PACS. Photometer SPT during Commissioning Phase. Herschel. PACS Commissioning Phase FFT/SPT report Page 1

Transcription

1 PACS Commissioning Phase FFT/SPT report Page 1 Photometer SPT during Commissioning Phase

2 PACS Commissioning Phase FFT/SPT report Page 2 Req Staring Measurement on Calibration Source for Low Frequency Noise A. History Version Date Author(s) Change description JUN 2009 K.Okumura First issue B. Summary The low frequency noise is measured using a spectral density of the measurements on a stable source over 30 minutes. The source used here is one of the PACS calibration sources (CSs) C. Data Reference Sheet OBSID Date Archive filename comment Phot LongBatch tm on VRL FIST SPT Phot tm DDCS FIST SPT Phot tm direct FIST SPT StaringMeasurementCsLowFreqPhot tm OD0026 OBSID PACS Calibration CPFFT 3- on VRL CPFFT IST414 nstdphotsetup nominal na 0001.tm OD0027 OBSID PACS Calibration CPSPTPHOT 2- DDCS CPSPTPHOT IST520 nstdlowfreq ddcs na 0001.tm OD0027 OBSID PACS Calibration CPSPTPHOT 2- direct CPSPTPHOT IST520 nstdlowfreq direct na 0001.tm D. Test Description The relatively well calibrated source available during IST at ESTEC is the PACS calibration sources (CSs). The noise spectrum is measured on CS2 during 30 minutes. The stability of the responsivity and offset is measured before and after the noise measurement using the chopper frequency of 0.5 Hz between CS1 and CS E. Results E.1. Obtained signal E.2. Spectral distribution of noise In order to assess the noise from the electronic chain without the bolometer, 1 hour data were measured in the configuration where the input power is only from the voltage VRL. Any input comes from the bolometer arrays. Figs 2 show the spectral density of the noise of these measurements. Except for the matrix9 of the red channel with VRL of 0.3V, which goes out of the AD converter s upper limit of 250mV, the other measurements provide meaningfull results. Two peaks around and Hz correspond to 30 and 60 seconds, but the origin of these components are presently unknown. The noise level around 3Hz reaches as low as 3µV/ Hz which is, as seen below, well below the noise level measured with the bolometer. However, this noise level is at the limit of the signal digitization, as shown in req (page??), and its value itsef should be taken with caution.

3 PACS Commissioning Phase FFT/SPT report Page 3 Figure 1: Signal of the pixel [8,8] of the blue [left column] and red detector [right column]. [top] Measurement of electronics (on VRL=0.45V) was obtained right after the PACS photometer switch-on and during the internal calibration sources (CSs) stabilisation. [middle]. DDCS mode and [bottom] direct mode with the standard bias setting.

4 PACS Commissioning Phase FFT/SPT report Page 4 Figure 2: Noise spectrum of blue detectors [left column] and red detector [right column]. The measurements during ILT (2007) with reference voltage VRL = 0.3 [top], 0.4 [middle]. During the commissioning phase VRL = 0.45V [bottom].

5 PACS Commissioning Phase FFT/SPT report Page 5 Figs. 3 comapare the noise spectra in direct mode and processed with and without the deglitching. Figs. 4 show the same for the DDCS mode. The comparison shows clearly the part of contribution of glitches in the noise spectra. The deglitching has been performed with photmmtdeglitching algorhithme with the parameters: scales=3, nsigma=8. It is seen that the contribution of glitches is seen on large frequency range below abour 3Hz. This increases of course the noise level, but the deglitching implemented in IA allows to recover quite well the noise level we had before on ground. Figure 3: Noise spectral density in direct mode, showing a comparison between [left column] with and [right column] without glitches Figs. 5 comapare the noise spectra in direct mode measured during IST (2008 Aug 27), those measured during TVTB (2008 Dec 4) both at ESTEC and those measured during the commissioning phase (2009 Jun 11). Figs. 6

6 PACS Commissioning Phase FFT/SPT report Page 6 Figure 4: Noise spectral density in DDCS mode, showing a comparison between [left column] with and [right column] without glitches

7 PACS Commissioning Phase FFT/SPT report Page 7 show the same for the DDCS mode. These results show that the signal after the glitch removal has a comparable noise spectra as on ground data before launch E.3. Stability of the response and the drift of the offset The chopped measurements before and after the noise measurement allows to assess the stability of the response and the offset within the time scale of the noise measurement. The variations measured during the commissioning phase is comparable with those before the launch. Table 1: Stability assessment of the response and offset Mode Blue [%] Blue [µv] Red [%] Red [µv] response offset response offset direct DDCS F. Conclusions G. IA scripts used / remarks on PCSS onetmtosavefile.py lowfreqmean.py lowfreqmeanplot.py ampli CSs.py

