#+end_export
* Introduction
** Block Diagram
The block diagram of the setup to be controlled is shown in Fig. [[fig:block_diagram_simplify]].
#+begin_src latex :file sercalo_diagram_simplify.pdf :exports results
\begin{tikzpicture}
\node[DAC] (dac) at (0, 0) {};
\node[block, right=1.2 of dac] (Gi) {$G_i$};
\node[block, right=1.5 of Gi] (Gc) {$G_c$};
\node[addb, right=0.5 of Gc] (add) {};
\node[block, above= of add] (Gn) {$G_n$};
\node[ADC, right=1.2 of add] (adc) {};
\coordinate (GiGc) at ($0.8*(Gi.east) + 0.2*(Gc.west)$);
\node[block] (Zc) at (GiGc|-Gn) {$Z_c$};
\node[block, above=1.2 of Zc] (Ga) {$G_a$};
\node[block, right= of adc] (Gd) {${G_{d}}^{-1}$};
\node[above, align=center] at (Gi.north) {Current\\Amplifier};
\node[left, align=right] at (Zc.west) {Sercalo's\\Impedance};
\node[left, align=right] at (Ga.west) {Voltage\\Amplifier};
\node[above, align=center] at (Gc.north) {Sercalo};
\node[left, align=right] at (Gn.west) {Newport};
\draw[->] ($(dac.west) + (-1, 0)$) --node[midway, sloped]{$/$} (dac.west);
\draw[->] (dac.east) -- (Gi.west) node[above left]{$\begin{bmatrix}U_{c,h} \\ U_{c,v}\end{bmatrix}$} node[below left]{$[V]$};
\draw[->] (Gi.east) -- (Gc.west) node[above left]{$\begin{bmatrix}I_{c,h} \\ I_{c,v}\end{bmatrix}$} node[below left]{$[A]$};
\draw[->] (Gc.east) -- (add.west);
\draw[->] (Gn.south) -- (add.north);
\draw[->] (GiGc)node[branch]{} -- (Zc.south);
\draw[->] (Zc.north) -- (Ga.south) node[below right]{$\begin{bmatrix}\tilde{V}_{c,h} \\ \tilde{V}_{c,v}\end{bmatrix}$} node[below left]{$[V]$};
\draw[->] (Ga.north) -- ++(0, 1.2) node[below right]{$\begin{bmatrix}V_{c,h} \\ V_{c,v}\end{bmatrix}$} node[below left]{$[V]$};
\draw[->] ($(Gn.north) + (0, 1.2)$) -- (Gn.north) node[above right]{$\begin{bmatrix}U_{n,h} \\ U_{n,v}\end{bmatrix}$} node[above left]{$[V]$};
\draw[->] (add.east) -- (adc.west) node[above left]{$\begin{bmatrix}V_{p,h} \\ V_{p,v}\end{bmatrix}$} node[below left]{$[V]$};
\draw[->] (adc.east) --node[midway, sloped]{$/$} (Gd.west);
\draw[->] (Gd.east) --node[midway, sloped]{$/$} ++(1, 0) node[above left]{$\begin{bmatrix} \theta_h \\ \theta_v \end{bmatrix}$} node[below left]{$[rad]$};
\end{tikzpicture}
#+end_src
#+name: fig:block_diagram_simplify
#+caption: Block Diagram of the Experimental Setup
#+RESULTS:
[[file:figs/sercalo_diagram_simplify.png]]
The transfer functions in the system are:
- *Current Amplifier*: from the voltage set by the DAC to the current going to the Sercalo's inductors
\[ G_i = \begin{bmatrix} G_{i,h} & 0 \\ 0 & G_{i,v} \end{bmatrix} \text{ in } \left[ \frac{A}{V} \right] \]
\[ \begin{bmatrix} I_{c,h} \\ I_{c,v} \end{bmatrix} = G_i \begin{bmatrix} U_{c,h} \\ U_{c,v} \end{bmatrix} \]
- *Impedance of the Sercalo* that converts the current going to the sercalo to the voltage across the sercalo:
\[ Z_c = \begin{bmatrix} Z_{c,h} & 0 \\ 0 & Z_{c,v} \end{bmatrix} \text{ in } \left[ \frac{V}{A} \right] \]
\[ \begin{bmatrix} \tilde{V}_{c,h} \\ \tilde{V}_{c,v} \end{bmatrix} = Z_c \begin{bmatrix} I_{c,h} \\ I_{c,v} \end{bmatrix} \]
- *Voltage Amplifier*: from the voltage across the Sercalo inductors to the measured voltage
\[ G_a = \begin{bmatrix} G_{a,h} & 0 \\ 0 & G_{a,v} \end{bmatrix} \text{ in } \left[ \frac{V}{V} \right] \]
\[ \begin{bmatrix} V_{c,h} \\ V_{c,v} \end{bmatrix} = G_a \begin{bmatrix} \tilde{V}_{c,h} \\ \tilde{V}_{c,v} \end{bmatrix} \]
- *Sercalo*: Transfer function from the current going through the sercalo inductors to the 4 quadrant measurement
\[ G_c = \begin{bmatrix} G_{\frac{V_{p,h}}{\tilde{U}_{c,h}}} & G_{\frac{V_{p,h}}{\tilde{U}_{c,v}}} \\ G_{\frac{V_{p,v}}{\tilde{U}_{c,h}}} & G_{\frac{V_{p,v}}{\tilde{U}_{c,v}}} \end{bmatrix} \text{ in } \left[ \frac{V}{A} \right] \]
\[ \begin{bmatrix} V_{p,h} \\ V_{p,v} \end{bmatrix} = G_c \begin{bmatrix} I_{c,h} \\ I_{c,v} \end{bmatrix} \]
- *Newport* Transfer function from the command signal of the Newport to the 4 quadrant measurement
\[ G_n = \begin{bmatrix} G_{\frac{V_{p,h}}{U_{n,h}}} & G_{\frac{V_{p,h}}{U_{n,v}}} \\ G_{\frac{V_{p,v}}{U_{n,h}}} & G_{\frac{V_{n,v}}{U_{n,v}}} \end{bmatrix} \text{ in } \left[ \frac{V}{V} \right] \]
\[ \begin{bmatrix} V_{p,h} \\ V_{p,v} \end{bmatrix} = G_c \begin{bmatrix} V_{n,h} \\ V_{n,v} \end{bmatrix} \]
- *4 Quadrant Diode*: the gain of the 4 quadrant diode in [V/rad] is inverse in order to obtain the physical angle of the beam
\[ G_d = \begin{bmatrix} G_{d,h} & 0 \\ 0 & G_{d,v} \end{bmatrix} \text{ in } \left[\frac{V}{rad}\right] \]
The block diagram with each transfer function is shown in Fig. [[fig:block_diagram]].
#+begin_src latex :file sercalo_diagram.pdf :exports results
\begin{tikzpicture}
\node[DAC] (dac) at (0, 0) {};
\node[block, right=1.5 of dac] (Gi) {$\begin{bmatrix} G_{i,h} & 0 \\ 0 & G_{i,v} \end{bmatrix}$};
\node[block, right=1.8 of Gi] (Gc) {$\begin{bmatrix} G_{\frac{V_{p,h}}{\tilde{U}_{c,h}}} & G_{\frac{V_{p,h}}{\tilde{U}_{c,v}}} \\ G_{\frac{V_{p,v}}{\tilde{U}_{c,h}}} & G_{\frac{V_{p,v}}{\tilde{U}_{c,v}}} \end{bmatrix}$};
\node[addb, right= of Gc] (add) {};
\node[block, above= of add] (Gn) {$\begin{bmatrix} G_{\frac{V_{p,h}}{U_{n,h}}} & G_{\frac{V_{p,h}}{U_{n,v}}} \\ G_{\frac{V_{p,v}}{U_{n,h}}} & G_{\frac{V_{n,v}}{U_{n,v}}} \end{bmatrix}$};
\node[ADC, right = 1.5 of add] (adc) {};
\coordinate (GiGc) at ($0.7*(Gi.east) + 0.3*(Gc.west)$);
\node[block] (Zc) at (GiGc|-Gn) {$\begin{bmatrix} Z_{c,h} & 0 \\ 0 & Z_{c,v} \end{bmatrix}$};
\node[block, above=1.2 of Zc] (Ga) {$\begin{bmatrix} G_{a,h} & 0 \\ 0 & G_{a,v} \end{bmatrix}$};
\node[block, right= of adc] (Gd) {$\begin{bmatrix} G_{d,h}^{-1} & 0 \\ 0 & G_{d,v}^{-1} \end{bmatrix}$};
\node[above, align=center] at (Gi.north) {Current\\Amplifier};
\node[above, align=center] at (Gc.north) {Sercalo};
\node[left, align=right] at (Gn.west) {Newport};
\draw[->] ($(dac.west) + (-1, 0)$) --node[midway, sloped]{$/$} (dac.west);
\draw[->] (dac.east) -- (Gi.west) node[above left]{$\begin{bmatrix}U_{c,h} \\ U_{c,v}\end{bmatrix}$};
\draw[->] (Gi.east) -- (Gc.west) node[above left]{$\begin{bmatrix}I_{c,h} \\ I_{c,v}\end{bmatrix}$};
\draw[->] (Gc.east) -- (add.west);
\draw[->] (Gn.south) -- (add.north);
\draw[->] (GiGc)node[branch]{} -- (Zc.south);
\draw[->] (Zc.north) -- (Ga.south) node[below right]{$\begin{bmatrix}\tilde{V}_{c,h} \\ \tilde{V}_{c,v}\end{bmatrix}$};
\draw[->] (Ga.north) -- ++(0, 1.5) node[below right]{$\begin{bmatrix}V_{c,h} \\ V_{c,v}\end{bmatrix}$};
\draw[->] ($(Gn.north) + (0, 1.5)$) -- (Gn.north) node[above right]{$\begin{bmatrix}U_{n,h} \\ U_{n,v}\end{bmatrix}$};
\draw[->] (add.east) -- (adc.west) node[above left]{$\begin{bmatrix}V_{p,h} \\ V_{p,v}\end{bmatrix}$};
\draw[->] (adc.east) --node[midway, sloped]{$/$} (Gd.west);
\draw[->] (Gd.east) --node[midway, sloped]{$/$} ++(1, 0) node[above left]{$\begin{bmatrix} \theta_h \\ \theta_v \end{bmatrix}$} node[below left]{$[rad]$};
\end{tikzpicture}
#+end_src
#+name: fig:block_diagram
#+attr_latex: :width \linewidth
#+caption: Block Diagram of the Experimental Setup with detailed dynamics
#+RESULTS:
[[file:figs/sercalo_diagram.png]]
** Sercalo
From the Sercalo documentation, we have the parameters shown on table [[tab:sercalo_parameters]].
#+name: tab:sercalo_parameters
#+attr_latex: :environment tabularx :width \linewidth :align lXXXXX
#+attr_latex: :center t :booktabs t :float t
#+caption: Sercalo Parameters
| | Max. Stroke | Res. Freq. | DC Gain | Gain at res. | RC Res. |
| | [deg] | [Hz] | [mA/deg] | [deg/V] | [Ohm] |
|------------------+-------------+------------+----------+--------------+---------|
| AX1 (Horizontal) | 5 | 411.13 | 28.4 | 382.9 | 9.41 |
| AX2 (Vertical) | 5 | 252.5 | 35.2 | 350.4 | |
The Inductance and DC resistance of the two axis of the Sercalo have been measured:
- $L_{c,h} = 0.1\ \text{mH}$
- $L_{c,v} = 0.1\ \text{mH}$
- $R_{c,h} = 9.3\ \Omega$
- $R_{c,v} = 8.3\ \Omega$
Let's first consider the *horizontal direction* and we try to model the Sercalo by a spring/mass/damper system (Fig. [[fig:mech_sercalo]]).
#+begin_src latex :file mech_sercalo.pdf :exports results
\begin{tikzpicture}
\def\massw{2.2} % Width of the masses
\def\massh{0.8} % Height of the masses
\def\spaceh{1.4} % Height of the springs/dampers
\def\dispw{0.3} % Width of the dashed line for the displacement
\def\disph{0.5} % Height of the arrow for the displacements
\def\bracs{0.05} % Brace spacing vertically
\def\brach{-10pt} % Brace shift horizontaly
\draw (-0.5*\massw, 0) -- (0.5*\massw, 0);
% Mass
\draw[fill=white] (-0.5*\massw, \spaceh) rectangle (0.5*\massw, \spaceh+\massh) node[pos=0.5]{$m$};
% Spring, Damper, and Actuator
\draw[spring] (-0.4*\massw, 0) -- (-0.4*\massw, \spaceh) node[midway, left=0.1]{$k$};
\draw[damper] (0, 0) -- ( 0, \spaceh) node[midway, left=0.2]{$c$};
\draw[actuator] ( 0.4*\massw, 0) -- ( 0.4*\massw, \spaceh) node[midway, left=0.1](F){$F$};
% Displacements
\draw[dashed] (0.5*\massw, \spaceh) -- ++(\dispw, 0);
\draw[->] (0.5*\massw+0.5*\dispw, \spaceh) -- ++(0, \disph) node[right]{$x$};
\end{tikzpicture}
#+end_src
#+name: fig:mech_sercalo
#+caption: 1 degree-of-freedom model of the Sercalo
#+RESULTS:
[[file:figs/mech_sercalo.png]]
The equation of motion is:
\begin{align*}
\frac{x}{F} &= \frac{1}{k + c s + m s^2} \\
&= \frac{G_0}{1 + 2 \xi \frac{s}{\omega_0} + \frac{s^2}{\omega_0^2}}
\end{align*}
with:
- $G_0 = 1/k$ is the gain at DC in rad/N
- $\xi = \frac{c}{2 \sqrt{km}}$ is the damping ratio of the system
- $\omega_0 = \sqrt{\frac{k}{m}}$ is the resonance frequency in rad
The force $F$ applied to the mass is proportional to the current $I$ flowing through the voice coils:
\[ \frac{F}{I} = \alpha \]
with $\alpha$ is in $N/A$ and is to be determined.
The current $I$ is also proportional to the voltage at the output of the buffer:
\begin{align*}
\frac{I_c}{U_c} &= \frac{1}{(R + R_c) + L_c s} \\
&\approx 0.02 \left[ \frac{A}{V} \right]
\end{align*}
Let's try to determine the equivalent mass and spring values.
From table [[tab:sercalo_parameters]], for the horizontal direction:
\[ \left| \frac{x}{I} \right|(0) = \left| \alpha \frac{x}{F} \right|(0) = 28.4\ \frac{mA}{deg} = 1.63\ \frac{A}{rad} \]
So:
\[ \alpha \frac{1}{k} = 1.63 \Longleftrightarrow k = \frac{\alpha}{1.63} \left[\frac{N}{rad}\right] \]
We also know the resonance frequency:
\[ \omega_0 = 411.1\ \text{Hz} = 2583\ \frac{rad}{s} \]
And the gain at resonance:
\begin{align*}
\left| \frac{x}{U_c} \right|(j\omega_0) &= \left| 0.02 \frac{x}{I_c} \right| (j\omega_0) \\
&= \left| 0.02 \alpha \frac{x}{F} \right| (j\omega_0) \\
&= 0.02 \alpha \frac{1/k}{2\xi} \\
&= 282.9\ \left[\frac{deg}{V}\right] \\
&= 4.938\ \left[\frac{rad}{V}\right]
\end{align*}
Thus:
\begin{align*}
& \frac{\alpha}{2 \xi k} = 245 \\
\Leftrightarrow & \frac{1.63}{2 \xi} = 245 \\
\Leftrightarrow & \xi = 0.0033 \\
\Leftrightarrow & \xi = 0.33 \%
\end{align*}
#+begin_important
\begin{align*}
G_0 &= \frac{1.63}{\alpha}\ \frac{rad}{N} \\
\xi &= 0.0033 \\
\omega_0 &= 2583\ \frac{rad}{s}
\end{align*}
and in terms of the physical properties:
\begin{align*}
k &= \frac{\alpha}{1.63}\ \frac{N}{rad} \\
\xi &= 0.0033 \\
m &= \frac{\alpha}{1.1 \cdot 10^7}\ \frac{kg}{m^2}
\end{align*}
Thus, we have to determine $\alpha$.
This can be done experimentally by determining the gain at DC or at resonance of the system.
For that, we need to know the angle of the mirror, thus we need to *calibrate* the photo-diodes.
This will be done using the Newport.
#+end_important
** Optical Setup
** Newport
Parameters of the Newport are shown in Fig. [[fig:newport_doc]].
It's dynamics for small angle excitation is shown in Fig. [[fig:newport_gain]].
And we have:
\begin{align*}
G_{n, h}(0) &= 2.62 \cdot 10^{-3}\ \frac{rad}{V} \\
G_{n, v}(0) &= 2.62 \cdot 10^{-3}\ \frac{rad}{V}
\end{align*}
#+name: fig:newport_doc
#+attr_latex: :width \linewidth
#+caption: Documentation of the Newport
[[file:figs/newport_doc.png]]
#+name: fig:newport_gain
#+attr_latex: :width \linewidth
#+caption: Transfer function of the Newport
[[file:figs/newport_gain.png]]
** 4 quadrant Diode
The front view of the 4 quadrant photo-diode is shown in Fig. [[fig:4qd_naming]].
