.384e-10 -1.208e-07 7.704e-11 -1.622e-14 0.7 255.1
5 0 6.912e-05 -9.210e-08 3.141e-11 9.390e-09 5.515e-12 -1.667e-15 0.2 233.4
1 3.082e-05 4.151e-08 -7.801e-11 5.876e-08 -1.659e-10 1.296e-13 0.7 214.8
2 5.852e-05 -9.880e-08 3.971e-11 2.002e-08 -4.465e-13 -9.606e-16 0.7 220.2
3 5.953e-05 -9.760e-08 4.011e-11 1.495e-08 1.210e-11 -8.288e-15 0.4 226.5
6 0 0.0004550 -3.578e-07 1.143e-10 -9.057e-08 8.505e-11 -2.831e-14 0.5 251.9
1 0.0005110 -3.916e-07 1.213e-10 -1.143e-07 1.048e-10 -3.460e-14 0.6 260.6
2 0.0005218 -4.117e-07 1.315e-10 -1.181e-07 1.046e-10 -3.445e-14 0.4 253.8
3 0.0004604 -3.437e-07 1.058e-10 -8.774e-08 6.973e-11 -2.156e-14 0.6 260.6
7 0 2.565e-05 -2.489e-08 8.933e-12 3.054e-08 -3.718e-11 1.277e-14 0.2 222.3
1 9.839e-06 -7.465e-09 3.136e-12 4.286e-08 -5.259e-11 1.866e-14 0.2 211.7
2 1.456e-05 -1.155e-08 3.436e-12 4.489e-08 -5.769e-11 2.157e-14 0.2 214.5
3 2.875e-05 -2.047e-08 4.313e-12 3.483e-08 -5.148e-11 2.150e-14 0.1 222.6
8 0 0.0003777 -3.177e-07 1.087e-10 -9.373e-08 9.469e-11 -3.368e-14 0.4 231.0
1 0.0004279 -3.731e-07 1.333e-10 -1.117e-07 1.122e-10 -4.278e-14 0.6 226.2
2 0.0004292 -3.691e-07 1.293e-10 -1.146e-07 1.090e-10 -3.832e-14 0.3 227.7
3 0.0003689 -3.059e-07 1.030e-10 -9.157e-08 9.514e-11 -3.396e-14 0.5 237.0

This entire procedure should be done separately for each node of each chip. Software used for this procedure is described here.

  1. Sum up several different data sets to get good signal-to noise. [Sample spectrum]

  2. Calculate the CTI for this summed data. We did this for blocks of 100 rows, across the chip. [Sample plots]

  3. Plot the coefficients for each line vs. the line centroids for the 100-row section closest to the node. Fit 3rd-degree polynomials to this data. These polynomials are the corrections given above in "CTI correction for centroids of peaks". [Sample plots]

  4. Plot line centroids vs. true line energies for the 100-row section closest to the node and fit a line. The coefficients of this fit are the energy scale correction coefficients. [Sample plots]

  5. apply these corrections to each individual data set. [Sample plots]

Back to the ACIS Calibration Page