8 PACS Commissioning Phase FFT/SPT report Page 8 Figure 5: Noise spectral density in direct mode. Comparison between [Top] IST, [Middle] TVTB both measured at ESTEC and [Bottom] deglitched Commissioning Phase data

9 PACS Commissioning Phase FFT/SPT report Page 9 Figure 6: Noise spectral density in DDCS mode. Comparison between [Top] IST, [Middle] TVTB both measured at ESTEC and [Bottom] deglitched Commissioning Phase data

10 PACS Commissioning Phase FFT/SPT report Page 10 Req Time constant for flux change in photometry A. History Version Date Author(s) Change description JUN 2009 K.Okumura First issue B. Summary The flux change is produced by chopper movements between the PACS calibration sources (CSs) which have different temperatures. The chopper is moved at different frequencies to explore the different amplitude of the signal modulation as a function of the chopper frequency. The general behaviour during TVTB was found here to be comparable to the results of IST at ESTEC on August However, the measured amplitude of the signal ascillation is between that of TVTB and IST. The reason is currently unknown C. Data Reference Sheet OBSID Date Archive filename FIST SPT Phot tm FIST SPT TimeConstantsFluxChangesPhot tm OD0027 OBSID PACS Calibration CPSPTPHOT 2- CPSPTPHOT IST521 nstdtimeconst fluxchange na 0001.tm D. Test Description During these measurements the chopper moves between 2 positions corresponding to the 2 calibration sources (CSs). Different chopper frequencies are set to analyse the effect of the flux change. The left column of Figs. 7 and Figs. 8 show some typical signals. The 4 different frequencies are repeated twice for 2 different filters of the blue side. The 4th frequency for each filter is 4Hz. This means that there are 5 readouts on each chopper plateau before the average of 4 succesive readouts by SPU. The 3 plateaus every 5 contain 4 succesive readouts which mixes the different plateau levels giving the SPU output without any obvious plateau anymore. With a relatively low bias biases (VH-VL) used here (2.0V for the blue and 1.5V for the red), the time constant is longer than with the nominal bias. The effect of the average of 4 readouts with these biases decreases the modulation amplitude as seen in the figures E. Results For each chopper frequency the amplitude of the signal modulation is measured by the minimum and maximum in the corresponding intervall for each pixel. The pixel to pixel dispersion is then computed as the standard deviation of the amplitudes over all the valid pixel of each matrix. Figs. 9 and Figs. 10 show the results for 3 wavelengths. The amplitude decreases slightly from 0.03Hz to 1Hz and then it drops at 2Hz as expected due to the global time constant of the bolometer and the electronics. This trend was also seen in both ILT and IST data. It is however noticeable that the modulation amplitude of the blue detectors at 3 low frequencies is systematically lower in TVTB data than IST data. The measurements from the commissioning phase show however the regain in amplitude. This amplitude fluctuation is unexplained so far F. Conclusions

11 PACS Commissioning Phase FFT/SPT report Page 11 Figure 7: [Left] Typical blue signal modulation over all the test sequence, [Right] Zoom on the signal with 1 Hz chopper movement. [Top] IST at ESTEC (August 2008) [Middle] TVTB at ESTEC (December 2008) [Bottom] Commissioning Phase (June 2009).

12 PACS Commissioning Phase FFT/SPT report Page 12 Figure 8: [Left] Typical blue signal modulation over all the test sequence, [Right] Zoom on the signal with 1 Hz chopper movement. [Top] IST at ESTEC (August 2008), [Middle] TVTB at ESTEC (December 2008), [Bottom] Commissioning Phase (June 2009).

13 PACS Commissioning Phase FFT/SPT report Page 13 Figure 9: Measured amplitudes of the blue signal modulation : [Left] 70µm filter, and [Right] 100µm filter. The error bars represent the pixel-to-pixel dispersion on each matrix. [Top] IST at ESTEC (August 2008), [Middle] TVTB at ESTEC (December 2008), [Bottom] Commissioning Phase (June 2009)

14 PACS Commissioning Phase FFT/SPT report Page 14 Figure 10: Measured amplitudes of the red signal modulation. The error bars represent the pixel-to-pixel dispersion on each matrix. [Top] IST at ESTEC (August 2008), [Middle] TVTB at ESTEC (December 2008), [Bottom] Commissioning Phase (June 2009)

15 PACS Commissioning Phase FFT/SPT report Page 15 The measurements during the commissioning phase are globally consistent with those of TVTB measurements at ESTEC in Decembre 2008 and those of IST at ESTEC in August However, the signal amplitude is varying from one test to another. The reason for this is currently unknown G. IA scripts used / remarks on PCSS timeconstist.py timeconstist py timeconstist py timeconstistod27.py