#+begin_src latex :file 4qd_naming.pdf :exports results
\begin{tikzpicture}
\node[draw, circle, minimum size=3cm] (c) at (0, 0){};
\draw[] (c.north) -- (c.south);
\draw[] (c.west) -- (c.east);
\node[] at (-0.6, 0.6){\huge 1};
\node[] at ( 0.6, 0.6){\huge 2};
\node[] at (-0.6, -0.6){\huge 3};
\node[] at ( 0.6, -0.6){\huge 4};
\end{tikzpicture}
#+end_src
#+name: fig:4qd_naming
#+caption: Front view of the 4QD
#+RESULTS:
[[file:figs/4qd_naming.png]]
Each of the photo-diode is amplified using a 4-channel amplifier as shown in Fig. [[fig:4qd_amplifier]].
#+begin_src latex :file 4qd_amplifier.pdf :exports results
\begin{tikzpicture}
\node[draw, minimum width=2cm, minimum height=1.5cm] (ampl) at (0, 0){Amp};
\node[above right] at (ampl.north west){\huge 2};
\node[above left] at (ampl.north east){\huge 1};
\node[below right] at (ampl.south west){\huge 4};
\node[below left] at (ampl.south east){\huge 3};
\end{tikzpicture}
#+end_src
#+name: fig:4qd_amplifier
#+caption: Wiring of the amplifier. The amplifier is located on the bottom right of the board
#+RESULTS:
[[file:figs/4qd_amplifier.png]]
** ADC/DAC
Let's compute the theoretical noise of the ADC/DAC.
\begin{align*}
\Delta V &= 20 V \\
n &= 16bits \\
q &= \Delta V/2^n = 305 \mu V \\
f_N &= 10kHz \\
\Gamma_n &= \frac{q^2}{12 f_N} = 7.76 \cdot 10^{-13} \frac{V^2}{Hz}
\end{align*}
with $\Delta V$ the total range of the ADC, $n$ its number of bits, $q$ the quantization, $f_N$ the sampling frequency and $\Gamma_n$ its theoretical Power Spectral Density.
* Identification of the system dynamics
:PROPERTIES:
:header-args:matlab+: :tangle matlab/sercalo_identification.m
:header-args:matlab+: :comments org :mkdirp yes
:END:
<>
** Introduction :ignore:
In this section, we seek to identify all the blocks as shown in Fig. [[fig:block_diagram_simplify]].
| Signal | Name | Unit |
|-----------------------------------------------------+-------+------|
| Voltage Sent to Sercalo - Horizontal | =Uch= | [V] |
| Voltage Sent to Sercalo - Vertical | =Ucv= | [V] |
| Voltage Sent to Newport - Horizontal | =Unh= | [V] |
| Voltage Sent to Newport - Vertical | =Unv= | [V] |
|-----------------------------------------------------+-------+------|
| 4Q Photodiode Measurement - Horizontal | =Vph= | [V] |
| 4Q Photodiode Measurement - Vertical | =Vpv= | [V] |
| Measured Voltage across the Inductance - Horizontal | =Vch= | [V] |
| Measured Voltage across the Inductance - Vertical | =Vcv= | [V] |
| Newport Metrology - Horizontal | =Vnh= | [V] |
| Newport Metrology - Vertical | =Vnv= | [V] |
|-----------------------------------------------------+-------+------|
| Attocube Measurement | =Va= | [m] |
** ZIP file containing the data and matlab files :ignore:
#+begin_src bash :exports none :results none
if [ matlab/sercalo_identification.m -nt data/sercalo_identification.zip ]; then
cp matlab/sercalo_identification.m sercalo_identification.m;
zip data/sercalo_identification \
mat/data_cal_pd_h.mat \
mat/data_cal_pd_v.mat \
mat/data_uch.mat \
mat/data_ucv.mat \
mat/data_unh.mat \
mat/data_unv.mat \
sercalo_identification.m
rm sercalo_identification.m;
fi
#+end_src
#+begin_note
All the files (data and Matlab scripts) are accessible [[file:data/sercalo_identification.zip][here]].
#+end_note
** Matlab Init :noexport:ignore:
#+begin_src matlab :tangle no :exports none :results silent :noweb yes :var current_dir=(file-name-directory buffer-file-name)
<>
#+end_src
#+begin_src matlab :exports none :results silent :noweb yes
<>
#+end_src
#+begin_src matlab
fs = 1e4;
Ts = 1/fs;
freqs = logspace(1, 3, 1000);
#+end_src
** Calibration of the 4 Quadrant Diode
*** Introduction :ignore:
Prior to any dynamic identification, we would like to be able to determine the meaning of the 4 quadrant diode measurement.
For instance, instead of obtaining transfer function in [V/V] from the input of the sercalo to the measurement voltage of the 4QD, we would like to obtain the transfer function in [rad/V].
This will give insight to physical interpretation.
To calibrate the 4 quadrant photo-diode, we can use the metrology included in the Newport.
We can choose precisely the angle of the Newport mirror and see what is the value measured by the 4 Quadrant Diode.
We then should be able to obtain the "gain" of the 4QD in [V/rad].
*** Input / Output data
The identification data is loaded
#+begin_src matlab
uh = load('mat/data_cal_pd_h.mat', 't', 'Vph', 'Vpv', 'Vnh');
uv = load('mat/data_cal_pd_v.mat', 't', 'Vph', 'Vpv', 'Vnv');
#+end_src
We remove the first seconds where the Sercalo is turned on.
#+begin_src matlab
t0 = 1;
uh.Vph(uh.t>
#+end_src
#+NAME: fig:calib_4qd_h
#+CAPTION: Identification signals when exciting the horizontal direction ([[./figs/calib_4qd_h.png][png]], [[./figs/calib_4qd_h.pdf][pdf]])
[[file:figs/calib_4qd_h.png]]
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
plot(uv.t, uv.Vnv, 'DisplayName', '$Vn_v$');
xlabel('Time [s]');
ylabel('Amplitude [V]');
ax2 = subplot(1, 2, 2);
hold on;
plot(uv.t, uv.Vpv, 'DisplayName', '$Vp_v$');
plot(uv.t, uv.Vph, 'DisplayName', '$Vp_h$');
hold off;
xlabel('Time [s]');
ylabel('Amplitude [V]');
legend();
linkaxes([ax1,ax2],'x');
xlim([uv.t(1), uv.t(end)]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/calib_4qd_v.pdf" :var figsize="full-normal" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:calib_4qd_v
#+CAPTION: Identification signals when exciting in the vertical direction ([[./figs/calib_4qd_v.png][png]], [[./figs/calib_4qd_v.pdf][pdf]])
[[file:figs/calib_4qd_v.png]]
*** Linear Regression to obtain the gain of the 4QD
We plot the angle of mirror
Gain of the Newport metrology in [rad/V].
#+begin_src matlab
gn0 = 2.62e-3;
#+end_src
The angular displacement of the beam is twice the angular displacement of the Newport mirror.
We do a linear regression
\[ y = a x + b \]
where:
- $y$ is the measured voltage of the 4QD in [V]
- $x$ is the beam angle (twice the mirror angle) in [rad]
- $a$ is the identified gain of the 4QD in [rad/V]
The linear regression is shown in Fig. [[fig:4qd_linear_reg]].
#+begin_src matlab
bh = [ones(size(uh.Vnh)) 2*gn0*uh.Vnh]\uh.Vph;
bv = [ones(size(uv.Vnv)) 2*gn0*uv.Vnv]\uv.Vpv;
#+end_src
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
hold on;
plot(2*gn0*uh.Vnh, uh.Vph, 'o', 'DisplayName', 'Exp. data');
plot(2*gn0*[min(uh.Vnh) max(uh.Vnh)], 2*gn0*[min(uh.Vnh) max(uh.Vnh)].*bh(2) + bh(1), 'k--', 'DisplayName', sprintf('%.1e x + %.1e', bh(2), bh(1)))
hold off;
xlabel('$\alpha_{0,h}$ [rad]'); ylabel('$Vp_h$ [V]');
legend();
ax2 = subplot(1, 2, 2);
hold on;
plot(2*gn0*uv.Vnv, uv.Vpv, 'o', 'DisplayName', 'Exp. data');
plot(2*gn0*[min(uv.Vnv) max(uv.Vnv)], 2*gn0*[min(uv.Vnv) max(uv.Vnv)].*bv(2) + bv(1), 'k--', 'DisplayName', sprintf('%.1e x + %.1e', bv(2), bv(1)))
hold off;
xlabel('$\alpha_{0,v}$ [rad]'); ylabel('$Vp_v$ [V]');
legend();
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/4qd_linear_reg.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:4qd_linear_reg
#+CAPTION: Linear Regression ([[./figs/4qd_linear_reg.png][png]], [[./figs/4qd_linear_reg.pdf][pdf]])
[[file:figs/4qd_linear_reg.png]]
Thus, we obtain the "gain of the 4 quadrant photo-diode as shown on table [[tab:gain_4qd]].
#+begin_src matlab :exports results :results value table replace :tangle no :post addhdr(*this*)
data2orgtable([bh(2), bv(2)], {}, {'Horizontal [V/rad]', 'Vertical [V/rad]'}, ' %.1f ');
#+end_src
#+name: tab:gain_4qd
#+attr_latex: :environment tabularx :width 0.5\linewidth :align lX
#+attr_latex: :center t :booktabs t :float t
#+caption: Identified Gain of the 4 quadrant diode
#+RESULTS:
| Horizontal [V/rad] | Vertical [V/rad] |
|--------------------+------------------|
| -31.0 | 36.3 |
#+begin_src matlab
Gd = tf([bh(2) 0 ;
0 bv(2)]);
#+end_src
We obtain:
\begin{align*}
\frac{V_{qd,h}}{\alpha_{0,h}} &\approx 0.032\ \left[ \frac{rad}{V} \right] \\
&\approx 32.3\ \left[ \frac{\mu rad}{mV} \right]
\end{align*}
\begin{align*}
\frac{V_{qd,v}}{\alpha_{0,v}} &\approx 0.028\ \left[ \frac{rad}{V} \right] \\
&\approx 27.6\ \left[ \frac{\mu rad}{mV} \right]
\end{align*}
** Identification of the Sercalo Impedance, Current Amplifier and Voltage Amplifier dynamics
*** Introduction :ignore:
We wish here to determine $G_i$ and $G_a$ shown in Fig. [[fig:block_diagram_simplify]].
We ignore the electro-mechanical coupling.
*** Electrical Schematic
The schematic of the electrical circuit used to drive the Sercalo is shown in Fig. [[fig:current_amplifier]].
#+begin_src latex :file sercalo_amplifier.pdf :exports results
\begin{circuitikz}[]
\ctikzset{bipoles/length=1.0cm}
\draw
(0, -2) node[ground]{} to[vco, V=$U_c$] (0, 0)
to [amp, t={1},i^>=$I_c$, l=BUF] ++(2, 0)
to [R=$R$] ++(2, 0) coordinate(A)
to [L=$L_c$] ++(0, -1)
to [R=$R_c$] ++(0, -1) coordinate(B) node[ground]{}
;
\draw (A) to [amp, i>^=$0$, l={60dB}] ++ (2, 0);
\draw[->] ($(B)+(-0.4, 0)$) -- node[midway, left]{$\tilde{V}_c$} ($(A)+(-0.4, 0)$);
\draw[->] ($(B)+(2, 0)$) -- node[midway, left]{$V_c$} ($(A)+(2, 0)$);
\end{circuitikz}
#+end_src
#+name: fig:current_amplifier
#+caption: Current Amplifier Schematic
#+RESULTS:
[[file:figs/sercalo_amplifier.png]]
The elements are:
- $U_c$: the voltage generated by the DAC
- BUF: is a unity-gain open-loop buffer that allows to increase the output current
- $R$: a chosen resistor that will determine the gain of the current amplifier
- $L_c$: inductor present in the Sercalo
- $R_c$: resistance of the inductor
- $\tilde{V}_c$: voltage measured across the Sercalo's inductor
- $V_c$: amplified voltage measured across the Sercalo's inductor
- $I_c$ is the current going through the Sercalo's inductor
The values of the components have been measured for the horizontal and vertical directions:
- $R_h = 41 \Omega$
- $L_{c,h} = 0.1 mH$
- $R_{c,h} = 9.3 \Omega$
- $R_v = 41 \Omega$
- $L_{c,v} = 0.1 mH$
- $R_{c,v} = 8.3 \Omega$
Let's first determine the transfer function from $U_c$ to $I_c$.
We have that:
\[ U_c = (R + R_c) I_c + L_c s I_c \]
Thus:
\begin{align}
G_i(s) &= \frac{I_c}{U_c} \\
&= \frac{1}{(R + R_c) + L_c s} \\
&= \frac{G_{i,0}}{1 + s/\omega_0}
\end{align}
with
- $G_{i,0} = \frac{1}{R + R_c}$
- $\omega_0 = \frac{R + R_c}{L_c}$
Now, determine the transfer function from $I_c$ to $\tilde{V}_c$:
\[ \tilde{V}_C = R_c I_c + L_c s I_c \]
Thus:
\begin{align}
Z_c(s) &= \frac{\tilde{V}_c}{I_c} \\
&= R_c + L_c s
\end{align}
Finally, the transfer function of the voltage amplifier $G_a$ is simply a low pass filter:
\begin{align}
G_a(s) &= \frac{V_c}{\tilde{V}_c} \\
&= \frac{G_{a,0}}{1 + s/\omega_c}
\end{align}
with
- $G_{a,0}$ is the gain 1000 (60dB)
- $\omega_c$ is the cut-off frequency of the voltage amplifier set to 1000Hz
*** Theoretical Transfer Functions
The values of the components in the current amplifier have been measured.
#+begin_src matlab
Rh = 41; % [Ohm]
Lch = 0.1e-3; % [H]
Rch = 9.3; % [Ohm]
Rv = 41; % [Ohm]
Lcv = 0.1e-3; % [H]
Rcv = 8.3; % [Ohm]
#+end_src
\begin{align*}
G_i(s) &= \frac{1}{(R + R_c) + L_c s} \\
Z_c(s) &= R_c + L_c s \\
G_a(s) &= \frac{1000}{1 + s/\omega_c}
\end{align*}
#+begin_src matlab
Gi = blkdiag(1/(Rh + Rch + Lch * s), 1/(Rv + Rcv + Lcv * s));
Zc = blkdiag(Rch+Lch*s, Rcv+Lcv*s);
Ga = blkdiag(1000/(1 + s/2/pi/1000), 1000/(1 + s/2/pi/1000));
#+end_src
#+begin_src matlab :exports none
freqs = logspace(1, 4, 1000);
figure;
ax1 = subplot(1, 3, 1);
hold on;
plot(freqs, abs(squeeze(freqresp(Gi(1,1), freqs, 'Hz'))), 'DisplayName', '$G_{i, h} = \frac{I_{c,h}}{U_{c,h}}$')
plot(freqs, abs(squeeze(freqresp(Gi(2,2), freqs, 'Hz'))), 'DisplayName', '$G_{i, v} = \frac{I_{c,v}}{U_{c,v}}$')
hold off;
set(gca, 'xscale', 'log'); set(gca, 'yscale', 'log');
xlabel('Frequency [Hz]'); ylabel('Amplitude [A/V]');
legend('location', 'northwest');
ylim([0.01, 1]);
ax2 = subplot(1, 3, 2);
hold on;
plot(freqs, abs(squeeze(freqresp(Zc(1,1), freqs, 'Hz'))), 'DisplayName', '$Z_{c, h} = \frac{\tilde{V}_{c,h}}{I_{c,h}}$')
plot(freqs, abs(squeeze(freqresp(Zc(2,2), freqs, 'Hz'))), 'DisplayName', '$Z_{c, v} = \frac{\tilde{V}_{c,v}}{I_{c,v}}$')
hold off;
set(gca, 'xscale', 'log'); set(gca, 'yscale', 'log');
xlabel('Frequency [Hz]'); ylabel('Amplitude [V/A]');
legend('location', 'southwest');
ylim([1, 100]);
ax3 = subplot(1, 3, 3);
hold on;
plot(freqs, abs(squeeze(freqresp(Ga(1,1), freqs, 'Hz'))), 'DisplayName', '$G_{a, h} = \frac{V_{c,h}}{\tilde{V}_{c,h}}$')
plot(freqs, abs(squeeze(freqresp(Ga(2,2), freqs, 'Hz'))), 'DisplayName', '$G_{a, v} = \frac{V_{c,v}}{\tilde{V}_{c,v}}$')
hold off;
set(gca, 'xscale', 'log'); set(gca, 'yscale', 'log');
xlabel('Frequency [Hz]'); ylabel('Amplitude [V/V]');
legend('location', 'southwest');
ylim([10, 1000]);
linkaxes([ax1, ax2, ax3], 'x');
xlim([freqs(1), freqs(end)]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/current_amplifier_tf.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:current_amplifier_tf
#+CAPTION: Transfer function for the current amplifier ([[./figs/current_amplifier_tf.png][png]], [[./figs/current_amplifier_tf.pdf][pdf]])
[[file:figs/current_amplifier_tf.png]]
#+begin_important
Over the frequency band of interest, the current amplifier transfer function $G_i$ can be considered as constant.
This is the same for the impedance $Z_c$.
#+end_important
#+begin_src matlab
Gi = tf(blkdiag(1/(Rh + Rch), 1/(Rv + Rcv)));
Zc = tf(blkdiag(Rch, Rcv));
#+end_src
*** Identified Transfer Functions
Noise is generated using the DAC ($[U_{c,h}\ U_{c,v}]$) and we measure the output of the voltage amplifier $[V_{c,h}, V_{c,v}]$.
From that, we should be able to identify $G_a Z_c G_i$.
The identification data is loaded.
