15
Istituto Nazionale di Fisica Nucleare Sezione di Pisa Largo Bruno Pontecorvo, 56127 Pisa – Italy. http://www.pi.infn.it VIRGO Collaboration Actuation Noise in Virgo+ and Advanced Virgo Document Number: VIR-0146A-10 Issue: 1.0 Date: 05/02/2010 Author: Alberto Gennai

VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

  • Upload
    others

  • View
    3

  • Download
    0

Embed Size (px)

Citation preview

Page 1: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Istituto Nazionale di Fisica Nucleare

Sezione di Pisa

Largo Bruno Pontecorvo, 56127 Pisa – Italy. http://www.pi.infn.it

VIRGO Collaboration

Actuation Noise in Virgo+ and Advanced Virgo

Document Number: VIR-0146A-10

Issue: 1.0

Date: 05/02/2010

Author: Alberto Gennai

Page 2: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146A-10 Issue: 1.0 Date: 05/02/2010 Page 2 of 15

Change Record

Issue Date Affected Paragraphs Reason/Remarks Author

1.0 05/02/2010 All First issue A. Gennai

Table Of Contents

1. INTRODUCTION ................................................................................................................. 3

1.1 PURPOSE & SCOPE ............................................................................................................ 3 1.2 ACRONYMS ...................................................................................................................... 3 1.3 REFERENCES .................................................................................................................... 3

2. VIRGO COIL DRIVER ........................................................................................................ 4

2.1 OVERVIEW ....................................................................................................................... 4 2.2 COIL DRIVER NOISE ......................................................................................................... 4

2.3 DAC NOISE ...................................................................................................................... 7 2.3.1 Old Coil Driver ............................................................................................................ 8

3. SIGNALS DYNAMIC ........................................................................................................... 9

3.1 REFERENCE MASS – MIRROR ACTUATORS ........................................................................ 9 3.2 FILTER #7 - MARIONETTE ............................................................................................... 11

4. PAYLOAD TRANSFER FUNCTIONS ............................................................................. 12

4.1 REFERENCE MASS – MIRROR .......................................................................................... 12

4.2 FILTER #7 – MARIONETTE ............................................................................................... 13

5. NOISE BUDGET ................................................................................................................ 13

5.1 VSR2 NOISE BUDGET ..................................................................................................... 13

5.2 VIRGO+ MS NOISE BUDGET ........................................................................................... 14

5.3 ADVANCED VIRGO ......................................................................................................... 14

Page 3: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146A-10 Issue: 1.0 Date: 05/02/2010 Page 3 of 15

1. INTRODUCTION

1.1 Purpose & Scope This document presents the noise budget for lower stage actuators for Virgo Science Run no. 2

(VSR2), Virgo+ with monolithic payloads (Virgo+MS) and Advanced Virgo (AdvV) . Main

purpose is the preparation of the Virgo+MS total noise budget.

The content of this document was presented to the Virgo Collaboration at the Advanced Virgo

meeting held in Cascina on February 4th, 2010 and at the Commissioning/Detector Meeting on

February 8th, 2010.

1.2 Acronyms This document contains several abbreviations and acronyms to refer concisely to an item after it has

been introduced. The following list is aimed to help the reader in recalling the extended meaning of

each short expression.

AdvV Advanced Virgo

CD Coil Driver

DAC Digital to Analog Converter

DSP Digital Signal Processor

RD Reference Document

Virgo+MS Virgo+ with Monolithic payloads

VSR2 Virgo Science Run No. 2

1.3 References This report refers to the following documents containing background or detailed information that

can be useful for the reader.

