Logged in as guest
Viewing EPICpn/2016 Full headers
Private message: yes no
Notes: Notification:
Date: Wed, 30 May 2001 09:26:51 GMT From: rs1@mssl.ucl.ac.uk To: xmmhelp@xvsoc01.vilspa.esa.es Subject: noise in PN power spectra
Full_Name: Roberto Soria Submission from: (NULL) (128.40.72.254) Greetings again. I am working on EPIC PN Timing mode of a bright source. I extracted a power spectrum with standard packages, but the result doesn't look good: the Fourier spectrum is dominated by spurious white noise > 10 Hz and strong spikes at ~ 24 and 48 Hz. [I am sending a couple of examples as Postscript attachments directly to xmmhelp@xvsoc01.vilspa.esa.es in a separate email, please have a look at them]. Having ascertained the noise is not from the source, it appears that it is caused by the CCD readout. I guess I could filter it out by using a dark or flat field exposure in Timing mode, but I don't know where to find them and how to do it properly. Can you comment and offer suggestions please? Thanks, Roberto Soria MSSL
Date: Wed, 30 May 2001 10:31:09 +0100 From: Roberto Soria <rs1@mssl.ucl.ac.uk> To: xmmhelp@xvsoc01.vilspa.esa.es Subject: Re: noise in PN power spectra (PR#2016)
This is a multi-part message in MIME format. --------------AC0C17FB68F04176DF126034 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit This is an example of power spectrum (EPIC PN, Timing Mode, rev 176) dominated by spurious noise (readout noise?). Roberto --------------AC0C17FB68F04176DF126034 Content-Type: application/postscript; name="plot1.ps" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="plot1.ps" %!PS-Adobe-3.0 EPSF-3.0 %%For: rs1 %%Title: PGPLOT PostScript plot %%Creator: PGPLOT %%CreationDate: 24-May-2001 18:38 %%BoundingBox: (atend) %%DocumentFonts: (atend) %%LanguageLevel: 1 %%Orientation: Landscape %%Pages: (atend) %%EndComments %%BeginProlog /L {moveto rlineto currentpoint stroke moveto} bind def /C {rlineto currentpoint stroke moveto} bind def /D {moveto 0 0 rlineto currentpoint stroke moveto} bind def /SLW {5 mul setlinewidth} bind def /SCF /pop load def /BP {newpath moveto} bind def /LP /rlineto load def /EP {rlineto closepath eofill} bind def /MB {gsave translate MFAC dup scale 1 setlinewidth 2 setlinecap 0 setlinejoin newpath} bind def /ME /grestore load def /CC {0 360 arc stroke} bind def /FC {0 360 arc fill} bind def %%EndProlog %%Page: 1 1 %%BeginPageSetup /PGPLOT save def 0.072 0.072 scale 8149 250 translate 90 rotate 1 setlinejoin 1 setlinecap 1 SLW 1 SCF %%EndPageSetup %%PageBoundingBox: (atend) 0.000 setgray 1 SLW 8399 0 1050 1365 L 0 5849 C -8399 0 C 0 -5849 C 0 45 1289 1365 L 0 45 1885 1365 L 0 45 2309 1365 L 0 45 2637 1365 L 0 45 2905 1365 L 0 45 3132 1365 L 0 45 3329 1365 L 0 45 3502 1365 L 0 90 3657 1365 L 0 45 4677 1365 L 0 45 5273 1365 L 0 45 5697 1365 L 0 45 6025 1365 L 0 45 6293 1365 L 0 45 6520 1365 L 0 45 6716 1365 L 0 45 6890 1365 L 0 90 7045 1365 L 0 45 8065 1365 L 0 45 8661 1365 L 0 45 9085 1365 L 0 45 9413 1365 L 0 45 1289 7169 L 0 45 1885 7169 L 0 45 2309 7169 L 0 45 2637 7169 L 0 45 2905 7169 L 0 45 3132 7169 L 0 45 3329 7169 L 0 45 3502 7169 L 0 90 3657 7124 L 0 45 4677 7169 L 0 45 5273 7169 L 0 45 5697 7169 L 0 45 6025 7169 L 0 45 6293 7169 L 0 45 6520 7169 L 0 45 6716 7169 L 0 45 6890 7169 L 0 90 7045 7124 L 0 45 8065 7169 L 0 45 8661 7169 L 0 45 9085 7169 L 0 45 9413 7169 L 12 6 3633 1233 L 18 18 C 0 -126 C 12 6 6961 1233 L 18 18 C 0 -126 C -18 -6 7099 1257 L -12 -18 C -6 -30 C 0 -18 C 6 -30 C 12 -18 C 18 -6 C 12 0 C 18 6 C 12 18 C 6 30 C 0 18 C -6 30 C -12 18 C -18 6 C -12 0 C 45 0 1050 2013 L 45 0 1050 3190 L 45 0 1050 4025 L 45 0 1050 4673 L 45 0 1050 5202 L 45 0 1050 5649 L 45 0 1050 6037 L 45 0 1050 6379 L 90 0 1050 6685 L 45 0 9404 2013 L 45 0 9404 3190 L 45 0 9404 4025 L 45 0 9404 4673 L 