#+begin_src matlab
uh = load('mat/data_uch.mat', 't', 'Uch', 'Vch');
uv = load('mat/data_ucv.mat', 't', 'Ucv', 'Vcv');
#+end_src
We remove the first seconds where the Sercalo is turned on.
#+begin_src matlab :exports none
t0 = 1;
uh.Uch(uh.t>
#+end_src
#+NAME: fig:current_amplifier_comp_theory_id
#+CAPTION: Identified and Theoretical Transfer Function $G_a G_i$ ([[./figs/current_amplifier_comp_theory_id.png][png]], [[./figs/current_amplifier_comp_theory_id.pdf][pdf]])
[[file:figs/current_amplifier_comp_theory_id.png]]
There is a gain mismatch, that is probably due to bad identification of the inductance and resistance measurement of the sercalo inductors.
Thus, we suppose $G_a$ is perfectly known (the gain and cut-off frequency of the voltage amplifier is very accurate) and that $G_i$ is also well determined as it mainly depends on the resistor used in the amplifier that is well measured.
#+begin_src matlab
Gi_resp_h = abs(GaZcGi_h)./squeeze(abs(freqresp(Ga(1,1)*Zc(1,1), f, 'Hz')));
Gi_resp_v = abs(GaZcGi_v)./squeeze(abs(freqresp(Ga(2,2)*Zc(2,2), f, 'Hz')));
Gi = tf(blkdiag(mean(Gi_resp_h(f>20 & f<200)), mean(Gi_resp_v(f>20 & f<200))));
#+end_src
#+begin_src matlab :exports none
GaZcGi_resp = abs(squeeze(freqresp(Ga*Zc*Gi, f, 'Hz')));
figure;
ax1 = subplot(1, 2, 1);
hold on;
plot(f, abs(GaZcGi_h), 'k-', 'DisplayName', 'Identified')
plot(f, squeeze(GaZcGi_resp(1,1,:)), 'k--', 'DisplayName', 'Theoretical')
title('FRF $G_{i,h} Z_{c,h} G_{a,h} = \frac{V_{c,h}}{U_{c,h}}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
ylabel('Amplitude'); xlabel('Frequency [Hz]');
legend();
hold off;
ax2 = subplot(1, 2, 2);
hold on;
plot(f, abs(GaZcGi_v), 'k-', 'DisplayName', 'Identified')
plot(f, squeeze(GaZcGi_resp(2,2,:)), 'k--', 'DisplayName', 'Theoretical')
title('FRF $G_{a,v} Z_{c,v} G_{i,v} = \frac{V_{c,v}}{U_{c,v}}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
ylabel('Amplitude'); xlabel('Frequency [Hz]');
legend();
hold off;
linkaxes([ax1,ax2],'xy');
xlim([1, 2000]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/current_amplifier_comp_theory_id_bis.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:current_amplifier_comp_theory_id_bis
#+CAPTION: Identified and Theoretical Transfer Function $G_a G_i$ ([[./figs/current_amplifier_comp_theory_id_bis.png][png]], [[./figs/current_amplifier_comp_theory_id_bis.pdf][pdf]])
[[file:figs/current_amplifier_comp_theory_id_bis.png]]
Finally, we have the following transfer functions:
#+begin_src matlab :results output replace :exports results
Gi,Zc,Ga
#+end_src
#+RESULTS:
#+begin_example
ans = filepath;
if ischar(ans), fid = fopen('/tmp/babel-ZKMGJu/matlab-FA7h5L', 'w'); fprintf(fid, '%s\n', ans); fclose(fid);
else, dlmwrite('/tmp/babel-ZKMGJu/matlab-FA7h5L', ans, '\t')
end
'org_babel_eoe'
Gi,Zc,Ga
'org_babel_eoe'
ans = filepath;
if ischar(ans), fid = fopen('/tmp/babel-ZKMGJu/matlab-FA7h5L', 'w'); fprintf(fid, '%s\n', ans); fclose(fid);
else, dlmwrite('/tmp/babel-ZKMGJu/matlab-FA7h5L', ans, '\t')
end
'org_babel_eoe'
ans =
'org_babel_eoe'
Gi,Zc,Ga
Gi =
From input 1 to output...
1: 0.01275
2: 0
From input 2 to output...
1: 0
2: 0.01382
Static gain.
Zc =
From input 1 to output...
1: 9.3
2: 0
From input 2 to output...
1: 0
2: 8.3
Static gain.
Ga =
From input 1 to output...
6.2832e+06
1: ----------
(s+6283)
2: 0
From input 2 to output...
1: 0
6.2832e+06
2: ----------
(s+6283)
Continuous-time zero/pole/gain model.
#+end_example
** Identification of the Sercalo Dynamics
*** Introduction :ignore:
We now wish to identify the dynamics of the Sercalo identified by $G_c$ on the block diagram in Fig. [[fig:block_diagram_simplify]].
To do so, we inject some noise at the input of the current amplifier $[U_{c,h},\ U_{c,v}]$ (one input after the other) and we measure simultaneously the output of the 4QD $[V_{p,h},\ V_{p,v}]$.
The transfer function obtained will be $G_c G_i$, and because we have already identified $G_i$, we can obtain $G_c$ by multiplying the obtained transfer function matrix by ${G_i}^{-1}$.
*** Input / Output data
The identification data is loaded
#+begin_src matlab
uh = load('mat/data_uch.mat', 't', 'Uch', 'Vph', 'Vpv');
uv = load('mat/data_ucv.mat', 't', 'Ucv', 'Vph', 'Vpv');
#+end_src
We remove the first seconds where the Sercalo is turned on.
#+begin_src matlab
t0 = 1;
uh.Uch(uh.t>
#+end_src
#+NAME: fig:identification_uh
#+CAPTION: Identification signals when exciting the horizontal direction ([[./figs/identification_uh.png][png]], [[./figs/identification_uh.pdf][pdf]])
[[file:figs/identification_uh.png]]
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
plot(uv.t, uv.Ucv, 'DisplayName', '$Uc_v$');
xlabel('Time [s]');
ylabel('Amplitude [V]');
ax2 = subplot(1, 2, 2);
hold on;
plot(uv.t, uv.Vpv, 'DisplayName', '$Vp_v$');
plot(uv.t, uv.Vph, 'DisplayName', '$Vp_h$');
hold off;
xlabel('Time [s]');
ylabel('Amplitude [V]');
legend();
linkaxes([ax1,ax2],'x');
xlim([uv.t(1), uv.t(end)]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/identification_uv.pdf" :var figsize="full-normal" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:identification_uv
#+CAPTION: Identification signals when exciting in the vertical direction ([[./figs/identification_uv.png][png]], [[./figs/identification_uv.pdf][pdf]])
[[file:figs/identification_uv.png]]
*** Coherence
The window used for the spectral analysis is an =hanning= windows with temporal size equal to 1 second.
#+begin_src matlab
win = hanning(ceil(1*fs));
#+end_src
#+begin_src matlab
[coh_Uch_Vph, f] = mscohere(uh.Uch, uh.Vph, win, [], [], fs);
[coh_Uch_Vpv, ~] = mscohere(uh.Uch, uh.Vpv, win, [], [], fs);
[coh_Ucv_Vph, ~] = mscohere(uv.Ucv, uv.Vph, win, [], [], fs);
[coh_Ucv_Vpv, ~] = mscohere(uv.Ucv, uv.Vpv, win, [], [], fs);
#+end_src
#+begin_src matlab :exports none
figure;
ax11 = subplot(2, 2, 1);
hold on;
plot(f, coh_Uch_Vph)
set(gca, 'Xscale', 'log');
title('Coherence $\frac{Vp_h}{Uc_h}$')
ylabel('Coherence')
hold off;
ax12 = subplot(2, 2, 2);
hold on;
plot(f, coh_Ucv_Vph)
set(gca, 'Xscale', 'log');
title('Coherence $\frac{Vp_h}{Uc_v}$')
hold off;
ax21 = subplot(2, 2, 3);
hold on;
plot(f, coh_Uch_Vpv)
set(gca, 'Xscale', 'log');
title('Coherence $\frac{Vp_v}{Uc_h}$')
ylabel('Coherence')
xlabel('Frequency [Hz]')
hold off;
ax22 = subplot(2, 2, 4);
hold on;
plot(f, coh_Ucv_Vpv)
set(gca, 'Xscale', 'log');
title('Coherence $\frac{Vp_v}{Uc_v}$')
xlabel('Frequency [Hz]')
hold off;
linkaxes([ax11,ax12,ax21,ax22],'xy');
xlim([10, 1000]); ylim([0, 1]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/coh_sercalo.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:coh_sercalo
#+CAPTION: Coherence ([[./figs/coh_sercalo.png][png]], [[./figs/coh_sercalo.pdf][pdf]])
[[file:figs/coh_sercalo.png]]
*** Estimation of the Frequency Response Function Matrix
We compute an estimate of the transfer functions.
#+begin_src matlab
[tf_Uch_Vph, f] = tfestimate(uh.Uch, uh.Vph, win, [], [], fs);
[tf_Uch_Vpv, ~] = tfestimate(uh.Uch, uh.Vpv, win, [], [], fs);
[tf_Ucv_Vph, ~] = tfestimate(uv.Ucv, uv.Vph, win, [], [], fs);
[tf_Ucv_Vpv, ~] = tfestimate(uv.Ucv, uv.Vpv, win, [], [], fs);
#+end_src
#+begin_src matlab :exports none
figure;
ax11 = subplot(2, 2, 1);
hold on;
plot(f, abs(tf_Uch_Vph))
title('Frequency Response Function $\frac{Vp_h}{Uc_h}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
ylabel('Amplitude [V/V]')
hold off;
ax12 = subplot(2, 2, 2);
hold on;
plot(f, abs(tf_Ucv_Vph))
title('Frequency Response Function $\frac{Vp_h}{Uc_v}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
hold off;
ax21 = subplot(2, 2, 3);
hold on;
plot(f, abs(tf_Uch_Vpv))
title('Frequency Response Function $\frac{Vp_v}{Uc_h}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
ylabel('Amplitude [V/V]')
xlabel('Frequency [Hz]')
hold off;
ax22 = subplot(2, 2, 4);
hold on;
plot(f, abs(tf_Ucv_Vpv))
title('Frequency Response Function $\frac{Vp_v}{Uc_v}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
xlabel('Frequency [Hz]')
hold off;
linkaxes([ax11,ax12,ax21,ax22],'xy');
xlim([10, 1000]); ylim([1e-2, 1e3]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/frf_sercalo_gain.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:frf_sercalo_gain
#+CAPTION: Frequency Response Matrix ([[./figs/frf_sercalo_gain.png][png]], [[./figs/frf_sercalo_gain.pdf][pdf]])
[[file:figs/frf_sercalo_gain.png]]
#+begin_src matlab :exports none
figure;
ax11 = subplot(2, 2, 1);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Uch_Vph)))
title('Frequency Response Function $\frac{Vp_h}{Uc_h}$')
set(gca, 'Xscale', 'log');
ylabel('Phase [deg]')
yticks(-180:90:180);
hold off;
ax12 = subplot(2, 2, 2);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Ucv_Vph)))
title('Frequency Response Function $\frac{Vp_h}{Uc_v}$')
set(gca, 'Xscale', 'log');
yticks(-180:90:180);
hold off;
ax21 = subplot(2, 2, 3);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Uch_Vpv)))
title('Frequency Response Function $\frac{Vp_v}{Uc_h}$')
set(gca, 'Xscale', 'log');
ylabel('Phase [deg]')
xlabel('Frequency [Hz]')
yticks(-180:90:180);
hold off;
ax22 = subplot(2, 2, 4);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Ucv_Vpv)))
title('Frequency Response Function $\frac{Vp_v}{Uc_v}$')
set(gca, 'Xscale', 'log');
xlabel('Frequency [Hz]')
yticks(-180:90:180);
hold off;
linkaxes([ax11,ax12,ax21,ax22],'xy');
xlim([10, 1000]); ylim([-200, 200]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/frf_sercalo_phase.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:frf_sercalo_phase
#+CAPTION: Frequency Response Matrix_Phase ([[./figs/frf_sercalo_phase.png][png]], [[./figs/frf_sercalo_phase.pdf][pdf]])
[[file:figs/frf_sercalo_phase.png]]
*** Time Delay
Now, we would like to remove the time delay included in the FRF prior to the model extraction.
Estimation of the time delay:
#+begin_src matlab
Ts_delay = Ts; % [s]
G_delay = tf(1, 1, 'InputDelay', Ts_delay);
G_delay_resp = squeeze(freqresp(G_delay, f, 'Hz'));
#+end_src
We then remove the time delay from the frequency response function.
#+begin_src matlab
tf_Uch_Vph = tf_Uch_Vph./G_delay_resp;
tf_Uch_Vpv = tf_Uch_Vpv./G_delay_resp;
tf_Ucv_Vph = tf_Ucv_Vph./G_delay_resp;
tf_Ucv_Vpv = tf_Ucv_Vpv./G_delay_resp;
#+end_src
*** Extraction of a transfer function matrix
First we define the initial guess for the resonance frequencies and the weights associated.
#+begin_src matlab
freqs_res_uh = [410]; % [Hz]
freqs_res_uv = [250]; % [Hz]
#+end_src
We then make an initial guess on the complex values of the poles.
#+begin_src matlab
xi = 0.001; % Approximate modal damping
poles_uh = [2*pi*freqs_res_uh*(xi + 1i), 2*pi*freqs_res_uh*(xi - 1i)];
poles_uv = [2*pi*freqs_res_uv*(xi + 1i), 2*pi*freqs_res_uv*(xi - 1i)];
#+end_src
We then define the weight that will be used for the fitting.
Basically, we want more weight around the resonance and at low frequency (below the first resonance).
Also, we want more importance where we have a better coherence.
Finally, we ignore data above some frequency.
#+begin_src matlab
weight_Uch_Vph = coh_Uch_Vph';
weight_Uch_Vpv = coh_Uch_Vpv';
weight_Ucv_Vph = coh_Ucv_Vph';
weight_Ucv_Vpv = coh_Ucv_Vpv';
alpha = 0.1;
for freq_i = 1:length(freqs_res_uh)
weight_Uch_Vph(f>(1-alpha)*freqs_res_uh(freq_i) & f<(1 + alpha)*freqs_res_uh(freq_i)) = 10;
weight_Uch_Vpv(f>(1-alpha)*freqs_res_uh(freq_i) & f<(1 + alpha)*freqs_res_uh(freq_i)) = 10;
weight_Ucv_Vph(f>(1-alpha)*freqs_res_uv(freq_i) & f<(1 + alpha)*freqs_res_uv(freq_i)) = 10;
weight_Ucv_Vpv(f>(1-alpha)*freqs_res_uv(freq_i) & f<(1 + alpha)*freqs_res_uv(freq_i)) = 10;
end
weight_Uch_Vph(f>1000) = 0;
weight_Uch_Vpv(f>1000) = 0;
weight_Ucv_Vph(f>1000) = 0;
weight_Ucv_Vpv(f>1000) = 0;
#+end_src
The weights are shown in Fig. [[fig:weights_sercalo]].
#+begin_src matlab :exports none
figure;
hold on;
plot(f, weight_Uch_Vph, 'DisplayName', '$W_{U_{ch},V_{ph}}$');
plot(f, weight_Uch_Vpv, 'DisplayName', '$W_{U_{ch},V_{pv}}$');
plot(f, weight_Ucv_Vph, 'DisplayName', '$W_{U_{cv},V_{ph}}$');
plot(f, weight_Ucv_Vpv, 'DisplayName', '$W_{U_{cv},V_{pv}}$');
hold off;
xlabel('Frequency [Hz]'); ylabel('Weight Amplitude');
set(gca, 'xscale', 'log');
xlim([f(1), f(end)]);
legend('location', 'northwest');
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/weights_sercalo.pdf" :var figsize="wide-normal" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:weights_sercalo
#+CAPTION: Weights amplitude ([[./figs/weights_sercalo.png][png]], [[./figs/weights_sercalo.pdf][pdf]])
[[file:figs/weights_sercalo.png]]
When we set some options for =vfit3=.
#+begin_src matlab
opts = struct();
opts.stable = 1; % Enforce stable poles
opts.asymp = 1; % Force D matrix to be null
opts.relax = 1; % Use vector fitting with relaxed non-triviality constraint
opts.skip_pole = 0; % Do NOT skip pole identification
opts.skip_res = 0; % Do NOT skip identification of residues (C,D,E)
opts.cmplx_ss = 0; % Create real state space model with block diagonal A
opts.spy1 = 0; % No plotting for first stage of vector fitting
opts.spy2 = 0; % Create magnitude plot for fitting of f(s)
#+end_src
We define the number of iteration.
#+begin_src matlab
Niter = 5;
#+end_src
An we run the =vectfit3= algorithm.