[RD1] A.Gennai, Actuators Noise in AdvV, VIR-0093A-10 Presentation to the AdvV Meeting,

February 4th, 2010

[RD2] M.Bitossi, A.Gennai, D. Passuello, Low Noise DAC Selection for Advanced Virgo, VIR-

0078A-10, January 20th, 2010

A. Gennai, DAC Noise Effect on Virgo Sensitivity, VIR-0006B-09, October 2nd, 2009

[RD3] Gennai, DAC Noise Contribution to Virgo Sensitivity, VIR-0072A-08, June 6th, 2008

[RD4] Gennai, New Coil Driver Measurements, VIR-0010C-08, March 23rd, 2008

[RD5] Gennai, Coil Driver Final Design Document, VIR-SPE-PIS-4900-121, September 30th, 2004

Page 4: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and Advanced Virgo

2. VIRGO COIL DRIVER

2.1 Overview In Virgo, mirrors actuation along laser beam direction is performed using 6 electromagnetic

actuators. Two horizontal actuators ac

Filter#7) while 4 actuators act directly on mirror from recoil mass (also known as ‘reference’ mass).

Starting from position error, the Suspension Control System computes forces

Processors (DSP).

In a magnet-coil actuator, force is proportional to the current flowing in the coil. The Coil Driver

(CD) is the electronic device that converts the voltage at the output of the Digital to Analog

Converter (DAC) into a current flowing int

Figure

New Coil Drivers, currently installed at terminal north and west suspensions, have a voltage to

current conversion factor that is remotely selectable with a smooth transition from one operation

mode to the next one. Five are the available operational modes: HIGHPOWER, used during cavities

lock acquisition phase, and LOWNOISE1 to LOWNOISE4 used during linear regime. Fur

details on coil drivers and their operational modes can be found in

2.2 Coil Driver Noise Coil Driver noise can be evaluated using an accurate PSPICE simulation and

measured values (see [RD4]). The f

simulation. Three distinct blocks put in evidence the input differential stage, the noise shaping filter

and the series resistor. Changing operating mode we simply change the value of the series resi

Monitoring network is reported only to keep into account of its noise contribution.

shown corresponds to the second order filter (two poles

coil driver. Actual design foresees the possibility

filter) to 5th.

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146AIssue: 1.0 Date: 05/02/2010 Page 4 of 15

VIRGO COIL DRIVER

In Virgo, mirrors actuation along laser beam direction is performed using 6 electromagnetic

actuators. Two horizontal actuators act on Marionette from the Steering Filter (also known as

Filter#7) while 4 actuators act directly on mirror from recoil mass (also known as ‘reference’ mass).

Starting from position error, the Suspension Control System computes forces using Digital Signal

coil actuator, force is proportional to the current flowing in the coil. The Coil Driver

(CD) is the electronic device that converts the voltage at the output of the Digital to Analog

Converter (DAC) into a current flowing into the coil.

Figure 1 Sketch of the control chain

New Coil Drivers, currently installed at terminal north and west suspensions, have a voltage to

factor that is remotely selectable with a smooth transition from one operation

mode to the next one. Five are the available operational modes: HIGHPOWER, used during cavities

lock acquisition phase, and LOWNOISE1 to LOWNOISE4 used during linear regime. Fur

details on coil drivers and their operational modes can be found in [RD4] and [RD5]

Coil Driver noise can be evaluated using an accurate PSPICE simulation and a comparison

. The following picture shows the coil driver schematic used in

simulation. Three distinct blocks put in evidence the input differential stage, the noise shaping filter

and the series resistor. Changing operating mode we simply change the value of the series resi

Monitoring network is reported only to keep into account of its noise contribution.

shown corresponds to the second order filter (two poles – two zeros) we are using in the installed

coil driver. Actual design foresees the possibility to implement filter with order ranging from 0 (no

0146A-10

In Virgo, mirrors actuation along laser beam direction is performed using 6 electromagnetic

t on Marionette from the Steering Filter (also known as

Filter#7) while 4 actuators act directly on mirror from recoil mass (also known as ‘reference’ mass).

using Digital Signal

coil actuator, force is proportional to the current flowing in the coil. The Coil Driver

(CD) is the electronic device that converts the voltage at the output of the Digital to Analog

New Coil Drivers, currently installed at terminal north and west suspensions, have a voltage to

factor that is remotely selectable with a smooth transition from one operation

mode to the next one. Five are the available operational modes: HIGHPOWER, used during cavities

lock acquisition phase, and LOWNOISE1 to LOWNOISE4 used during linear regime. Further

[RD5].

a comparison with

ollowing picture shows the coil driver schematic used in

simulation. Three distinct blocks put in evidence the input differential stage, the noise shaping filter

and the series resistor. Changing operating mode we simply change the value of the series resistor.