45 0 9404 5202 L 45 0 9404 5649 L 45 0 9404 6037 L 45 0 9404 6379 L 90 0 9359 6685 L -6 12 811 6601 L -18 18 C 126 0 C 6 -18 787 6739 L 18 -12 C 30 -6 C 18 0 C 30 6 C 18 12 C 6 18 C 0 12 C -6 18 C -18 12 C -30 6 C -18 0 C -30 -6 C -18 -12 C -6 -18 C 0 -12 C -6 0 817 1977 L -12 6 C -6 6 C -6 12 C 0 24 C 6 12 C 6 6 C 12 6 C 12 0 C 12 -6 C 18 -12 C 60 -60 C 0 84 C 0 -60 787 4703 L 54 -6 C -6 6 C -6 18 C 0 18 C 6 18 C 12 12 C 18 6 C 12 0 C 18 -6 C 12 -12 C 6 -18 C 0 -18 C -6 -18 C -6 -6 C -12 -6 C 800 0 1050 6362 L 0 77 C 752 0 C 0 -69 C 495 0 C 0 -95 C 370 0 C 0 -240 C 295 0 C 0 -178 C 246 0 C 0 63 C 211 0 C 0 -325 C 184 0 C 0 -124 C 163 0 C 0 -268 C 148 0 C 0 -153 C 133 0 C 0 -273 C 123 0 C 0 -198 C 113 0 C 0 -75 C 106 0 C 0 -143 C 98 0 C 0 -280 C 92 0 C 0 -249 C 86 0 C 0 91 C 82 0 C 0 -349 C 77 0 C 0 -74 C 74 0 C 0 53 C 70 0 C 0 -252 C 67 0 C 0 -179 C 64 0 C 0 17 C 61 0 C 0 -325 C 59 0 C 0 31 C 57 0 C 0 -1 C 54 0 C 0 -239 C 53 0 C 0 -34 C 51 0 C 0 52 C 49 0 C 0 -109 C 47 0 C 0 -172 C 46 0 C 0 21 C 45 0 C 0 -55 C 43 0 C 0 -36 C 42 0 C 0 80 C 41 0 C 0 -25 C 40 0 C 0 -132 C 38 0 C 0 -120 C 38 0 C 0 -5 C 37 0 C 0 167 C 36 0 C 0 -20 C 35 0 C 0 -98 C 34 0 C 0 -16 C 33 0 C 0 -77 C 33 0 C 0 56 C 32 0 C 0 -39 C 31 0 C 0 -16 C 31 0 C 0 -96 C 30 0 C 0 86 C 30 0 C 0 -19 C 28 0 C 0 -1 C 29 0 C 0 83 C 27 0 C 0 -72 C 28 0 C 0 30 C 26 0 C 0 -13 C 27 0 C 0 -4 C 26 0 C 0 -72 C 25 0 C 0 55 C 25 0 C 0 -63 C 24 0 C 0 256 C 25 0 C 0 -45 C 23 0 C 0 -252 C 24 0 C 0 207 C 23 0 C 0 -210 C 22 0 C 0 32 C 23 0 C 0 24 C 21 0 C 0 60 C 22 0 C 0 -155 C 21 0 C 0 77 C 21 0 C 0 81 C 21 0 C 0 -67 C 21 0 C 0 -24 C 20 0 C 0 -15 C 20 0 C 0 124 C 19 0 C 0 -187 C 20 0 C 0 -106 C 19 0 C 0 114 C 19 0 C 0 238 C 18 0 C 0 -1 C 19 0 C 0 -184 C 18 0 C 0 37 C 18 0 C 0 25 C 17 0 C 0 -179 C 18 0 C 0 3 C 17 0 C 0 274 C 17 0 C 0 -235 C 17 0 C 0 172 C 17 0 C 0 -21 C 17 0 C 0 29 C 16 0 C 0 -149 C 16 0 C 0 5 C 16 0 C 0 15 C 16 0 C 0 101 C 16 0 C 0 -156 C 15 0 C 0 65 C 15 0 C 0 149 C 16 0 C 0 -161 C 15 0 C 0 55 C 14 0 C 0 -11 C 15 0 C 0 -6 C 15 0 C 0 72 C 14 0 C 0 -35 C 14 0 C 0 -62 C 15 0 C 0 43 C 14 0 C 0 -34 C 13 0 C 0 31 C 14 0 C 0 -94 C 14 0 C 0 151 C 13 0 C 0 -30 C 14 0 C 0 -39 C 13 0 C 0 -174 C 13 0 C 0 278 C 13 0 C 0 -164 C 13 0 C 0 63 C 13 0 C 0 -58 C 13 0 C 0 29 C 12 0 C 0 -13 C 13 0 C 0 51 C 12 0 C 0 32 C 12 0 C 0 46 C 12 0 C 0 66 C 12 0 C 0 1265 C 12 0 C 0 -124
Date: Thu, 31 May 2001 09:40:54 +0100 From: Roberto Soria <rs1@mssl.ucl.ac.uk> To: Maria Santos-Lleo <xmmhelp@xvsoc01.vilspa.esa.es> Subject: Re: noise in PN power spectra (PR#2016)
> I have sent your question to the EPIC calibration scientist. > He tells me that you already contacted him directly and he > asked you to contact the MPI people. Dear Maria, I did contact Dave Lumb, and he sent me the following reply, also cc ugb@xray.mpe.mpg.de, kendziorra@astro.uni-tuebingen.de It wasn't clear from his email who I should have contacted in MPE, so I referred my problem to you at the help desk for directions. ************************************* DAVE LUMB'S EMAIL: I assume your data on LMCX-3 also refers to the PN? You should be aware that the is a genarally more noisy operation in TIMING mode and if some component of that is instrumental it might show up in yur power spectra. Can you let us know which observation this was from, as there have been quite a number of tests on different instrument settings. Because of the cyclical nature of the readout, one could expect different timescales to appear. The most likely is cycle of one CCD readout of the 199 rows, but that is <6msec or a freq ~170 Hz. There is a logical sequence of 20 such cycles, which would translate to 8.5 Hz, so neither fit the bill. Perhaps there could be some interference with Housekeeping or other spacecraft clocks (we know something like that happens on MOS) I have copied this reply to the PN experts to see if they can identify these frequencies with something intrinsic to the PN camera design ************************************* I haven't heard anything more after this message, so I was hoping someone could actually offer a practical suggestion on how to fix it, or what parameters to choose in xmmselect to minimise the problem, or something. Many thanks for your effort & best regards, Roberto