#+begin_src matlab
for iter = 1:Niter
[SER_Uch_Vph, poles, ~, fit_Uch_Vph] = vectfit3(tf_Uch_Vph.', 1i*2*pi*f, poles_uh, weight_Uch_Vph, opts);
end
for iter = 1:Niter
[SER_Uch_Vpv, poles, ~, fit_Uch_Vpv] = vectfit3(tf_Uch_Vpv.', 1i*2*pi*f, poles_uh, weight_Uch_Vpv, opts);
end
for iter = 1:Niter
[SER_Ucv_Vph, poles, ~, fit_Ucv_Vph] = vectfit3(tf_Ucv_Vph.', 1i*2*pi*f, poles_uv, weight_Ucv_Vph, opts);
end
for iter = 1:Niter
[SER_Ucv_Vpv, poles, ~, fit_Ucv_Vpv] = vectfit3(tf_Ucv_Vpv.', 1i*2*pi*f, poles_uv, weight_Ucv_Vpv, opts);
end
#+end_src
#+begin_src matlab :exports none
figure;
ax11 = subplot(2, 2, 1);
hold on;
plot(f, abs(tf_Uch_Vph))
plot(f, abs(fit_Uch_Vph))
title('Frequency Response Function $\frac{Vp_h}{Uc_h}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
ylabel('Amplitude [V/V]')
hold off;
ax12 = subplot(2, 2, 2);
hold on;
plot(f, abs(tf_Ucv_Vph))
plot(f, abs(fit_Ucv_Vph))
title('Frequency Response Function $\frac{Vp_h}{Uc_v}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
hold off;
ax21 = subplot(2, 2, 3);
hold on;
plot(f, abs(tf_Uch_Vpv))
plot(f, abs(fit_Uch_Vpv))
title('Frequency Response Function $\frac{Vp_v}{Uc_h}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
ylabel('Amplitude [V/V]')
xlabel('Frequency [Hz]')
hold off;
ax22 = subplot(2, 2, 4);
hold on;
plot(f, abs(tf_Ucv_Vpv))
plot(f, abs(fit_Ucv_Vpv))
title('Frequency Response Function $\frac{Vp_v}{Uc_v}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
xlabel('Frequency [Hz]')
hold off;
linkaxes([ax11,ax12,ax21,ax22],'xy');
xlim([10, 1000]); ylim([1e-2, 1e3]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/identification_matrix_fit.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:identification_matrix_fit
#+CAPTION: Transfer Function Extraction of the FRF matrix ([[./figs/identification_matrix_fit.png][png]], [[./figs/identification_matrix_fit.pdf][pdf]])
[[file:figs/identification_matrix_fit.png]]
#+begin_src matlab :exports none
figure;
ax11 = subplot(2, 2, 1);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Uch_Vph)))
plot(f, 180/pi*unwrap(angle(fit_Uch_Vph)))
title('Frequency Response Function $\frac{Vp_h}{Uc_h}$')
set(gca, 'Xscale', 'log');
ylabel('Phase [deg]')
yticks(-180:90:180);
hold off;
ax12 = subplot(2, 2, 2);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Ucv_Vph)))
plot(f, 180/pi*unwrap(angle(fit_Ucv_Vph)))
title('Frequency Response Function $\frac{Vp_h}{Uc_v}$')
set(gca, 'Xscale', 'log');
yticks(-180:90:180);
hold off;
ax21 = subplot(2, 2, 3);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Uch_Vpv)))
plot(f, 180/pi*unwrap(angle(fit_Uch_Vpv)))
title('Frequency Response Function $\frac{Vp_v}{Uc_h}$')
set(gca, 'Xscale', 'log');
ylabel('Phase [deg]')
xlabel('Frequency [Hz]')
yticks(-180:90:180);
hold off;
ax22 = subplot(2, 2, 4);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Ucv_Vpv)))
plot(f, 180/pi*unwrap(angle(fit_Ucv_Vpv)))
title('Frequency Response Function $\frac{Vp_v}{Uc_v}$')
set(gca, 'Xscale', 'log');
xlabel('Frequency [Hz]')
yticks(-180:90:180);
hold off;
linkaxes([ax11,ax12,ax21,ax22],'xy');
xlim([10, 1000]); ylim([-200, 200]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/identification_matrix_fit_phase.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:identification_matrix_fit_phase
#+CAPTION: Transfer Function Extraction of the FRF matrix ([[./figs/identification_matrix_fit_phase.png][png]], [[./figs/identification_matrix_fit_phase.pdf][pdf]])
[[file:figs/identification_matrix_fit_phase.png]]
And finally, we create the identified $G_c$ matrix by multiplying by ${G_i}^{-1}$.
#+begin_src matlab
G_Uch_Vph = tf(minreal(ss(full(SER_Uch_Vph.A),SER_Uch_Vph.B,SER_Uch_Vph.C,SER_Uch_Vph.D)));
G_Ucv_Vph = tf(minreal(ss(full(SER_Ucv_Vph.A),SER_Ucv_Vph.B,SER_Ucv_Vph.C,SER_Ucv_Vph.D)));
G_Uch_Vpv = tf(minreal(ss(full(SER_Uch_Vpv.A),SER_Uch_Vpv.B,SER_Uch_Vpv.C,SER_Uch_Vpv.D)));
G_Ucv_Vpv = tf(minreal(ss(full(SER_Ucv_Vpv.A),SER_Ucv_Vpv.B,SER_Ucv_Vpv.C,SER_Ucv_Vpv.D)));
Gc = [G_Uch_Vph, G_Ucv_Vph;
G_Uch_Vpv, G_Ucv_Vpv]*inv(Gi);
#+end_src
** Identification of the Newport Dynamics
*** Introduction :ignore:
We here identify the transfer function from a reference sent to the Newport $[U_{n,h},\ U_{n,v}]$ to the measurement made by the 4QD $[V_{p,h},\ V_{p,v}]$.
To do so, we inject noise to the Newport $[U_{n,h},\ U_{n,v}]$ and we record the 4QD measurement $[V_{p,h},\ V_{p,v}]$.
*** Input / Output data
The identification data is loaded
#+begin_src matlab
uh = load('mat/data_unh.mat', 't', 'Unh', 'Vph', 'Vpv');
uv = load('mat/data_unv.mat', 't', 'Unv', 'Vph', 'Vpv');
#+end_src
We remove the first seconds where the Sercalo is turned on.
#+begin_src matlab
t0 = 3;
uh.Unh(uh.t>
#+end_src
#+NAME: fig:identification_unh
#+CAPTION: Identification signals when exciting the horizontal direction ([[./figs/identification_unh.png][png]], [[./figs/identification_unh.pdf][pdf]])
[[file:figs/identification_unh.png]]
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
plot(uv.t, uv.Unv, 'DisplayName', '$Un_v$');
xlabel('Time [s]');
ylabel('Amplitude [V]');
ax2 = subplot(1, 2, 2);
hold on;
plot(uv.t, uv.Vpv, 'DisplayName', '$Vp_v$');
plot(uv.t, uv.Vph, 'DisplayName', '$Vp_h$');
hold off;
xlabel('Time [s]');
ylabel('Amplitude [V]');
legend();
linkaxes([ax1,ax2],'x');
xlim([uv.t(1), uv.t(end)]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/identification_unv.pdf" :var figsize="full-normal" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:identification_unv
#+CAPTION: Identification signals when exciting in the vertical direction ([[./figs/identification_unv.png][png]], [[./figs/identification_unv.pdf][pdf]])
[[file:figs/identification_unv.png]]
*** Coherence
The window used for the spectral analysis is an =hanning= windows with temporal size equal to 1 second.
#+begin_src matlab
win = hanning(ceil(1*fs));
#+end_src
#+begin_src matlab
[coh_Unh_Vph, f] = mscohere(uh.Unh, uh.Vph, win, [], [], fs);
[coh_Unh_Vpv, ~] = mscohere(uh.Unh, uh.Vpv, win, [], [], fs);
[coh_Unv_Vph, ~] = mscohere(uv.Unv, uv.Vph, win, [], [], fs);
[coh_Unv_Vpv, ~] = mscohere(uv.Unv, uv.Vpv, win, [], [], fs);
#+end_src
#+begin_src matlab :exports none
figure;
ax11 = subplot(2, 2, 1);
hold on;
plot(f, coh_Unh_Vph)
set(gca, 'Xscale', 'log');
title('Coherence $\frac{Vp_h}{Un_h}$')
ylabel('Coherence')
hold off;
ax12 = subplot(2, 2, 2);
hold on;
plot(f, coh_Unv_Vph)
set(gca, 'Xscale', 'log');
title('Coherence $\frac{Vp_h}{Un_v}$')
hold off;
ax21 = subplot(2, 2, 3);
hold on;
plot(f, coh_Unh_Vpv)
set(gca, 'Xscale', 'log');
title('Coherence $\frac{Vp_v}{Un_h}$')
ylabel('Coherence')
xlabel('Frequency [Hz]')
hold off;
ax22 = subplot(2, 2, 4);
hold on;
plot(f, coh_Unv_Vpv)
set(gca, 'Xscale', 'log');
title('Coherence $\frac{Vp_v}{Un_v}$')
xlabel('Frequency [Hz]')
hold off;
linkaxes([ax11,ax12,ax21,ax22],'xy');
xlim([10, 1000]); ylim([0, 1]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/id_newport_coherence.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:id_newport_coherence
#+CAPTION: Coherence ([[./figs/id_newport_coherence.png][png]], [[./figs/id_newport_coherence.pdf][pdf]])
[[file:figs/id_newport_coherence.png]]
*** Estimation of the Frequency Response Function Matrix
We compute an estimate of the transfer functions.
#+begin_src matlab
[tf_Unh_Vph, f] = tfestimate(uh.Unh, uh.Vph, win, [], [], fs);
[tf_Unh_Vpv, ~] = tfestimate(uh.Unh, uh.Vpv, win, [], [], fs);
[tf_Unv_Vph, ~] = tfestimate(uv.Unv, uv.Vph, win, [], [], fs);
[tf_Unv_Vpv, ~] = tfestimate(uv.Unv, uv.Vpv, win, [], [], fs);
#+end_src
#+begin_src matlab :exports none
figure;
ax11 = subplot(2, 2, 1);
hold on;
plot(f, abs(tf_Unh_Vph))
title('Frequency Response Function $\frac{Vp_h}{Un_h}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
ylabel('Amplitude [V/V]')
hold off;
ax12 = subplot(2, 2, 2);
hold on;
plot(f, abs(tf_Unv_Vph))
title('Frequency Response Function $\frac{Vp_h}{Un_v}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
hold off;
ax21 = subplot(2, 2, 3);
hold on;
plot(f, abs(tf_Unh_Vpv))
title('Frequency Response Function $\frac{Vp_v}{Un_h}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
ylabel('Amplitude [V/V]')
xlabel('Frequency [Hz]')
hold off;
ax22 = subplot(2, 2, 4);
hold on;
plot(f, abs(tf_Unv_Vpv))
title('Frequency Response Function $\frac{Vp_v}{Un_v}$')
set(gca, 'Xscale', 'log'); set(gca, 'Yscale', 'log');
xlabel('Frequency [Hz]')
hold off;
linkaxes([ax11,ax12,ax21,ax22],'xy');
xlim([10, 1000]); ylim([1e-4, 1e1]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/frf_newport_gain.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:frf_newport_gain
#+CAPTION: Frequency Response Matrix ([[./figs/frf_newport_gain.png][png]], [[./figs/frf_newport_gain.pdf][pdf]])
[[file:figs/frf_newport_gain.png]]
#+begin_src matlab :exports none
figure;
ax11 = subplot(2, 2, 1);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Unh_Vph)))
title('Frequency Response Function $\frac{Vp_h}{Un_h}$')
set(gca, 'Xscale', 'log');
ylabel('Phase [deg]')
yticks(-180:90:180);
hold off;
ax12 = subplot(2, 2, 2);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Unv_Vph)))
title('Frequency Response Function $\frac{Vp_h}{Un_v}$')
set(gca, 'Xscale', 'log');
yticks(-180:90:180);
hold off;
ax21 = subplot(2, 2, 3);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Unh_Vpv)))
title('Frequency Response Function $\frac{Vp_v}{Un_h}$')
set(gca, 'Xscale', 'log');
ylabel('Phase [deg]'); xlabel('Frequency [Hz]');
yticks(-180:90:180);
hold off;
ax22 = subplot(2, 2, 4);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Unv_Vpv)))
title('Frequency Response Function $\frac{Vp_v}{Un_v}$')
set(gca, 'Xscale', 'log');
xlabel('Frequency [Hz]');
yticks(-180:90:180);
hold off;
linkaxes([ax11,ax12,ax21,ax22],'xy');
xlim([10, 1000]); ylim([-200, 200]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/frf_newport_phase.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:frf_newport_phase
#+CAPTION: Frequency Response Matrix Phase ([[./figs/frf_newport_phase.png][png]], [[./figs/frf_newport_phase.pdf][pdf]])
[[file:figs/frf_newport_phase.png]]
*** Time Delay
Now, we would like to remove the time delay included in the FRF prior to the model extraction.
Estimation of the time delay:
#+begin_src matlab
Ts_delay = 0.0005; % [s]
G_delay = tf(1, 1, 'InputDelay', Ts_delay);
G_delay_resp = squeeze(freqresp(G_delay, f, 'Hz'));
#+end_src
We then remove the time delay from the frequency response function.
#+begin_src matlab :exports none
figure;
ax11 = subplot(2, 2, 1);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Unh_Vph)))
plot(f, 180/pi*unwrap(angle(tf_Unh_Vph./G_delay_resp)))
title('Frequency Response Function $\frac{Vp_h}{Un_h}$')
set(gca, 'Xscale', 'log');
ylabel('Phase [deg]')
yticks(-180:90:180);
hold off;
ax12 = subplot(2, 2, 2);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Unv_Vph)))
plot(f, 180/pi*unwrap(angle(tf_Unv_Vph./G_delay_resp)))
title('Frequency Response Function $\frac{Vp_h}{Un_v}$')
set(gca, 'Xscale', 'log');
yticks(-180:90:180);
hold off;
ax21 = subplot(2, 2, 3);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Unh_Vpv)))
plot(f, 180/pi*unwrap(angle(tf_Unh_Vpv./G_delay_resp)))
title('Frequency Response Function $\frac{Vp_v}{Un_h}$')
set(gca, 'Xscale', 'log');
ylabel('Phase [deg]')
xlabel('Frequency [Hz]')
yticks(-180:90:180);
hold off;
ax22 = subplot(2, 2, 4);
hold on;
plot(f, 180/pi*unwrap(angle(tf_Unv_Vpv)))
plot(f, 180/pi*unwrap(angle(tf_Unv_Vpv./G_delay_resp)))
title('Frequency Response Function $\frac{Vp_v}{Un_v}$')
set(gca, 'Xscale', 'log');
xlabel('Frequency [Hz]')
yticks(-180:90:180);
hold off;
linkaxes([ax11,ax12,ax21,ax22],'xy');
xlim([10, 1000]); ylim([-200, 200]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/time_delay_newport.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:time_delay_newport
#+CAPTION: Phase change due to time-delay in the Newport dynamics ([[./figs/time_delay_newport.png][png]], [[./figs/time_delay_newport.pdf][pdf]])
[[file:figs/time_delay_newport.png]]
*** Extraction of a transfer function matrix
From Fig. [[fig:frf_newport_gain]], it seems reasonable to model the Newport dynamics as diagonal and constant.
#+begin_src matlab
Gn = blkdiag(tf(mean(abs(tf_Unh_Vph(f>10 & f<100)))), tf(mean(abs(tf_Unv_Vpv(f>10 & f<100)))));
#+end_src
** Full System
We now have identified:
- $G_i$
- $G_a$
- $G_c$
- $G_n$
- $G_d$
We name the input and output of each transfer function:
#+begin_src matlab
Gi.InputName = {'Uch', 'Ucv'};
Gi.OutputName = {'Ich', 'Icv'};
Zc.InputName = {'Ich', 'Icv'};
Zc.OutputName = {'Vtch', 'Vtcv'};
Ga.InputName = {'Vtch', 'Vtcv'};
Ga.OutputName = {'Vch', 'Vcv'};
Gc.InputName = {'Ich', 'Icv'};
Gc.OutputName = {'Vpch', 'Vpcv'};
Gn.InputName = {'Unh', 'Unv'};
Gn.OutputName = {'Vpnh', 'Vpnv'};
Gd.InputName = {'Rh', 'Rv'};
Gd.OutputName = {'Vph', 'Vpv'};
#+end_src
#+begin_src matlab
Sh = sumblk('Vph = Vpch + Vpnh');
Sv = sumblk('Vpv = Vpcv + Vpnv');
#+end_src
#+begin_src matlab
inputs = {'Uch', 'Ucv', 'Unh', 'Unv'};
outputs = {'Vch', 'Vcv', 'Ich', 'Icv', 'Rh', 'Rv', 'Vph', 'Vpv'};
sys = connect(Gi, Zc, Ga, Gc, Gn, inv(Gd), Sh, Sv, inputs, outputs);
#+end_src
The file =mat/plant.mat= is accessible [[./mat/plant.mat][here]].
#+begin_src matlab
save('mat/plant.mat', 'sys', 'Gi', 'Zc', 'Ga', 'Gc', 'Gn', 'Gd');
#+end_src
* Huddle Test
:PROPERTIES:
:header-args:matlab+: :tangle matlab/huddle_test.m
:header-args:matlab+: :comments org :mkdirp yes
:END:
<>
** Introduction :ignore:
The goal is to determine the noise of the photodiodes as well as the noise of the Attocube interferometer.
