40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  PSL  Not logged in ELOG logo
Entry  Wed Nov 3 01:22:57 2010, tara, Notes, PMC, TF from PMC servo LISO_compare.pngPMC_TF_2.pngPMC_TF.m.zippmc_data.matpmc.fil.zip
    Reply  Wed Nov 3 02:08:06 2010, Koji, Notes, PMC, TF from PMC servo 
       Reply  Wed Nov 3 15:38:05 2010, tara, Notes, PMC, TF from PMC servo mod_TF.pngmisnomber.png
          Reply  Thu Nov 4 04:13:59 2010, Koji, Notes, PMC, TF from PMC servo 
             Reply  Thu Nov 4 21:13:32 2010, tara, Notes, PMC, TF from PMC servo compare_TF1.pngcompare_TF.png
Message ID: 383     Entry time: Thu Nov 4 21:13:32 2010     In reply to: 382
Author: tara 
Type: Notes 
Category: PMC 
Subject: TF from PMC servo 

I got the calibration from [here]

1) DC ext channel on PMC servo: 32.82 MHz/ V

The DC gain between DC ext channel and the voltage at PZT is 27.65 dB (x24.13),

so the Actuator gain will be 32.82/24.13 = 1.36 MHz/ V;

 The plot on fig1 is the Transfer function of the PZT actuator in MHz/ Volt.


The liso plot, [fig1] offset by 30.5 dB, match the result from the measurement.

This means that the gain from AD602 is 30.5 dB, even though the gain slider says 30dB.


Assuming that from DC to 100kHz, the TF from optic is flat.

The OLG TF measurement must equal The TF from servo(From LISO) + gain slider(30.5 dB) + PZT(LISO) + optics(flat offset)

The offset in the plot is 25.5 dB. With the 30.5 dB from gain slider, TF from optics is -5dB flat, with 180 degree phase shift see fig2. [add calibration from Hz -> V] [plot2]

The result from previous entry which gives the optic's TF to be flat at 1 dB is wrong because I did not use the whole TF from the servo

when I compare the model and the measurement, so I missed -6 dB from AD797.





What are the units of the vert axes?

Separate the open loop gain into three part:

- Optical Gain, Unit [V/m] or [V/Hz], usually flat or simple low path shape

- Servo Filter Gain, Unit [V/V], various shape

- Actuator Gain, Unit [m/V] or [Hz/V], flat or low path filter like up to kHz~100kHz (depending on the time constant of the RC filter),
mechanical resonances above that freq region, which usually determin the highest UGF.

You can change the servo gain by modifying the circuit.

You can change the optical gain by changing the amount of the light in the cavity / on the PD as well as changing the cavity finesse etc.

You can change the actuator gain by replacing the actuator.


Sorry for the confusion, PZT actuator is included in the optical TF. 

The plot on fig2 below shows the TF of PZT part, offset by 1 dB to match the misnomer optical path TF.

Thus, the real optical TF is rather flat with magnitude~ 1 dB, the phase shift is 180 degree,

 and the modifiable TF (LISO model) is plot on fig1. This plot has not taken the gain from the slider into account yet.



Why is the optical TF not (kinda) flat?

Why does the PZT actuator completely ignored?

You need to talk to me tomorrow afternoon when I am in ATF.


I determined the OLG TF of the whole PMC loop, and TFs from servo paths and optical path.


We want to modify the PMC servo to optimize the PMC loop, so we have to know what are the TFs from part where we can modify,

and where we can't (optical path).


The whole TF is measured before, but I remeasured again just to make sure that there won't be any problem from the laser.

How I measure the whole TF is [here].


 I measured the OLG TF from the PMC servo. The results agree well with the LISO model, see fig 1.


 Optical TF = Whole TF - Servo TF.

The Optical TF won't be modified. It will be used to compute the whole TF after the PMC servo modification. 

The measurement at low frequency does not look nice because the signal was suppressed by the gain.

But the TF around UGF still looks fine to work with.





Attachment 1: compare_TF1.png  22 kB  Uploaded Fri Nov 5 01:21:37 2010  | Hide | Hide all
Attachment 2: compare_TF.png  28 kB  Uploaded Fri Nov 5 01:23:10 2010  | Hide | Hide all
ELOG V3.1.3-