Background:
Data aquisition system is fixed, and now we can use the Dataviewer to measure Q-values of the ringdowns for each DOF, each optics.
First of all, I measured MC1 suspention damping servo for a test.
What I did:
1. Used DAQ channels activated in this entry(#3690) to see and compare the ringdowns when the damping servo is on and off with the Dataviewer.
2. Plotted the data and fitted the ringdown using this formula;
p[0]*exp(-p[1]*t)*sin(p[2]*t+p[3])+p[4]
I used python's scipy.optimize.leastsq for the fitting.
3. Calculated the resonant frequency f0 and Q-value using following formulas;
f0=2*pi*sqrt(p[1]**2+p[2]**2)
Q=f0/(2*pi)/(2*p[1])
4. For plotting, I subtracted the offset(=p[4]).
All parameters I used for this measurement are automatically saved here;
/cvs/cds/caltech/burt/autoburt/snapshots/2010/Oct/12/13:07/c1mcs.epics
(-1,0,1 for all matrix elements, GAIN=3,3,3,150 for POS,PIT,YAW,SIDE)
Result:
Attached is the plot of each 4 DOF ringdown when servo is off and on.
"servo off" means off for that DOF. Servo for the other 3 DOFs are on.
As you can see clearly, the damping servo is working.
The resonant frequencies and Q-values calculated from the fitting are as follows;
|
servo off |
servo on |
f0 (Hz) |
Q |
f0 (Hz) |
Q |
POS |
0.97 |
large |
0.97 |
16 |
PIT |
0.71 |
96 |
0.73 |
6.9 |
YAW |
0.80 |
100 |
0.82 |
8.9 |
SIDE |
0.99 |
large |
0.99 |
27 |
Resonant frequencies and Q-values have about 1% and 10% error respectively.
I estimated it from my 2-time measurement of the POS ringdown.
Next work:
- Find and modify some scripts to optimize the matrix elements
- Calibrate the displacement
- Do the same thing for other optics
|