Multiple measurements are done with different experimental configuration as follow:
#+name: tab:huddle_tests
#+attr_latex: :environment tabularx :width 0.7\linewidth :align lXXX
#+attr_latex: :center t :booktabs t :float t
#+caption: Experimental Configuration for the various Huddle test
| Number | OL/CL | Compensation Unit | Aluminum |
|--------+-------------+-------------------+----------|
| 1 | Open Loop | | |
| 2 | Open Loop | Compensation Unit | |
| 3 | Closed Loop | Compensation Unit | |
| 4 | Open Loop | Compensation Unit | Aluminum |
| 5 | Closed Loop | Compensation Unit | Aluminum |
** Matlab Init :noexport:ignore:
#+begin_src matlab :tangle no :exports none :results silent :noweb yes :var current_dir=(file-name-directory buffer-file-name)
<>
#+end_src
#+begin_src matlab :exports none :results silent :noweb yes
<>
#+end_src
** Load Data
#+begin_src matlab
ht_1 = load('./mat/data_huddle_test_1.mat', 't', 'Vph', 'Vpv', 'Va');
ht_2 = load('./mat/data_huddle_test_2.mat', 't', 'Vph', 'Vpv', 'Va');
ht_3 = load('./mat/data_huddle_test_3.mat', 't', 'Uch', 'Ucv', 'Vph', 'Vpv', 'Va');
ht_4 = load('./mat/data_huddle_test_4.mat', 't', 'Vph', 'Vpv', 'Va');
% ht_5 = load('./mat/data_huddle_test_5.mat', 't', 'Uch', 'Ucv', 'Vph', 'Vpv', 'Va');
#+end_src
#+begin_src matlab
fs = 1e4;
#+end_src
** Pre-processing
#+begin_src matlab
t0 = 1; % [s]
tend = 100; % [s]
ht_s = {ht_1 ht_2 ht_3 ht_4}
for i = 1:length(ht_s)
ht_s{i}.Vph(ht_s{i}.t>
#+end_src
#+NAME: fig:huddle_test_Va
#+CAPTION: Measurement of the Attocube during Huddle Test ([[./figs/huddle_test_Va.png][png]], [[./figs/huddle_test_Va.pdf][pdf]])
[[file:figs/huddle_test_Va.png]]
#+begin_src matlab :exports none
figure;
ax1 = subplot(2, 2, 1)
hold on;
plot(ht_1.t, ht_1.Vph);
plot(ht_1.t, ht_1.Vpv);
hold off;
ylabel('Voltage [V]');
set(gca, 'XTickLabel',[]);
title('OL');
ax2 = subplot(2, 2, 2)
hold on;
plot(ht_2.t, ht_2.Vph);
plot(ht_2.t, ht_2.Vpv);
hold off;
set(gca, 'XTickLabel',[]);
set(gca, 'YTickLabel',[]);
title('OL + CU');
ax3 = subplot(2, 2, 3)
hold on;
plot(ht_3.t, ht_3.Vph);
plot(ht_3.t, ht_3.Vpv);
hold off;
xlabel('Time [s]');
ylabel('Voltage [V]');
title('CL + CU');
ax4 = subplot(2, 2, 4)
hold on;
plot(ht_4.t, ht_4.Vph);
plot(ht_4.t, ht_4.Vpv);
hold off;
xlabel('Time [s]');
set(gca, 'YTickLabel',[]);
title('OL + CU + AL');
linkaxes([ax1 ax2 ax3 ax4], 'xy');
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/huddle_test_4qd.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:huddle_test_4qd
#+CAPTION: Measurement of the 4QD during the Huddle tests ([[./figs/huddle_test_4qd.png][png]], [[./figs/huddle_test_4qd.pdf][pdf]])
[[file:figs/huddle_test_4qd.png]]
** Power Spectral Density
#+begin_src matlab
win = hanning(ceil(1*fs));
#+end_src
#+begin_src matlab
[psd_Va1, f] = pwelch(ht_1.Va, win, [], [], fs);
[psd_Va2, ~] = pwelch(ht_2.Va, win, [], [], fs);
[psd_Va3, ~] = pwelch(ht_3.Va, win, [], [], fs);
[psd_Va4, ~] = pwelch(ht_4.Va, win, [], [], fs);
#+end_src
#+begin_src matlab
[psd_Vph1, ~] = pwelch(ht_1.Vph, win, [], [], fs);
[psd_Vph2, ~] = pwelch(ht_2.Vph, win, [], [], fs);
[psd_Vph3, ~] = pwelch(ht_3.Vph, win, [], [], fs);
[psd_Vph4, ~] = pwelch(ht_4.Vph, win, [], [], fs);
#+end_src
#+begin_src matlab
[psd_Vpv1, ~] = pwelch(ht_1.Vpv, win, [], [], fs);
[psd_Vpv2, ~] = pwelch(ht_2.Vpv, win, [], [], fs);
[psd_Vpv3, ~] = pwelch(ht_3.Vpv, win, [], [], fs);
[psd_Vpv4, ~] = pwelch(ht_4.Vpv, win, [], [], fs);
#+end_src
#+begin_src matlab :exports none
figure;
hold on;
plot(f, sqrt(psd_Va1), 'DisplayName', 'OL');
plot(f, sqrt(psd_Va2), 'DisplayName', 'OL + CU');
plot(f, sqrt(psd_Va3), 'DisplayName', 'CL + CU');
plot(f, sqrt(psd_Va4), 'DisplayName', 'OL + CU + AL');
hold off;
set(gca, 'xscale', 'log'); set(gca, 'yscale', 'log');
xlabel('Frequency [Hz]');
ylabel('ASD $\left[\frac{m}{\sqrt{Hz}}\right]$');
legend('location', 'northeast');
xlim([1, 1000]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/huddle_test_psd_va.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:huddle_test_psd_va
#+CAPTION: PSD of the Interferometer measurement during Huddle tests ([[./figs/huddle_test_psd_va.png][png]], [[./figs/huddle_test_psd_va.pdf][pdf]])
[[file:figs/huddle_test_psd_va.png]]
#+begin_src matlab :exports none
figure;
hold on;
plot(f, sqrt(psd_Vph1), 'DisplayName', 'OL');
plot(f, sqrt(psd_Vph2), 'DisplayName', 'OL + CU');
plot(f, sqrt(psd_Vph3), 'DisplayName', 'CL + CU');
plot(f, sqrt(psd_Vph4), 'DisplayName', 'OL + CU + AL');
hold off;
set(gca, 'xscale', 'log'); set(gca, 'yscale', 'log');
xlabel('Frequency [Hz]');
ylabel('ASD $\left[\frac{1}{\sqrt{Hz}}\right]$');
legend('location', 'northeast');
xlim([1, 1000]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/huddle_test_4qd_psd.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:huddle_test_4qd_psd
#+CAPTION: PSD of the 4QD signal during Huddle tests ([[./figs/huddle_test_4qd_psd.png][png]], [[./figs/huddle_test_4qd_psd.pdf][pdf]])
[[file:figs/huddle_test_4qd_psd.png]]
** Conclusion
The Attocube's "Environmental Compensation Unit" does not have a significant effect on the stability of the measurement.
* Budget Error
<>
** Introduction :ignore:
*Goals*:
- List all sources of error and compute their effects on the Attocube measurement
- Think about how to determine the value of the individual sources of error
- Sum all the sources of error and determine the limiting ones
*Sources of error for the Attocube measurement*:
- Beam non-perpendicularity to the concave mirror is linked to the non-perfect feedback loop:
- We have only finite gain / limited bandwidth so the Sercalo mirror angle will not be perfect
- The non-perpendicularity is measured by the 4QD and is used as the feedback signal, however this signal is noisy and even with infinite gain, this noise will be transmitted to the angle of the beam
- Sercalo/Newport unwanted translation perpendicular to its surface.
This can be due to:
- Non idealities in the mechanics of the Sercalo
- Temperature variations
- The reproducible part of the perpendicular translation with respect to the angle of the Sercalo can be taken into account and subtracted from the Attocube measurement
- Temperature variations of the metrology frame
- Change in the refractive air index in the beam path.
This can be due to change of Temperature, Pressure and Humidity of the air in the beam path
*Procedure*:
- in section [[sec:sercalo_angle_error]]:
We estimate the effect of an angle error of the Sercalo mirror on the Attocube measurement
- in section [[sec:mirror_perpendicular_motion]]:
The effect of perpendicular motion of the Newport and Sercalo mirrors on the Attocube measurement is determined.
- in section [[sec:effect_refractive_index]]:
We estimate the expected change of refractive index of the air in the beam path and the resulting Attocube measurement error
- in section [[sec:feedback_error]]:
The feedback system using the 4 quadrant diode and the Sercalo is studied.
Sensor noise, actuator noise and their effects on the control error is discussed.
** Matlab Init :noexport:ignore:
#+begin_src matlab :tangle no :exports none :results silent :noweb yes :var current_dir=(file-name-directory buffer-file-name)
<>
#+end_src
#+begin_src matlab :exports none :results silent :noweb yes
<>
#+end_src
** Effect of the Sercalo angle error on the measured distance by the Attocube
<>
To simplify, we suppose that the Newport mirror is a flat mirror (instead of a concave one).
The geometry of the setup is shown in Fig. [[fig:angle_error_schematic_sercalo]] where:
- $O$ is the reference surface of the Attocube
- $S$ is the point where the beam first hits the Sercalo mirror
- $X$ is the point where the beam first hits the Newport mirror
- $\delta \theta_c$ is the angle error from its ideal 45 degrees
We define the angle error range $\delta \theta_c$ where we want to evaluate the distance error $\delta L$ measured by the Attocube.
#+begin_src matlab
thetas_c = logspace(-7, -4, 100); % [rad]
#+end_src
The geometrical parameters of the setup are defined below.
#+begin_src matlab
H = 0.05; % [m]
L = 0.05; % [m]
#+end_src
#+begin_src latex :file angle_error_schematic_sercalo.pdf :exports results
\begin{tikzpicture}
\draw[->] (0, 0)coordinate(O)node[below]{$O (0,0)$} -- ++(1, 0) node[above left]{$x$};
\draw[->] (0, 0) -- ++(0, 1) node[below right]{$y$};
\draw[] (4, 0)coordinate(S)node[below right]{$S (L,0)$} -- ++(45:1);
\draw[] (4, 0) -- ++(225:1);
\draw[] (3, 2) --coordinate[midway](X)node[above]{$X (0,H)$} (5, 2);
\draw[<->] ([shift=(30:1.2)]S.center) arc (30:60:1.2) node[midway, above right]{$\delta \theta_c$};
\draw[red, ->-=.7, -<-=0.3] (O) -- (S);
\draw[red, ->-=.7, -<-=0.3] (S) -- (X);
\end{tikzpicture}
#+end_src
#+NAME: fig:angle_error_schematic_sercalo
#+CAPTION: Schematic of the geometry used to evaluate the effect of $\delta \theta_c$ on the measured distance $\delta L$
#+RESULTS:
[[file:figs/angle_error_schematic_sercalo.png]]
The nominal points $O$, $S$ and $X$ are defined.
#+begin_src matlab
O = [-L, 0];
S = [0, 0];
X = [0, H];
#+end_src
Thus, the initial path length $L$ is:
#+begin_src matlab
path_nominal = norm(S-O) + norm(X-S) + norm(S-X) + norm(O-S);
#+end_src
We now compute the new path length when there is an error angle $\delta \theta_c$ on the Sercalo mirror angle.
#+begin_src matlab
path_length = zeros(size(thetas_c));
for i = 1:length(thetas_c)
theta_c = thetas_c(i);
Y = [H*tan(2*theta_c), H];
M = 2*H/(tan(pi/4-theta_c)+1/tan(2*theta_c))*[1, tan(pi/4-theta_c)];
T = [-L, M(2)+(L+M(1))*tan(4*theta_c)];
path_length(i) = norm(S-O) + norm(Y-S) + norm(M-Y) + norm(T-M);
end
#+end_src
We then compute the distance error and we plot it as a function of the Sercalo angle error (Fig. [[fig:effect_sercalo_angle_distance_meas]]).
#+begin_src matlab
path_error = path_length - path_nominal;
#+end_src
#+begin_src matlab :exports none
figure;
plot(thetas_c, path_error)
set(gca,'xscale','log');
set(gca,'yscale','log');
xlabel('Sercalo angle error [rad]');
ylabel('Attocube measurement error [m]');
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/effect_sercalo_angle_distance_meas.pdf" :var figsize="wide-normal" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:effect_sercalo_angle_distance_meas
#+CAPTION: Effect of an angle error of the Sercalo on the distance error measured by the Attocube ([[./figs/effect_sercalo_angle_distance_meas.png][png]], [[./figs/effect_sercalo_angle_distance_meas.pdf][pdf]])
[[file:figs/effect_sercalo_angle_distance_meas.png]]
And we plot the beam path using Matlab for an high angle to verify that the code is working (Fig. [[fig:simulation_beam_path_high_angle]]).
#+begin_src matlab
theta = 2*2*pi/360; % [rad]
H = 0.05; % [m]
L = 0.05; % [m]
O = [-L, 0];
S = [0, 0];
X = [0, H];
Y = [H*tan(2*theta), H];
M = 2*H/(tan(pi/4-theta)+1/tan(2*theta))*[1, tan(pi/4-theta)];
T = [-L, M(2)+(L+M(1))*tan(4*theta)];
#+end_src
#+begin_src matlab :exports none
figure;
hold on;
plot([-L, -L], [0, H], 'k-'); % Interferometer
plot([-L, 0.1*L], [H, H], 'k-'); % Reflector
plot(0.5*min(L, H)*[-cos(pi/4-theta), cos(pi/4-theta)], 0.5*min(L, H)*[-sin(pi/4-theta), sin(pi/4-theta)], 'k-'); % Tilt-Mirror
plot(0.5*min(L, H)*[-cos(pi/4), cos(pi/4)], 0.5*min(L, H)*[-sin(pi/4), sin(pi/4)], 'k--'); % Initial position of tilt mirror
plot([O(1), S(1), Y(1), M(1), T(1)], [O(2), S(2), Y(2), M(2), T(2)], 'r-');
plot([O(1), S(1), X(1), S(1), O(1)], [O(2), S(2), X(2), S(2), O(2)], 'b--');
hold off;
xlabel('X [m]'); ylabel('Y [m]');
axis equal
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/simulation_beam_path_high_angle.pdf" :var figsize="wide-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:simulation_beam_path_high_angle
#+CAPTION: Simulation of a beam path for high angle error ([[./figs/simulation_beam_path_high_angle.png][png]], [[./figs/simulation_beam_path_high_angle.pdf][pdf]])
[[file:figs/simulation_beam_path_high_angle.png]]
#+begin_important
Based on Fig. [[fig:effect_sercalo_angle_distance_meas]], we see that an angle error $\delta\theta_c$ of the Sercalo mirror induces a distance error $\delta L$ measured by the Attocube which is dependent of the square of $\delta \theta_c$:
\begin{equation}
\delta L = \delta\theta_c^2
\end{equation}
with:
- $\delta L$ expressed in [m]
- $\delta \theta_c$ in [rad]
Some example are shown in table [[tab:effect_angle_error]].
The tracking error of the feedback system used to position the Sercalo mirror should thus be limited to few micro-meters.
#+end_important
#+name: tab:effect_angle_error
#+attr_latex: :environment tabularx :width 0.6\linewidth :align lX
#+attr_latex: :center t :booktabs t :float t
#+caption: Effect of an angle error $\delta \theta_c$ of the Sercalo's mirror on the measurement error $\delta L$ by the Attocube
| Angle Error $\delta \theta_c$ | Distance measurement error $\delta L$ |
|-------------------------------+---------------------------------------|
| $1\,\mu\text{rad}$ | $1\, nm$ |
| $5\,\mu\text{rad}$ | $25\, nm$ |
| $10\,\mu\text{rad}$ | $100\, nm$ |
** Unwanted motion of Sercalo/Newport mirrors perpendicular to its surface
<>
From Figs [[fig:sercalo_perpendicular_motion]] and [[fig:newport_perpendicular_motion]], it is clear that perpendicular motions of the Sercalo mirror and of the Newport mirror have an impact on the measured distance by the Attocube interferometer.
More precisely, if the note:
- $\delta d_c$ the perpendicular motion of the Sercalo's mirror
- $\delta d_n$ the perpendicular motion of the Newport's mirror
We have that:
\begin{align}
\delta L &= 2 \cdot \delta d_c \\
\delta L &= 2 \cdot \delta d_n
\end{align}
Note here that $\delta L$ denote the change of beam traveled distance.
The error in measured distance by the Attocube will we $\delta L/2$.