Date: Thu, 31 May 2001 14:54:54 +0100 From: Roberto Soria <rs1@mssl.ucl.ac.uk> To: Maria Santos-Lleo <xmmhelp@xvsoc01.vilspa.esa.es> Subject: Re: noise in PN power spectra (PR#2016)
> Though I understand from PN experts at MPE that you have got the answer > directly from them, I copy it here for completeness. Thanks, I contacted Markus directly yesterday, I didn't know you would also end up asking him. Hopefully Markus' paper will offer hints for a solution. If not I'll ask him again directly, without overloading the help desk again! Regards, Roberto
From: Maria Santos-Lleo <xmmhelp@xmm.vilspa.esa.es> To: rs1@mssl.ucl.ac.uk Subject: Re: noise in PN power spectra (PR#2016) Date: Thu May 31 08:18:01 2001
Dear Roberto, I have sent your question to the EPIC calibration scientist. He tells me that you already contacted him directly and he asked you to contact the MPI people. In case you did not contact them , please let me know and I will do it for you. Otherwise, could you please send us the answer so it becames clear in the helpdesk as well ? With many thanks and best regards Maria
From: Maria Santos-Lleo <xmmhelp@xmm.vilspa.esa.es> To: rs1@mssl.ucl.ac.uk Subject: Re: noise in PN power spectra (PR#2016) Date: Thu May 31 09:53:04 2001
Dear Roberto I have asked the PN experts whether they can provide some clues on your problem. I will let you know Best regards Maria Maria Santos-Lleo, XMM-Newton SOC User Support Group
From: Maria Santos-Lleo <xmmhelp@xmm.vilspa.esa.es> To: rs1@mssl.ucl.ac.uk Subject: Re: noise in PN power spectra (PR#2016) Date: Thu May 31 13:40:52 2001
Dear Roberto, Though I understand from PN experts at MPE that you have got the answer directly from them, I copy it here for completeness. I hope this solves your problems With best regards Maria -------------------------------------------------------------------------- Hi Roberto, (...) I've attached a technical paper to this mail describing the readout of the CCD. The more interesting part what your question concerns is the simulated readout of an observation in timing mode (Fig. 3a). For the simulation we took into account the readout window function of the CCD plus additional Poissonian noise. The first peak you can see in the PSD (full range plot) corresponds to the readout cycle of the CCD (5.9ms). As you can see the noise is increasing above 10 Hz which is agreement with your data. You can't see any of the peaks you've in your PSD in both plots. I suppose that the two peaks you are seeing, are caused due to an aliasing effect. The readout of the CCD is a fix window function. Whenever you choose a binning different from the readout time of the CCD, you'll get aliasing effects which are caused due to the superposition of the readout window and the binning window you are using. To prevent this effect you should use multiples of the readout time as bin size. Anyhow, I will check for the effect you saw with some data I have in timing mode as soon as I've extracted the data. I hope this will help you. Markus Kuster
From: Maria Santos-Lleo <xmmhelp@xmm.vilspa.esa.es> To: rs1@mssl.ucl.ac.uk Subject: Re: noise in PN power spectra (PR#2016) Date: Thu May 31 15:23:55 2001
Dear Roberto, Thanks for your answer. I hope you will finally fully understand your data. I wish you exiting XMM-Newton science !. Best regards Maria