Caveat Emptor:

While we freely admit that the fudged matrix being released is not the best solution to the problem of the residual systematics in spectral fits, it is the only possibility available to us at this time. The present release is intended only as an intermediate update until more calibration data have been collected and analyzed. We have certainly noted the overwhelming desire for a ``better'' calibration of the XRT/PSPC, we ourselves certainly share this concern.

Channel Matrix Channel Matrix

92mar11 93jan12 92mar11 93jan12

10 0.245 0.306 100 -0.000 0.000

12 0.164 0.187 110 -0.011 -0.013

14 0.107 0.103 120 -0.000 -0.043

16 0.071 0.049 130 0.007 0.004

18 0.052 0.019 140 0.003 0.005

20 0.043 0.006 150 -0.002 0.003

25 0.038 0.014 160 -0.002 0.004

30 -0.002 0.007 170 0.002 0.006

35 -0.078 -0.034 180 0.011 0.014

40 -0.111 -0.045 190 0.033 0.034

45 -0.071 -0.016 200 0.060 0.059

50 -0.014 0.008 210 0.089 0.088

55 -0.002 -0.005 220 0.123 0.120

60 0.018 0.003 230 0.161 0.155

65 0.033 0.017 240 0.203 0.192

70 0.028 0.013 250 0.246 0.228

75 0.019 0.007 260 0.288 0.261

80 0.009 0.001 270 0.328 0.289

85 0.002 -0.002 280 0.362 0.311

90 0.002 0.000 290 0.390 0.323

95 0.003 0.003 300 0.410 0.324

Using a broken power law with fixed to cm

final fit: Slope_1: 2.06, Break 0.77keV, Slope_2: 2.61

Table 3.4: Spline fit to the fractional residuals of Mkn 421

**Table 3.4:** Spline fit to the fractional residuals of Mkn 421
Channel	Matrix	Channel	Matrix
	92mar11	93jan12		92mar11	93jan12
10	0.245	0.306	100	-0.000	0.000
12	0.164	0.187	110	-0.011	-0.013
14	0.107	0.103	120	-0.000	-0.043
16	0.071	0.049	130	0.007	0.004
18	0.052	0.019	140	0.003	0.005
20	0.043	0.006	150	-0.002	0.003
25	0.038	0.014	160	-0.002	0.004
30	-0.002	0.007	170	0.002	0.006
35	-0.078	-0.034	180	0.011	0.014
40	-0.111	-0.045	190	0.033	0.034
45	-0.071	-0.016	200	0.060	0.059
50	-0.014	0.008	210	0.089	0.088
55	-0.002	-0.005	220	0.123	0.120
60	0.018	0.003	230	0.161	0.155
65	0.033	0.017	240	0.203	0.192
70	0.028	0.013	250	0.246	0.228
75	0.019	0.007	260	0.288	0.261
80	0.009	0.001	270	0.328	0.289
85	0.002	-0.002	280	0.362	0.311
90	0.002	0.000	290	0.390	0.323
95	0.003	0.003	300	0.410	0.324
Using a broken power law with fixed to cm
final fit: Slope_1: 2.06, Break 0.77keV, Slope_2: 2.61

Source Name Type Courtesy Type of Fit, Remarks

Mkn 421 BL Lac Thomas BPL, gal.NH, used to fudge DRM_31

PKS 2155-304 BL Lac Brinkmann,Comastri BPL, gal.NH, good fit, no features

H1426+42 BL Lac George BPL, gal.NH, good fit, no features

Sum of RMSS QSOs Hasinger, Comastri PL, no features >0.2 (NH ambiguous)

Mkn 766 Sy1 Molendi, Comastri PL+BB, no features, gal. NH

Chi Per Be-HMXB Haberl PL, highly absorbed, good fit

? OB bin Haberl RS+RS, good fit

HZ 43 WD Fleming WD-model, 1 parameter, good fit

Crab SNR PL, 1 parameter, 5% lower gain

Table 3.5: Tests of the 93jan12 DRM on a variety of different sources

Source Name	Type	Courtesy	Type of Fit, Remarks
Mkn 421	BL Lac	Thomas	BPL, gal.NH, used to fudge DRM_31
PKS 2155-304	BL Lac	Brinkmann,Comastri	BPL, gal.NH, good fit, no features
H1426+42	BL Lac	George	BPL, gal.NH, good fit, no features
Sum of RMSS	QSOs	Hasinger, Comastri	PL, no features >0.2 (NH ambiguous)
Mkn 766	Sy1	Molendi, Comastri	PL+BB, no features, gal. NH
Chi Per	Be-HMXB	Haberl	PL, highly absorbed, good fit
?	OB bin	Haberl	RS+RS, good fit
HZ 43	WD	Fleming	WD-model, 1 parameter, good fit
Crab	SNR		PL, 1 parameter, 5% lower gain