#+begin_src latex :file sercalo_perpendicular_motion.pdf :exports results
\begin{tikzpicture}
% X-Y axis
\draw[->] (0, 0)coordinate(O) -- ++(1, 0) node[above left]{$x$};
\draw[->] (0, 0) -- ++(0, 1) node[below right]{$y$};
% Sercalo Mirror
\draw[] ($(4, 0)+(225:1)$)coordinate(a) --node[midway, below, rotate=45]{Sercalo}coordinate[midway](S) ($(4, 0)+(45:1)$);
\draw[dashed, name path=Cc--Cd] ($(4, 0)+(135:0.5)+(225:1)$)coordinate(b) -- ($(4, 0)+(135:0.5)+(45:1)$);
% Mirror displacement
\draw[->] (a) --node[midway, below left]{$\delta d_c$} (b);
% Newport Mirror
\draw[] (3, 2) --coordinate[midway](X)node[above]{Newport} (5, 2);
% Nominal Beam path
\draw[red, ->-=.7, -<-=0.3, name path=O--S] (O.center) -- (S.center);
\draw[red, ->-=.7, -<-=0.3] (S.center) --coordinate[midway](c1) (X.center);
% Changed beam path
\path [name intersections={of=O--S and Cc--Cd,by=E}];
\draw[red, dashed, ->-=.7, -<-=0.3] (E) --coordinate[midway](c2) (E|-X);
\draw[<->] (c1) --node[midway, above]{$\frac{\delta L}{2}$} (c2);
\end{tikzpicture}
#+end_src
#+name: fig:sercalo_perpendicular_motion
#+caption: Effect of a Perpendicular motion of the Sercalo Mirror
#+RESULTS:
[[file:figs/sercalo_perpendicular_motion.png]]
#+begin_src latex :file newport_perpendicular_motion.pdf :exports results
\begin{tikzpicture}
% X-Y axis
\draw[->] (0, 0)coordinate(O) -- ++(1, 0) node[above left]{$x$};
\draw[->] (0, 0) -- ++(0, 1) node[below right]{$y$};
% Sercalo Mirror
\draw[] ($(4, 0)+(225:1)$) --node[midway, below, rotate=45]{Sercalo}coordinate[midway](S) ($(4, 0)+(45:1)$);
% Newport Mirror
\draw[] (3, 2)coordinate(a) --coordinate[midway](X) (5, 2);
\draw[dashed] (3, 2.5)coordinate(b) --coordinate[midway](X2)node[above]{Newport} (5, 2.5);
% Mirror displacement
\draw[->] (a) --node[midway, left]{$\frac{\delta L}{2} = \delta d_n$} (b);
% Nominal Beam path
\draw[red, ->-=.7, -<-=0.3, name path=O--S] (O) -- (S);
\draw[red, ->-=.7, -<-=0.3] (S) -- (X);
\draw[dashed, red] (X) -- (X2);
\end{tikzpicture}
#+end_src
#+name: fig:newport_perpendicular_motion
#+caption: Effect of a Perpendicular motion of the Newport Mirror
#+RESULTS:
[[file:figs/newport_perpendicular_motion.png]]
#+begin_important
The motion of the both Sercalo's and Newport's mirrors perpendicular to its surface is fully transmitted to the measured distance by the Attocube interferometer.
This motion can be measured and the repeatable part can be compensated.
However, the non repeatability of this motion should be less than few nano-meters.
#+end_important
** Change in refractive index of the air in the beam path
<>
Three physical properties of the air makes change of the Attocube measurement:
- Temperature: $K_T \approx 1 ppmK^{-1}$
- Pressure: $K_P \approx 0.27 ppm hPa^{-1}$
- Humidity: $K_{HR} \approx 0.01 ppm \% RH^{-1}$
These physical properties should change relatively slowly, however, for a beam path of 100mm:
| Air property Variations | Measurement error |
|-------------------------+-------------------|
| $\Delta T = 1^oC$ | 100nm |
| $\Delta P = 1hPa$ | 27nm |
| $\Delta 10\%RH$ | 10nm |
An *Environmental Compensation Unit* is used and can compensate for variations or air properties up to:
| Air property Variations | Measurement error |
|-------------------------+---------------------|
| $\Delta T = \pm 0.1^oC$ | $\pm 10\,\text{nm}$ |
| $\Delta P = \pm 1hPa$ | $\pm 25\,\text{nm}$ |
| $\Delta \pm 2\%RH$ | $\pm 2\,\text{nm}$ |
#+begin_important
The total measurement error induced by air properties variations is then:
\begin{equation}
\sqrt{20^2 + 50^2 + 4^2} = 54nm
\end{equation}
The beam path should be protected using aluminum to minimize the change in refractive index of the air in the beam path.
#+end_important
** Thermal Expansion of the Metrology Frame
The material used for the metrology frame is Aluminum.
Its linear thermal expansion coefficient is $\alpha = 23 \cdot 10^{-6} K^{-1}$.
The distance between the Attocube head and the Attocube is approximatively equal to 5cm.
\[ \frac{\delta L}{\delta T} \approx 0.05 \cdot 23 \cdot 10^{-6} \approx 1\,\frac{\mu m}{{}^oC} \]
If invar is used ($\alpha = 1.2 \cdot 10^{-6} \, K^{-1}$):
\[ \frac{\delta L}{\delta T} \approx 60 \frac{nm}{{}^oC} \]
Thus, the temperature of the metrology frame should be kept constant to less than $0.1\,^oC$.
** Estimation of the Sercalo angle error due to Noise
<>
*** Introduction :ignore:
In this section, we seek to estimate the angle error $\delta \theta$
Consider the block diagram in Fig. [[fig:feedback_diagram]] with:
- $G$: represents the transfer function from a voltage applied by the Speedgoat DAC used for the Sercalo to the Beam angle
- $K$: is the control law used
The signals are:
- $\delta \theta$: is the "true" laser beam angle
- $\delta \theta_m$: is the measured beam angle ($\delta \theta_m = \delta \theta + n_\theta$)
- $n_\theta$: is the measurement noise of the laser beam angle using the 4 quadrant diode.
It includes:
- ADC noise
- $1/f$ noise, Shot noise, Ambian noise, Intensity noise...
- $d_u$: is noise at the input of the Sercalo.
It includes:
- DAC noise of the speadgoat
- $d$: is disturbance on the angle of the beam.
It includes:
- Angle variations of the Newport mirror
#+begin_src latex :file feedback_diagram.pdf :exports results
\begin{tikzpicture}
\node[block] (K) at (0,0) {$K$};
\node[addb={+}{}{}{}{}, right=1 of K] (addu) {};
\node[block, right=1 of addu] (G){$G$};
\node[addb={+}{}{}{}{}, right=1 of G] (adddy){};
\node[addb={+}{}{}{}{}, below right=1 and 1 of adddy] (addn) {};
\draw[->] (K.east) -- (addu.west);
\draw[->] (addu.east) -- (G.west) node[above left]{$u$};
\draw[<-] (addu.north) -- ++(0, 0.7) node[below right]{$d_u$};
\draw[->] (G.east) -- (adddy.west);
\draw[<-] (addn.east) -- ++(0.8, 0) coordinate[](endpos) node[above left]{$n_\theta$};
\draw[->] (adddy.east) -- (G-|endpos) node[above left]{$\delta\theta$};
\draw[->] (G-|addn) node[branch]{} -- (addn.north);
\draw[->] (addn.west) -| ($(K.west)+(-1, 0)$) -- (K.west) node[above left]{$\delta\theta_m$};
\draw[<-] (adddy.north) -- ++(0, 0.7) node[below right]{$d$};
\end{tikzpicture}
#+end_src
#+name: fig:feedback_diagram
#+caption: Block Diagram of the Feedback system
#+RESULTS:
[[file:figs/feedback_diagram.png]]
*** Estimation of sources of noise and disturbances
Let's estimate the values of $d_u$, $d$ and $n_\theta$.
**** ADC Quantization Noise
The ADC quantization noise is:
\begin{equation}
\Gamma_\text{ADC} = \frac{\left(\frac{\Delta V}{2^n}\right)^2}{12 f_s} \text{ in } \left[ \frac{V^2}{Hz} \right]
\end{equation}
with:
- $\Delta V$: is the range of the ADC in $[V]$
- $n$: is the number of ADC's bits
- $f_s$: is the sampling frequency in $[Hz]$
For the ADC used:
- $\Delta V = 20\, V$
- $n = 16$
- $f_s = 10\, kHz$
#+begin_important
\begin{equation}
\Gamma_\text{ADC}(f) = 7.76 \cdot 10^{-13}\,\left[ \frac{V^2}{Hz} \right]
\end{equation}
#+end_important
**** DAC Quantization Noise
The DAC quantization noise is:
\begin{equation}
\Gamma_\text{DAC} = \frac{\left(\frac{\Delta V}{2^n}\right)^2}{12 f_s} \text{ in } \left[ \frac{V^2}{Hz} \right]
\end{equation}
with:
- $\Delta V$: is the range of the DAC in $[V]$
- $n$: is the number of DAC's bits
- $f_s$: is the sampling frequency in $[Hz]$
For the DAC used:
- $\Delta V = 20\, V$
- $n = 16$
- $f_s = 10\, kHz$
#+begin_important
\begin{equation}
\Gamma_\text{DAC}(f) = 7.76 \cdot 10^{-13}\,\left[ \frac{V^2}{Hz} \right]
\end{equation}
#+end_important
**** Noise of the Newport Mirror angle
Plus, we estimate the effect of DAC quantization noise on the angle error on the Newport mirror.
The gain of the Newport is:
\begin{align*}
\frac{\theta_n}{V_n} &= \frac{26.2}{10}\ \left[ \frac{mrad}{V} \right] \\
&= 2.62 \cdot 10^{-3}\ \left[ \frac{rad}{V} \right]
\end{align*}
\begin{align*}
\Gamma_{\theta_n}(f) &= \left(\frac{\theta_n}{V_n}\right)^2 \cdot \Gamma_\text{DAC}(f) \\
&= (2.62 \cdot 10^{-3})^2 \cdot 7.76 \cdot 10^{-13} \\
&= 3.96 \cdot 10^{-18}\,\left[ \frac{rad^2}{Hz} \right]
\end{align*}
If we integrate that to obtain an rms value:
\begin{align*}
\theta_{n, rms} &= \sqrt{\int_{-f_s/2}^{f_s/2} \Gamma_{\theta_n}(f) df} \\
&= 0.2\, \mu rad
\end{align*}
Which is much less than the noise equivalent angle specified by Newport: $3\, \mu rad\,[rms]$.
Thus, quantization error of the DAC is not a problem.
We expect the angle noise of the Newport mirror to be around $3\, \mu rad\,[rms]$ which is $6\, \mu rad\,[rms]$ for the beam angle.
#+begin_important
If we suppose a white noise, the power spectral density of the beam angle due to the noise of the Newport mirror corresponds to:
\begin{align*}
\Gamma_{d} &= \frac{(6 \cdot 10^{-6})^2}{f_s}\ \left[ \frac{rad^2}{Hz} \right] \\
&= 3.6 \cdot 10^{-15}\ \left[ \frac{rad^2}{Hz} \right]
\end{align*}
#+end_important
**** Disturbances due the Newport Mirror Rotation
We will rotate the Newport mirror in order to simulate a displacement of the Sample:
- The angle range for the Newport mirror is $\pm 26.2\ mrad = \pm 1.5^o$
- The radius of the concave mirror is 200 mm
#+begin_src latex :file newport_angle_concave_mirror.pdf :post pdf2svg(file=*this*, ext="png") :exports results
\begin{tikzpicture}
% X-Y axis
\draw[->] (0, 0)coordinate(O) -- ++(1, 0) node[above left]{$x$};
\draw[->] (0, 0) -- ++(0, 1) node[below right]{$y$};
% Sercalo Mirror
\draw[] ($(4, 0)+(225:1)$) --node[midway, below, rotate=45]{Sercalo}coordinate[midway](S) ($(4, 0)+(45:1)$);
% Concave Newport Mirror
\draw[] ([shift=(260:6)]4, 10)coordinate(a) arc (260:280:6)coordinate(b)coordinate[midway](X);
\node[branch] (C) at (4, 10){};
\draw[dashed, <->] (a) -- node[midway, left]{$R = 200\,mm$} (C);
\draw[dashed] (X) -- (O|-X);
\draw[dashed, <->] (O) -- node[midway, left]{$H = 50\,mm$} (O|-X);
% Nominal Beam path
\draw[red, ->-=.7, -<-=0.3, name path=O--S] (O) -- (S);
\draw[red, ->-=.7, -<-=0.3] (S) -- (X);
\draw[red, dashed] (X) -- (C);
\begin{scope}[rotate around={-10:(X)}]
\draw[draw=blue!50!white, name path=arcb] ([shift=(260:6)]4, 10) arc (260:280:6) coordinate[midway](Xb);
\node[branch, color=blue!50!white] (Cb) at (4, 10){};
\draw[dashed, draw=blue!50!white] (Xb) -- (Cb);
\end{scope}
\path[name path=S--Cb] (S) -- (Cb);
% Changed beam path
\path [name intersections={of=arcb and S--Cb,by=E}];
\draw[red, dashed, ->-=.7, -<-=0.3] (S) -- (E);
\draw[red, dashed] (E) -- (Cb);
% Center of rotation
\node[branch] at (X){};
\node[below left] at (X){$O_M$};
\end{tikzpicture}
#+end_src
#+name: fig:newport_angle_concave_mirror
#+caption: Rotation of the (concave) Newport mirror
#+RESULTS:
[[file:figs/newport_angle_concave_mirror.png]]
If we suppose small angles, the corresponding beam deviation is:
\[ \delta \theta \approx 2*\frac{\alpha R}{H + R} = 1.6 \alpha \]
where $\alpha$ is the rotation of the Newport mirror.
*** Perfect Control
If the feedback is perfect, the Sercalo angle error will be equal to the 4 quadrant diode noise.
Let's estimate the 4QD noise in radians.
If we note $V_1$, $V_2$, $V_3$ and $V_4$ the voltage of each of the quadrant, a measurement error $\delta V_i$ of one of the quadrant will have an effect $\delta \theta$ on the measured angle:
\[ \delta\theta = G \frac{\delta V_i}{V_1 + V_2 + V_3 + V_4} \]
with $G$ is the gain of the 4QD in [rad].
We should then have that the voltage of each quadrant is as large as possible.
Suppose here that $V_i \approx 5V$, $\delta V_i = 1mV$ and $G = 0.03\,rad$, we obtain:
\[ \delta \theta = 0.03 \frac{0.001}{20} = 1.5\, \mu\text{rad} \]
This then corresponds to
\[ \delta L = 10^{-6} \cdot \delta \theta = 1.5\,nm \]
If we just consider the ADC noise:
- the ADC range is $\pm 10V$ with $16\text{ bits}$.
- thus, the LSB corresponds to:
\[ \frac{20}{2^{16}} \approx 0.000305\,V = 0.305\,mV \]
- this corresponds to an error $\delta L \approx 0.5 nm$
*** Error due to DAC noise used for the Sercalo
#+begin_src matlab
load('./mat/plant.mat', 'Gi', 'Gc', 'Gd');
#+end_src
#+begin_src matlab
G = inv(Gd)*Gc*Gi;
#+end_src
Dynamical estimation:
- ASD of DAC noise used for the Sercalo
- Multiply by transfer function from Sercalo voltage to angle estimation using the 4QD
#+begin_src matlab
freqs = logspace(1, 3, 1000);
fs = 1e4;
% ASD of the DAC voltage going to the Sercalo in [V/sqrt(Hz)]
asd_uc = (20/2^16)/sqrt(12*fs)*ones(length(freqs), 1);
% ASD of the measured angle by the QD in [rad/sqrt(Hz)]
asd_theta = asd_uc.*abs(squeeze(freqresp(G(1,1), freqs, 'Hz')));
figure;
loglog(freqs, asd_theta)
#+end_src
Then the corresponding ASD of the measured displacement by the interferometer is:
#+begin_src matlab
asd_L = asd_theta*10^(-6); % [m/sqrt(Hz)]
#+end_src
And we integrate that to have the RMS value:
#+begin_src matlab
cps_L = 1/pi*cumtrapz(2*pi*freqs, (asd_L).^2);
#+end_src
The RMS value is:
#+begin_src matlab :results value replace
sqrt(cps_L(end))
#+end_src
#+RESULTS:
: 1.647e-11
#+begin_src matlab
figure;
loglog(freqs, cps_L)
#+end_src
Let's estimate the beam angle error corresponding to 1 LSB of the sercalo's DAC.
Gain of the Sercalo is approximatively 5 degrees for 10V.
However the beam angle deviation is 4 times the angle deviation of the sercalo mirror, thus:
#+begin_src matlab :results value replace
d_alpha = 4*(20/2^16)*(5*pi/180)/10 % [rad]
#+end_src
#+RESULTS:
: 1.0653e-05
This corresponds to a measurement error of the Attocube equals to (in [m])
#+begin_src matlab :results value replace
1e-6*d_alpha % [m]
#+end_src
#+RESULTS:
: 1.0653e-11
#+begin_important
The DAC noise use for the Sercalo does not limit the performance of the system.
#+end_important
* Plant Uncertainty
** Matlab Init :noexport:ignore:
#+begin_src matlab :tangle no :exports none :results silent :noweb yes :var current_dir=(file-name-directory buffer-file-name)
<>
#+end_src
#+begin_src matlab :exports none :results silent :noweb yes
<>
#+end_src
** Coprime Factorization
#+begin_src matlab
load('mat/plant.mat', 'sys', 'Gi', 'Zc', 'Ga', 'Gc', 'Gn', 'Gd');
#+end_src
#+begin_src matlab
[fact, Ml, Nl] = lncf(Gc*Gi);
#+end_src
* Plant Scaling
<>
** Introduction :ignore:
The goal is the scale the plant prior to control synthesis.
This will simplify the choice of weighting functions and will yield useful insight on the controllability of the plant.
#+name: tab:plant_scaling_values
#+attr_latex: :environment tabularx :width 0.7\linewidth :align lXXX
#+attr_latex: :center t :booktabs t :float t
#+caption: Maximum wanted values for various signals
| | Value | Unit | Variable Name |
|------------------------+-------+-------------+---------------|
| Expected perturbations | 1 | [V] | $U_n$ |
| Maximum input usage | 10 | [V] | $U_c$ |
| Maximum wanted error | 10 | [$\mu rad$] | $\theta$ |
| Measured noise | 5 | [$\mu rad$] | |
** Control Objective
The maximum expected stroke is $y_\text{max} = 3mm \approx 5e^{-2} rad$ at $1Hz$.
The maximum wanted error is $e_\text{max} = 10 \mu rad$.
