Reports 1-1 of 1 Clear search Modify search
AdV-DAQ (Calibration)
rolland, bonnand, gouaty, tacca - 9:09 Friday 18 January 2019 (44428) Print this report
Measurement of B1p_PD1,2_DC frequency dependent response.

On 2019/01/17, when SDB2 minitower was opened, we measured the frequency dependent response of B1p_PD1_DC and B1p_PD2_DC channels.

These channels are used for actuator calibration in free swinging Michelson. They were assumed flat in frequency in the past, but we noticed it is not the case, since the impedance vary with frequency due to the presence of the Audio channel.

To measure it, we have installed a LED in front of the photodidoes, and excited the LED with 16 lines from 0.05 Hz to 210 Hz. The transfer function from the LED excitation to the DC channel gives the DC response shape. We have checked at LAPP that the DAC+LED has a flat response: applying the same excitation to the LED in front of a PCAl photodiode (a simple photodiode with a series resistor for sensing), the TF from the excitation to the photodiode readout was flat.

Time of the injections and measured transfer functions:

  • injection on B1p_PD1:  17/01/2019, 12h58 LT-> 13h07m00     (from shutter open to shutter close) . --> figure 1.  PD1 Audio channel its analog high-pass filter around 16 Hz.
  • injection on B1p_PD2:  17/01/2019, 13h15 LT-> 13h27m00  --> figure 2. PD2 Audio channel its analog high-pass around 2 to 3 Hz.

The figure 3 gives a picture of the LED installed in front of the photodiode B1p_PD2.

The transfer functions have been fitted with a simple pole and simple zero model. The parameteres of the fit are the following (the gain and delay are not relevant for the photodiode calibration and are not reported):

  B1p_PD1_DC B1p_PD2_DC
Pole (Hz) 11.221 +/- 0.022 1.824 +/- 0.024
Zero (Hz) 17.247 +/- 0.035 2.216 +/- 0.028
Chi2/ndf (proba) 98.8/92 (29.5%) 41.6/92 (99.9%)


The cable and LED that was connected to the DAC1955 output has been left in the DET lab board for future measurements.



For the record, here is the list of injected signals, generated in the Acl config EDB

ACL_CONST_CH    LED_offset        "V"    1.0            SAMP_FREQ        +4.0         #LED start blinking above 2.5-3 V

# nameOut - unit - rampTime - sampleRate -         sinus ampl - sinus Freq(Hz)  - phi0(rad) - offset(unit)
ACL_SINEWAVE_CH    LED_sine1        "V"    1            SAMP_FREQ    0.1          0.05        0    0
ACL_SINEWAVE_CH    LED_sine2        "V"    1            SAMP_FREQ    0.1          0.11        0    0
ACL_SINEWAVE_CH    LED_sine3        "V"    1            SAMP_FREQ    0.1          0.51        0    0
ACL_SINEWAVE_CH    LED_sine4        "V"    1            SAMP_FREQ    0.1          0.91        0    0
ACL_SINEWAVE_CH    LED_sine5        "V"    1            SAMP_FREQ    0.05      1.51        0    0
ACL_SINEWAVE_CH    LED_sine6        "V"    1            SAMP_FREQ    0.05      2.11        0    0
ACL_SINEWAVE_CH    LED_sine7        "V"    1            SAMP_FREQ    0.05      3.11        0    0
ACL_SINEWAVE_CH    LED_sine8        "V"    1            SAMP_FREQ    0.03      4.11        0    0
ACL_SINEWAVE_CH    LED_sine9        "V"    1            SAMP_FREQ    0.03      5.11        0    0
ACL_SINEWAVE_CH    LED_sine10        "V"    1            SAMP_FREQ    0.03      7.11        0    0
ACL_SINEWAVE_CH    LED_sine11        "V"    1            SAMP_FREQ    0.05     13.10        0    0
ACL_SINEWAVE_CH    LED_sine12        "V"    1            SAMP_FREQ    0.01     23.10        0    0
ACL_SINEWAVE_CH    LED_sine13        "V"    1            SAMP_FREQ    0.01     43.10        0    0
ACL_SINEWAVE_CH    LED_sine14        "V"    1            SAMP_FREQ    0.01     63.10        0    0
ACL_SINEWAVE_CH    LED_sine15        "V"    1            SAMP_FREQ    0.05    110.10        0    0
ACL_SINEWAVE_CH    LED_sine16        "V"    1            SAMP_FREQ    0.01    210.10        0    0

ACL_SUM_CH        LED_sine        "V"    1    1 LED_sine1 1 LED_sine2 1 LED_sine3 1 LED_sine4 1 LED_sine5 1 LED_sine6 1 LED_sine7 1 LED_sine8 1 LED_sine9 1 LED_sine10 1 LED_sine11 1   LED_sine12 1 LED_sine13 1 LED_sine14 1 LED_sine15 1 LED_sine16

ACL_SUM_CH        LED_cmd_tmp        "V"    1            1 LED_offset         1 LED_sine
ACL_CLIP_CH        LED_cmd            LED_cmd_tmp        0    +5

And channel sent to the DAC1955, with gain 2.  A cable LEMO3 to BNC adapted to the DAC1955 output was used for the DAC.