Thus, we require the sensitivity function at $\omega_0 = 1\text{ Hz}$:
\begin{align*}
|S(j\omega_0)| &< \left| \frac{e_\text{max}}{y_\text{max}} \right| \\
&< 2 \cdot 10^{-4}
\end{align*}
In terms of loop gain, this is equivalent to:
\[ |L(j\omega_0)| > 5 \cdot 10^{3} \]
** General Configuration
The plant is put in a general configuration as shown in Fig. [[fig:general_control_names]].
#+name: fig:general_control_names
#+caption: General Control Configuration
[[file:figs/general_control_names.png]]
* Plant Analysis
<>
** Introduction :ignore:
** Matlab Init :noexport:ignore:
#+begin_src matlab :tangle no :exports none :results silent :noweb yes :var current_dir=(file-name-directory buffer-file-name)
<>
#+end_src
#+begin_src matlab :exports none :results silent :noweb yes
<>
#+end_src
** Load Plant
#+begin_src matlab
load('mat/plant.mat', 'G');
#+end_src
** RGA-Number
#+begin_src matlab
freqs = logspace(2, 4, 1000);
G_resp = freqresp(G, freqs, 'Hz');
A = zeros(size(G_resp));
RGAnum = zeros(1, length(freqs));
for i = 1:length(freqs)
A(:, :, i) = G_resp(:, :, i).*inv(G_resp(:, :, i))';
RGAnum(i) = sum(sum(abs(A(:, :, i)-eye(2))));
end
% RGA = G0.*inv(G0)';
#+end_src
#+begin_src matlab
figure;
plot(freqs, RGAnum);
set(gca, 'xscale', 'log');
#+end_src
#+begin_src matlab
U = zeros(2, 2, length(freqs));
S = zeros(2, 2, length(freqs))
V = zeros(2, 2, length(freqs));
for i = 1:length(freqs)
[Ui, Si, Vi] = svd(G_resp(:, :, i));
U(:, :, i) = Ui;
S(:, :, i) = Si;
V(:, :, i) = Vi;
end
#+end_src
** Rotation Matrix
#+begin_src matlab
G0 = freqresp(G, 0);
#+end_src
* Active Damping
:PROPERTIES:
:header-args:matlab+: :tangle matlab/active_damping.m
:header-args:matlab+: :comments org :mkdirp yes
:END:
<>
** Introduction :ignore:
** Matlab Init :noexport:ignore:
#+begin_src matlab :tangle no :exports none :results silent :noweb yes :var current_dir=(file-name-directory buffer-file-name)
<>
#+end_src
#+begin_src matlab :exports none :results silent :noweb yes
<>
#+end_src
#+begin_src matlab
freqs = logspace(1, 3, 1000);
#+end_src
** Load Plant
#+begin_src matlab
load('mat/plant.mat', 'sys', 'Gi', 'Zc', 'Ga', 'Gc', 'Gn', 'Gd');
#+end_src
** Integral Force Feedback
#+begin_src matlab
bode(sys({'Vch', 'Vcv'}, {'Uch', 'Ucv'}));
#+end_src
#+begin_src matlab
Kppf = blkdiag(-10000/s, tf(0));
Kppf.InputName = {'Vch', 'Vcv'};
Kppf.OutputName = {'Uch', 'Ucv'};
#+end_src
#+begin_src matlab :exports none
figure;
% Magnitude
ax1 = subplot(2,1,1);
hold on;
plot(freqs, abs(squeeze(freqresp(G, freqs, 'Hz'))), 'k-');
set(gca, 'XScale', 'log'); set(gca, 'YScale', 'log');
set(gca, 'XTickLabel',[]);
ylabel('Magnitude [dB]');
hold off;
% Phase
ax2 = subplot(2,1,2);
hold on;
plot(freqs, 180/pi*angle(squeeze(freqresp(G, freqs, 'Hz'))), 'k-');
set(gca,'xscale','log');
yticks(-360:90:180);
ylim([-360 0]);
xlabel('Frequency [Hz]'); ylabel('Phase [deg]');
hold off;
linkaxes([ax1,ax2],'x');
xlim([freqs(1), freqs(end)]);
#+end_src
#+begin_src matlab
inputs = {'Uch', 'Ucv', 'Unh', 'Unv'};
outputs = {'Ich', 'Icv', 'Rh', 'Rv', 'Vph', 'Vpv'};
sys_cl = connect(sys, Kppf, inputs, outputs);
figure; bode(sys_cl({'Vph', 'Vpv'}, {'Uch', 'Ucv'}), sys({'Vph', 'Vpv'}, {'Uch', 'Ucv'}))
#+end_src
** Conclusion
Active damping does not seems to be applicable here.
* Decentralized Control of the Sercalo
:PROPERTIES:
:header-args:matlab+: :tangle matlab/decentralized_control.m
:header-args:matlab+: :comments org :mkdirp yes
:END:
<>
** Introduction :ignore:
In this section, we try to implement a simple decentralized controller.
** ZIP file containing the data and matlab files :ignore:
#+begin_src bash :exports none :results none
if [ matlab/decentralized_control.m -nt data/decentralized_control.zip ]; then
cp matlab/decentralized_control.m decentralized_control.m;
zip data/decentralized_control \
mat/plant.mat \
decentralized_control.m
rm decentralized_control.m;
fi
#+end_src
#+begin_note
All the files (data and Matlab scripts) are accessible [[file:data/decentralized_control.zip][here]].
#+end_note
** Matlab Init :noexport:ignore:
#+begin_src matlab :tangle no :exports none :results silent :noweb yes :var current_dir=(file-name-directory buffer-file-name)
<>
#+end_src
#+begin_src matlab :exports none :results silent :noweb yes
<>
#+end_src
#+begin_src matlab
freqs = logspace(0, 3, 1000);
#+end_src
** Load Plant
#+begin_src matlab
load('mat/plant.mat', 'sys', 'Gi', 'Zc', 'Ga', 'Gc', 'Gn', 'Gd');
#+end_src
** Diagonal Controller
Using =SISOTOOL=, a diagonal controller is designed.
The two SISO loop gains are shown in Fig. [[fig:diag_contr_loop_gain]].
#+begin_src matlab
Kh = -0.25598*(s+112)*(s^2 + 15.93*s + 6.686e06)/((s^2*(s+352.5)*(1+s/2/pi/2000)));
Kv = 10207*(s+55.15)*(s^2 + 17.45*s + 2.491e06)/(s^2*(s+491.2)*(s+7695));
K = blkdiag(Kh, Kv);
K.InputName = {'Rh', 'Rv'};
K.OutputName = {'Uch', 'Ucv'};
#+end_src
#+begin_src matlab :exports none
figure;
% Magnitude
ax1 = subplot(2,1,1);
hold on;
plot(freqs, abs(squeeze(freqresp(Kh*sys('Rh', 'Uch'), freqs, 'Hz'))), 'DisplayName', '$L_h = K_h G_{d,h}^{-1} G_{\frac{V_{p,h}}{\tilde{U}_{c,h}}} G_{i,h} $');
plot(freqs, abs(squeeze(freqresp(Kv*sys('Rv', 'Ucv'), freqs, 'Hz'))), 'DisplayName', '$L_v = K_v G_{d,v}^{-1} G_{\frac{V_{p,v}}{\tilde{U}_{c,v}}} G_{i,v} $');
set(gca, 'XScale', 'log'); set(gca, 'YScale', 'log');
set(gca, 'XTickLabel',[]);
ylabel('Magnitude [dB]');
hold off;
legend('location', 'northeast');
% Phase
ax2 = subplot(2,1,2);
hold on;
plot(freqs, 180/pi*angle(squeeze(freqresp(Kh*sys('Rh', 'Uch'), freqs, 'Hz'))));
plot(freqs, 180/pi*angle(squeeze(freqresp(Kv*sys('Rv', 'Ucv'), freqs, 'Hz'))));
set(gca,'xscale','log');
yticks(-180:90:180);
ylim([-180 180]);
xlabel('Frequency [Hz]'); ylabel('Phase [deg]');
hold off;
linkaxes([ax1,ax2],'x');
xlim([freqs(1), freqs(end)]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/diag_contr_loop_gain.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:diag_contr_loop_gain
#+CAPTION: Loop Gain using the Decentralized Diagonal Controller ([[./figs/diag_contr_loop_gain.png][png]], [[./figs/diag_contr_loop_gain.pdf][pdf]])
[[file:figs/diag_contr_loop_gain.png]]
We then close the loop and we look at the transfer function from the Newport rotation signal to the beam angle (Fig. [[fig:diag_contr_effect_newport]]).
#+begin_src matlab
inputs = {'Uch', 'Ucv', 'Unh', 'Unv'};
outputs = {'Vch', 'Vcv', 'Ich', 'Icv', 'Rh', 'Rv', 'Vph', 'Vpv'};
sys_cl = connect(sys, -K, inputs, outputs);
#+end_src
#+begin_src matlab :exports none
figure;
hold on;
set(gca,'ColorOrderIndex',1);
plot(freqs, abs(squeeze(freqresp(sys('Rh', 'Unh'), freqs, 'Hz'))), '-', 'DisplayName', 'OL - $R_h/U_{n,h}$');
set(gca,'ColorOrderIndex',1);
plot(freqs, abs(squeeze(freqresp(sys_cl('Rh', 'Unh'), freqs, 'Hz'))), '--', 'DisplayName', 'CL - $R_h/U_{n,h}$');
set(gca,'ColorOrderIndex',2);
plot(freqs, abs(squeeze(freqresp(sys('Rv', 'Unv'), freqs, 'Hz'))), '-', 'DisplayName', 'OL - $R_v/U_{n,v}$');
set(gca,'ColorOrderIndex',2);
plot(freqs, abs(squeeze(freqresp(sys_cl('Rv', 'Unv'), freqs, 'Hz'))), '--', 'DisplayName', 'CL - $R_v/U_{n,v}$');
set(gca, 'XScale', 'log'); set(gca, 'YScale', 'log');
xlabel('Frequency [Hz]'); ylabel('Magnitude [dB]');
hold off;
xlim([freqs(1), freqs(end)]);
legend('location', 'southeast');
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/diag_contr_effect_newport.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:diag_contr_effect_newport
#+CAPTION: Effect of the Newport rotation on the beam position when the loop is closed using the Decentralized Diagonal Controller ([[./figs/diag_contr_effect_newport.png][png]], [[./figs/diag_contr_effect_newport.pdf][pdf]])
[[file:figs/diag_contr_effect_newport.png]]
** Save the Controller
#+begin_src matlab
Kd = c2d(K, 1e-4, 'tustin');
#+end_src
The diagonal controller is accessible [[./mat/K_diag.mat][here]].
#+begin_src matlab
save('mat/K_diag.mat', 'K', 'Kd');
#+end_src
* Newport Control
<>
** Introduction :ignore:
In this section, we try to implement a simple decentralized controller for the Newport.
This can be used to align the 4QD:
- once there is a signal from the 4QD, the Newport feedback loop is closed
- thus, the Newport is positioned such that the beam hits the center of the 4QD
- then we can move the 4QD manually in X-Y plane in order to cancel the command signal of the Newport
- finally, we are sure to be aligned when the command signal of the Newport is 0
** Matlab Init :noexport:ignore:
#+begin_src matlab :tangle no :exports none :results silent :noweb yes :var current_dir=(file-name-directory buffer-file-name)
<>
#+end_src
#+begin_src matlab :exports none :results silent :noweb yes
<>
#+end_src
#+begin_src matlab
freqs = logspace(0, 2, 1000);
#+end_src
** Load Plant
#+begin_src matlab
load('mat/plant.mat', 'Gn', 'Gd');
#+end_src
** Analysis
The plant is basically a constant until frequencies up to the required bandwidth.
We get that constant value.
#+begin_src matlab
Gn0 = freqresp(inv(Gd)*Gn, 0);
#+end_src
We design two controller containing 2 integrators and one lead near the crossover frequency set to 10Hz.
#+begin_src matlab
h = 2;
w0 = 2*pi*10;
Knh = 1/Gn0(1,1) * (w0/s)^2 * (1 + s/w0*h)/(1 + s/w0/h)/h;
Knv = 1/Gn0(2,2) * (w0/s)^2 * (1 + s/w0*h)/(1 + s/w0/h)/h;
#+end_src
#+begin_src matlab :exports none
figure;
hold on;
plot(freqs, abs(squeeze(freqresp(Gn0(1,1)*Knh, freqs, 'Hz'))))
hold off;
set(gca, 'xscale', 'log'); set(gca, 'yscale', 'log');
xlabel('Frequency [Hz]'); ylabel('Loop Gain');
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/loop_gain_newport.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:loop_gain_newport
#+CAPTION: Diagonal Loop Gain for the Newport ([[./figs/loop_gain_newport.png][png]], [[./figs/loop_gain_newport.pdf][pdf]])
[[file:figs/loop_gain_newport.png]]
** Save
#+begin_src matlab
Kn = blkdiag(Knh, Knv);
Knd = c2d(Kn, 1e-4, 'tustin');
#+end_src
The controllers can be downloaded [[./mat/K_newport.mat][here]].
#+begin_src matlab
save('mat/K_newport.mat', 'Kn', 'Knd');
#+end_src
* Measurement of the non-repeatability
<>
** Introduction :ignore:
The goal here is the measure the non-repeatability of the setup.
All sources of error (detailed in the budget error in Section [[sec:budget_error]]) will contribute to the non-repeatability of the system.
** Matlab Init :noexport:ignore:
#+begin_src matlab :tangle no :exports none :results silent :noweb yes :var current_dir=(file-name-directory buffer-file-name)
<>
#+end_src
#+begin_src matlab :exports none :results silent :noweb yes
<>
#+end_src
#+begin_src matlab
fs = 1e4;
#+end_src
** Data Load and pre-processing
#+begin_src matlab
uh = load('mat/data_rep_h.mat', ...
't', 'Uch', 'Ucv', ...
'Unh', 'Unv', ...
'Vph', 'Vpv', ...
'Vnh', 'Vnv', ...
'Va');
uv = load('mat/data_rep_v.mat', ...
't', 'Uch', 'Ucv', ...
'Unh', 'Unv', ...
'Vph', 'Vpv', ...
'Vnh', 'Vnv', ...
'Va');
#+end_src
#+begin_src matlab
% Let's start one second after the first command in the system
i0 = find(uh.Unh ~= 0, 1) + fs;
iend = i0+fs*floor((length(uh.t)-i0)/fs);
uh.Uch([1:i0-1, iend:end]) = [];
uh.Ucv([1:i0-1, iend:end]) = [];
uh.Unh([1:i0-1, iend:end]) = [];
uh.Unv([1:i0-1, iend:end]) = [];
uh.Vph([1:i0-1, iend:end]) = [];
uh.Vpv([1:i0-1, iend:end]) = [];
uh.Vnh([1:i0-1, iend:end]) = [];
uh.Vnv([1:i0-1, iend:end]) = [];
uh.Va ([1:i0-1, iend:end]) = [];
uh.t ([1:i0-1, iend:end]) = [];
% We reset the time t
uh.t = uh.t - uh.t(1);
#+end_src
#+begin_src matlab
% Let's start one second after the first command in the system
i0 = find(uv.Unv ~= 0, 1) + fs;
iend = i0+fs*floor((length(uv.t)-i0)/fs);
uv.Uch([1:i0-1, iend:end]) = [];
uv.Ucv([1:i0-1, iend:end]) = [];
uv.Unh([1:i0-1, iend:end]) = [];
uv.Unv([1:i0-1, iend:end]) = [];
uv.Vph([1:i0-1, iend:end]) = [];
uv.Vpv([1:i0-1, iend:end]) = [];
uv.Vnh([1:i0-1, iend:end]) = [];
uv.Vnv([1:i0-1, iend:end]) = [];
uv.Va ([1:i0-1, iend:end]) = [];
uv.t ([1:i0-1, iend:end]) = [];
% We reset the time t
uv.t = uv.t - uv.t(1);
#+end_src
** Some Time domain plots
#+begin_src matlab :exports none
tend = 5; % [s]
figure;
ax1 = subplot(2, 2, 1);
hold on;
plot(uh.t(1:tend*fs), uh.Unh(1:tend*fs));
hold off;
xlabel('Time [s]'); ylabel('Voltage [V]');
title('Newport Tilt - Horizontal Direction');
ax3 = subplot(2, 2, 3);
hold on;
plot(uh.t(1:tend*fs), 1e9*uh.Va(1:tend*fs));
hold off;
xlabel('Time [s]'); ylabel('Distance [nm]');
title('Attocube - Horizontal Direction');
ax2 = subplot(2, 2, 2);
hold on;
plot(uv.t(1:tend*fs), uv.Unv(1:tend*fs));
hold off;
xlabel('Time [s]'); ylabel('Voltage [V]');
title('Newport Tilt - Vertical Direction');
ax4 = subplot(2, 2, 4);
hold on;
plot(uv.t(1:tend*fs), 1e9*uv.Va(1:tend*fs));
hold off;
xlabel('Time [s]'); ylabel('Distance [nm]');
title('Attocube - Vertical Direction');
linkaxes([ax1,ax2,ax3,ax4],'x');
linkaxes([ax1,ax2],'xy');
linkaxes([ax3,ax4],'xy');
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/repeat_time_signals.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:repeat_time_signals
#+CAPTION: Time domain Signals for the repeatability measurement ([[./figs/repeat_time_signals.png][png]], [[./figs/repeat_time_signals.pdf][pdf]])
[[file:figs/repeat_time_signals.png]]
** Verify Tracking Angle Error
Let's verify that the positioning error of the beam is small and what could be the effect on the distance measured by the intereferometer.