ACL_DAC_CH    dac1955_EDB_DAC_ch01        1        DAC1955_FREQ        LED_cmd                    0    2    "None"                    "ad1955-10v"    ""

Images attached to this report
Comments to this report:
rolland - 15:39 Wednesday 06 February 2019 (44756) Print this report

Today around 11h30 UTC, we have added the whitened DC and Blended channels in SDB2_Photodiodes configuration. Before the whitening and blending processing, the raw 100 kHz ADC DC and Audio channels are decimated directly down to 10 kHz (instead of the still standard two-step decimations from 100 kHz to 20 kHz and then to 10 kHz). These new channels will be used for the deltaL reconstruction of free swinging Michelson data.

The figure shows the coherences and TFs from the not-whitened 20 kHz channels to the new whitened 10 kHz channels, for B1p_PD1 and B1p_PD2.

Images attached to this comment
rolland, verkindt, bonnand, gouaty - 15:27 Thursday 14 March 2019 (45259) Print this report

B1p_PD1_DC  frequency-dependent response (vs temperature)

The temperature of B1p_PD1 during the measurement done in previous entry was 38°C.

Profiting of the opening of SDB2 on Tuesday for HQE photodiode installation, we made again the measurement of the B1p_PD1_DC response putting the LED with the same lines in front of the photodiode. The LED has been put quickly after SDB2 opening to be able to measure it during the temperature transient, from 09:55 UTC to 10h50 UTC (on March 12th 2019), while PD1 temperature decreased from 48°C to 42°C. We did the same measurement and fit as in previous entry at different times during that period.

The photodiode temperature when we take free swinging Michelson data is higher, around to 53°C. It is also possible to measure the shape of the B1p_PD1_DC response doing the TF B1p_PD1_DC/B1p_PD2_DC, and fitting both the pole and zero of PD1 and (inversed) pole and zero of PD2. This has been done using free swinging Michelson data from February 15th, see figures 1 and 2 (zoom on the residual y-axis).   (GPS 1234299920, during 300 s ; FFTs of 20 s ; fit using points with coherence higher than 90%).

The figure 3 shows the pole and zero frequencies of the B1p_PD1_DC response shape as a function of the temperature T. The linear behaviour is fitted by:

  • f_pole = 13.56 - 0.0606 x T
  • f_zero = 20.29 - 0.079 x T


B1p_PD2_DC frequency-dependent response (vs temperature)

Figure 4 shows the same analysis done for B1p_PD2_DC, but we have only two measurements, one with cold photodiode and one during free swinging Michelson data. The "fit" gives:

  • f_pole = 2.533 - 0.0184 x T
  • f_zero = 3.137 - 0.0239 x T

Note that the B1_PD1_DC and B1_PD2_DC channels have similar poles and zeros (see logbook 45250), and must have a similar temperature dependent behavior.


Update of pole and zero B1p values in SDB2_Photodiodes

We have updated the pole and zero values of the whitening filters for B1p_PD1 and PD2 for the channels SDB2_CAL_B1p_* that are being used for free swinging Michelson analysis.

Instead of using the values measured at 38°C, we now use the values measured around 53°C. We will check in next free swinging Michelson data if the temperature is always similar and if the new whitening filters indeed reduce the 5% bias that we found between 10 Hz and 30 Hz in the mirror actuation measurements.


Timing of B1p and B1 photodiodes with LED flashing the IRIG-B signal

The timing of the photodiodes (after replacement of B1 photodiodes by the HQE ones) have been remeasured using the LED flashing the IRIG-B coming from a GPS receiver starting around 10h52 UTC on Tuesday:

  • B1p_PD1: 1236423160 +10 s
  • B1p_PD2:1236423220 +10 s
  • B1_PD1:1236423280 +10 s
  • B1_PD2:1236423300 +10 s

10 s of data have been extracted from the raw_full.ffl for each photodiode and saved in /data/procdata/calibration/rawdataCalibration/2019_03_12_photodiodeHQETiming/


Images attached to this comment