#+begin_src matlab
load('./mat/plant.mat', 'Gd');
#+end_src
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
hold on;
plot(uh.t(1:2*fs), 1e6*uh.Vph(1:2*fs)/freqresp(Gd(1,1), 0), 'DisplayName', '$\theta_{h}$');
plot(uh.t(1:2*fs), 1e6*uh.Vpv(1:2*fs)/freqresp(Gd(2,2), 0), 'DisplayName', '$\theta_{v}$');
hold off;
xlabel('Time [s]'); ylabel('$\theta$ [$\mu$ rad]');
title('Newport Tilt - Horizontal Direction');
legend();
ax2 = subplot(1, 2, 2);
hold on;
plot(uv.t(1:2*fs), 1e6*uv.Vph(1:2*fs)/freqresp(Gd(1,1), 0), 'DisplayName', '$\theta_{h}$');
plot(uv.t(1:2*fs), 1e6*uv.Vpv(1:2*fs)/freqresp(Gd(2,2), 0), 'DisplayName', '$\theta_{v}$');
hold off;
xlabel('Time [s]'); ylabel('$\theta$ [$\mu$ rad]');
title('Newport Tilt - Vertical Direction');
legend();
linkaxes([ax1,ax2],'xy');
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/repeat_tracking_errors.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:repeat_tracking_errors
#+CAPTION: Tracking errors during the repeatability measurement ([[./figs/repeat_tracking_errors.png][png]], [[./figs/repeat_tracking_errors.pdf][pdf]])
[[file:figs/repeat_tracking_errors.png]]
Let's compute the PSD of the error to see the frequency content.
#+begin_src matlab
[psd_UhRh, f] = pwelch(uh.Vph/freqresp(Gd(1,1), 0), hanning(ceil(1*fs)), [], [], fs);
[psd_UhRv, ~] = pwelch(uh.Vpv/freqresp(Gd(2,2), 0), hanning(ceil(1*fs)), [], [], fs);
[psd_UvRh, ~] = pwelch(uv.Vph/freqresp(Gd(1,1), 0), hanning(ceil(1*fs)), [], [], fs);
[psd_UvRv, ~] = pwelch(uv.Vpv/freqresp(Gd(2,2), 0), hanning(ceil(1*fs)), [], [], fs);
#+end_src
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
hold on;
plot(f, sqrt(psd_UhRh), 'DisplayName', '$\Gamma_{\theta_h}$');
plot(f, sqrt(psd_UhRv), 'DisplayName', '$\Gamma_{\theta_v}$');
hold off;
set(gca, 'xscale', 'log'); set(gca, 'yscale', 'log');
xlabel('Frequency [Hz]'); ylabel('ASD $\left[\frac{rad}{\sqrt{Hz}}\right]$')
legend('Location', 'southwest');
title('Newport Tilt - Horizontal Direction');
ax2 = subplot(1, 2, 2);
hold on;
plot(f, sqrt(psd_UvRh), 'DisplayName', '$\Gamma_{\theta_h}$');
plot(f, sqrt(psd_UvRv), 'DisplayName', '$\Gamma_{\theta_v}$');
hold off;
set(gca, 'xscale', 'log'); set(gca, 'yscale', 'log');
xlabel('Frequency [Hz]'); ylabel('ASD $\left[\frac{rad}{\sqrt{Hz}}\right]$')
legend('Location', 'southwest');
title('Newport Tilt - Vertical Direction');
linkaxes([ax1,ax2],'xy');
xlim([1, 1000]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/psd_tracking_error_rad.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:psd_tracking_error_rad
#+CAPTION: Power Spectral Density of the tracking errors ([[./figs/psd_tracking_error_rad.png][png]], [[./figs/psd_tracking_error_rad.pdf][pdf]])
[[file:figs/psd_tracking_error_rad.png]]
Let's convert that to errors in distance
\[ \Delta L = L^\prime - L = \frac{L}{\cos(\alpha)} - L \approx \frac{L \alpha^2}{2} \]
with
- $L$ is the nominal distance traveled by the beam
- $L^\prime$ is the distance traveled by the beam with an angle error
- $\alpha$ is the angle error
#+begin_src matlab
L = 0.1; % [m]
#+end_src
#+begin_src matlab
[psd_UhLh, f] = pwelch(0.5*L*(uh.Vph/freqresp(Gd(1,1), 0)).^2, hanning(ceil(1*fs)), [], [], fs);
[psd_UhLv, ~] = pwelch(0.5*L*(uh.Vpv/freqresp(Gd(2,2), 0)).^2, hanning(ceil(1*fs)), [], [], fs);
[psd_UvLh, ~] = pwelch(0.5*L*(uv.Vph/freqresp(Gd(1,1), 0)).^2, hanning(ceil(1*fs)), [], [], fs);
[psd_UvLv, ~] = pwelch(0.5*L*(uv.Vpv/freqresp(Gd(2,2), 0)).^2, hanning(ceil(1*fs)), [], [], fs);
#+end_src
Now, compare that with the PSD of the measured distance by the interferometer (Fig. [[fig:compare_tracking_error_attocube_meas]]).
#+begin_src matlab
[psd_Lh, f] = pwelch(uh.Va, hanning(ceil(1*fs)), [], [], fs);
[psd_Lv, ~] = pwelch(uv.Va, hanning(ceil(1*fs)), [], [], fs);
#+end_src
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
hold on;
plot(f, sqrt(psd_UhLh), 'DisplayName', '$\Gamma_{L_h}$');
plot(f, sqrt(psd_UhLv), 'DisplayName', '$\Gamma_{L_v}$');
plot(f, sqrt(psd_Lh), '--k', 'DisplayName', '$\Gamma_{L_h}$');
hold off;
set(gca, 'xscale', 'log'); set(gca, 'yscale', 'log');
xlabel('Frequency [Hz]'); ylabel('ASD $\left[\frac{m}{\sqrt{Hz}}\right]$')
legend('Location', 'southwest');
title('Newport Tilt - Horizontal Direction');
ax2 = subplot(1, 2, 2);
hold on;
plot(f, sqrt(psd_UvLh), 'DisplayName', '$\Gamma_{L_h}$');
plot(f, sqrt(psd_UvLv), 'DisplayName', '$\Gamma_{L_v}$');
plot(f, sqrt(psd_Lv), '--k', 'DisplayName', '$\Gamma_{L_h}$');
hold off;
set(gca, 'xscale', 'log'); set(gca, 'yscale', 'log');
xlabel('Frequency [Hz]'); ylabel('ASD $\left[\frac{m}{\sqrt{Hz}}\right]$')
legend('Location', 'southwest');
title('Newport Tilt - Vertical Direction');
linkaxes([ax1,ax2],'xy');
xlim([1, 1000]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/compare_tracking_error_attocube_meas.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:compare_tracking_error_attocube_meas
#+CAPTION: Comparison of the effect of tracking error on the measured distance and the measured distance by the Attocube ([[./figs/compare_tracking_error_attocube_meas.png][png]], [[./figs/compare_tracking_error_attocube_meas.pdf][pdf]])
[[file:figs/compare_tracking_error_attocube_meas.png]]
#+begin_important
The tracking errors are a limiting factor.
#+end_important
** Processing
First, we get the mean value as measured by the interferometer for each value of the Newport angle.
#+begin_src matlab
Vahm = mean(reshape(uh.Va, [fs floor(length(uh.t)/fs)]),2);
Unhm = mean(reshape(uh.Unh, [fs floor(length(uh.t)/fs)]),2);
Vavm = mean(reshape(uv.Va, [fs floor(length(uv.t)/fs)]),2);
Unvm = mean(reshape(uv.Unv, [fs floor(length(uv.t)/fs)]),2);
#+end_src
And we can compute the RMS value of the non-repeatable part:
#+begin_src matlab :exports results :results value table replace :tangle no :post addhdr(*this*)
data2orgtable([rms(1e9*(uh.Va - repmat(Vahm, length(uh.t)/length(Vahm),1))), rms(1e9*(uv.Va - repmat(Vavm, length(uv.t)/length(Vavm),1)))], {}, {'Va - Horizontal [nm rms]', 'Va - Vertical [nm rms]'}, ' %.1f ');
#+end_src
#+RESULTS:
| Va - Horizontal [nm rms] | Va - Vertical [nm rms] |
|--------------------------+------------------------|
| 19.6 | 13.9 |
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
hold on;
plot(uh.Unh, uh.Vaf);
plot(Unhm, Vahm)
hold off;
xlabel('$V_{n,h}$ [V]'); ylabel('$V_a$ [m]');
ax2 = subplot(1, 2, 2);
hold on;
plot(uv.Unv, uv.Vaf);
plot(Unvm, Vavm)
hold off;
xlabel('$V_{n,v}$ [V]'); ylabel('$V_a$ [m]');
linkaxes([ax1,ax2],'xy');
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/repeat_plot_raw.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:repeat_plot_raw
#+CAPTION: Repeatability of the measurement ([[./figs/repeat_plot_raw.png][png]], [[./figs/repeat_plot_raw.pdf][pdf]])
[[file:figs/repeat_plot_raw.png]]
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
hold on;
plot(uh.Unh, 1e9*(uh.Vaf - repmat(Vahm, length(uh.t)/length(Vahm),1)));
hold off;
xlabel('$V_{n,h}$ [V]'); ylabel('$V_a$ [nm]');
ax2 = subplot(1, 2, 2);
hold on;
plot(uv.Unv, 1e9*(uv.Vaf - repmat(Vavm, length(uv.t)/length(Vavm),1)));
hold off;
xlabel('$V_{n,v}$ [V]'); ylabel('$V_a$ [nm]');
linkaxes([ax1,ax2],'xy');
ylim([-100 100]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/repeat_plot_subtract_mean.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:repeat_plot_subtract_mean
#+CAPTION: Repeatability of the measurement after subtracting the mean value ([[./figs/repeat_plot_subtract_mean.png][png]], [[./figs/repeat_plot_subtract_mean.pdf][pdf]])
[[file:figs/repeat_plot_subtract_mean.png]]
** Analysis of the non-repeatable contributions
Let's know try to determine where does the non-repeatability comes from.
From the 4QD signal, we can compute the angle error of the beam and thus determine the corresponding displacement measured by the attocube.
We then take the non-repeatable part of this displacement and we compare that with the total non-repeatability.
We also plot the displacement measured during the huddle test.
All the signals are shown on Fig. [[fig:non-repeatability-parts]].
#+begin_src matlab :exports none
Vphm = mean(reshape(uh.Vph/freqresp(Gd(1,1), 0), [fs floor(length(uh.t)/fs)]),2);
Unhm = mean(reshape(uh.Unh, [fs floor(length(uh.t)/fs)]),2);
Vpvm = mean(reshape(uv.Vpv/freqresp(Gd(2,2), 0), [fs floor(length(uv.t)/fs)]),2);
Unvm = mean(reshape(uv.Unv, [fs floor(length(uv.t)/fs)]),2);
#+end_src
#+begin_src matlab :exports none
% =Vaheq= is the equivalent measurement error in [m] due to error angle of the Sercalo.
Vaheq = uh.Vph/freqresp(Gd(1,1), 0) - repmat(Vphm, length(uh.t)/length(Vphm),1);
Vaveq = uv.Vpv/freqresp(Gd(2,2), 0) - repmat(Vpvm, length(uv.t)/length(Vpvm),1);
Vaheq = sign(Vaheq).*Vaheq.^2;
Vaveq = sign(Vaveq).*Vaveq.^2;
#+end_src
#+begin_src matlab :exports none
ht = load('./mat/data_huddle_test_3.mat', 't', 'Va');
#+end_src
#+begin_src matlab :exports none
htm = 1e9*ht.Va(1:length(Vaheq)) - repmat(mean(1e9*ht.Va(1:length(Vaheq))), length(uh.t)/length(Vaheq),1);
#+end_src
#+begin_src matlab :exports none
figure;
hold on;
plot(uh.Unh, 1e9*(uh.Va - repmat(Vahm, length(uh.t)/length(Vahm),1)), 'DisplayName', 'Measured Non-Repeatability');
plot(uh.Unh, 1e9*ht.Va(1:length(Vaheq))-mean(1e9*ht.Va(1:length(Vaheq))), 'DisplayName', 'Huddle Test');
plot(uh.Unh, 1e9*Vaheq, 'DisplayName', 'Due to Sercalo Angle Error');
hold off;
xlabel('$V_{n,h}$ [V]'); ylabel('$V_a$ [nm]');
ylim([-100 100]);
legend();
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/non-repeatability-parts-half.pdf" :var figsize="normal-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
hold on;
plot(uh.Unh, 1e9*(uh.Va - repmat(Vahm, length(uh.t)/length(Vahm),1)));
plot(uh.Unh, 1e9*ht.Va(1:length(Vaheq))-mean(1e9*ht.Va(1:length(Vaheq))));
plot(uh.Unh, 1e9*Vaheq);
hold off;
xlabel('$V_{n,h}$ [V]'); ylabel('$V_a$ [nm]');
ax2 = subplot(1, 2, 2);
hold on;
plot(uv.Unv, 1e9*(uv.Va - repmat(Vavm, length(uv.t)/length(Vavm),1)), 'DisplayName', 'Measured Non-Repeatability');
plot(uv.Unv, 1e9*ht.Va(1:length(Vaveq))-mean(1e9*ht.Va(1:length(Vaveq))), 'DisplayName', 'Huddle Test');
plot(uv.Unv, 1e9*Vaveq, 'DisplayName', 'Due to Sercalo Angle Error');
hold off;
xlabel('$V_{n,v}$ [V]'); ylabel('$V_a$ [nm]');
legend('location', 'northeast');
linkaxes([ax1,ax2],'xy');
ylim([-100 100]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/non-repeatability-parts.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:non-repeatability-parts
#+CAPTION: Non repeatabilities ([[./figs/non-repeatability-parts.png][png]], [[./figs/non-repeatability-parts.pdf][pdf]])
[[file:figs/non-repeatability-parts.png]]
** Results with a low pass filter
We filter the data with a first order low pass filter with a crossover frequency of $\omega_0$.
#+begin_src matlab
w0 = 10; % [Hz]
G_lpf = 1/(1 + s/2/pi/w0);
uh.Vaf = lsim(G_lpf, uh.Va, uh.t);
uv.Vaf = lsim(G_lpf, uv.Va, uv.t);
#+end_src
** Processing
First, we get the mean value as measured by the interferometer for each value of the Newport angle.
#+begin_src matlab
Vahm = mean(reshape(uh.Vaf, [fs floor(length(uh.t)/fs)]),2);
Unhm = mean(reshape(uh.Unh, [fs floor(length(uh.t)/fs)]),2);
Vavm = mean(reshape(uv.Vaf, [fs floor(length(uv.t)/fs)]),2);
Unvm = mean(reshape(uv.Unv, [fs floor(length(uv.t)/fs)]),2);
#+end_src
And we can compute the RMS value of the non-repeatable part:
#+begin_src matlab :exports results :results value table replace :tangle no :post addhdr(*this*)
data2orgtable([rms(1e9*(uh.Vaf - repmat(Vahm, length(uh.t)/length(Vahm),1))), rms(1e9*(uv.Vaf - repmat(Vavm, length(uv.t)/length(Vavm),1)))], {}, {'Va - Horizontal [nm rms]', 'Va - Vertical [nm rms]'}, ' %.1f ');
#+end_src
#+RESULTS:
| Va - Horizontal [nm rms] | Va - Vertical [nm rms] |
|--------------------------+------------------------|
| 22.9 | 13.9 |
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
hold on;
plot(uh.Unh, uh.Vaf);
plot(Unhm, Vahm)
hold off;
xlabel('$V_{n,h}$ [V]'); ylabel('$V_a$ [m]');
ax2 = subplot(1, 2, 2);
hold on;
plot(uv.Unv, uv.Vaf);
plot(Unvm, Vavm)
hold off;
xlabel('$V_{n,v}$ [V]'); ylabel('$V_a$ [m]');
linkaxes([ax1,ax2],'xy');
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/repeat_plot_lpf.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:repeat_plot_raw
#+CAPTION: Repeatability of the measurement ([[./figs/repeat_plot_lpf.png][png]], [[./figs/repeat_plot_lpf.pdf][pdf]])
[[file:figs/repeat_plot_lpf.png]]
#+begin_src matlab :exports none
figure;
ax1 = subplot(1, 2, 1);
hold on;
plot(uh.Unh, 1e9*(uh.Vaf - repmat(Vahm, length(uh.t)/length(Vahm),1)));
hold off;
xlabel('$V_{n,h}$ [V]'); ylabel('$V_a$ [nm]');
ax2 = subplot(1, 2, 2);
hold on;
plot(uv.Unv, 1e9*(uv.Vaf - repmat(Vavm, length(uv.t)/length(Vavm),1)));
hold off;
xlabel('$V_{n,v}$ [V]'); ylabel('$V_a$ [nm]');
linkaxes([ax1,ax2],'xy');
ylim([-60 60]);
#+end_src
#+HEADER: :tangle no :exports results :results none :noweb yes
#+begin_src matlab :var filepath="figs/repeat_plot_subtract_mean_lpf.pdf" :var figsize="full-tall" :post pdf2svg(file=*this*, ext="png")
<>
#+end_src
#+NAME: fig:repeat_plot_subtract_mean_lpf
#+CAPTION: Repeatability of the measurement after subtracting the mean value ([[./figs/repeat_plot_subtract_mean_lpf.png][png]], [[./figs/repeat_plot_subtract_mean_lpf.pdf][pdf]])
[[file:figs/repeat_plot_subtract_mean_lpf.png]]