Monitoring network is reported only to keep into account of its noise contribution. Shaping filter

two zeros) we are using in the installed

to implement filter with order ranging from 0 (no

Page 5: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and Advanced Virgo

Noise refereed to the output of the coil driver, expressed in Ampere/sqrt(Hz) is of course a function

of the operating mode since changing operating mode we change the voltage to current conversion

factor.

Figure

Due to the second order low pass filter, referring the Coil Driver noise to its input, we obtain the

following picture where at frequencies bigger than about 10 Hz noise is ‘amplified’ by the filter

C10

1n

0

R2122k

C7

1n

U3B

OPA2604E/BB

+5

-6

V+8

V-4

OUT7

+15

ilnmoni

-15

R20

47k

R17

47k

U12

AD8671

+1

-2

V+3

V-4

OUT5

U13

AD8671

+1

-2

V+3

V-4

OUT5

DIFFERENTIAL INPUT STAGE

-15

R1

1k

R10

1k

R2

560

0

R11

560R13100k

OP-27A/AD

+3

-2

C110n

C510n

C410n

0

R14100k

VIN11Vac0Vdc

+15

+15

-15

V115Vdc

0V2

15Vdc

+15

-15

U11A

OPA2604E/BB

OUT1

-15

ihpmoni

R35

47k

0

R2322k

C8

1n

U7A

OPA2604E/BB

+3

-2

V+8

V-4

OUT1

+15

-15

R22

47k

R19

1.6k

C6

5u

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146AIssue: 1.0 Date: 05/02/2010 Page 5 of 15

Figure 2 Coil Driver Schematic

Noise refereed to the output of the coil driver, expressed in Ampere/sqrt(Hz) is of course a function

changing operating mode we change the voltage to current conversion

Figure 3 Coil Driver Output Noise

Due to the second order low pass filter, referring the Coil Driver noise to its input, we obtain the

following picture where at frequencies bigger than about 10 Hz noise is ‘amplified’ by the filter

R9

{rval}

R3622k

0

U14

AD8671

+1

-2

V+3

V-4

OUT5

R26

100k

R27100k

U15

AD8671

+1

-2

V+3

V-4

OUT5

0

U10

AD743J/AD

+3

-2

V+7

V-4

OUT6

N11

N25

+15

-15

vmoni

R300.1

0

R12

560

R3

560

0

U7B

OPA2604E/BB

+5

-6

V+8

V-4

OUT7

R4

33k

+15

R53.6k

-15

0

C24.7u

+15

-15

U2

OP-27A/AD

V+7

V-4

OUT6

N11

N28

-15

R16

560

R25

560

R24

560

R15

560

0

+15

-15

U8

OP-27A/AD

+3

-2

V+7

V-4

OUT6

N11

N28

+15

+15

-15

R28

1k

R31

1k

R3250k

0

R34

50k

U3A

OPA2604E/BB

+3

-2

V+8

V-4

OUT1

U11A

OPA2604E/BB

+3

-2

V+8

V-4

OUT

+15

-15

+15

R35

47k

-15

SHAPING FILTER

R29

47k

R6

33k R73.6k

0

C34.7u

VOLTAGE AND CURRENT MONITORING

R33

1.6k

U4

AD743J/AD

+3

-2

V+7

V-4

OUT6

N11

N25

C9

5u

+15

-15

SERIES RESISTOR

R18

47kR19

1.6k

R8

{rval}

PARAMETERS:rval = 150

LN1 --> rval = 150LN2 --> rval = 600

LN3 --> rval = 1200LN4 --> rval = 2400

0146A-10

Noise refereed to the output of the coil driver, expressed in Ampere/sqrt(Hz) is of course a function

changing operating mode we change the voltage to current conversion

Due to the second order low pass filter, referring the Coil Driver noise to its input, we obtain the

following picture where at frequencies bigger than about 10 Hz noise is ‘amplified’ by the filter

RCOIL110

+-

H1

H

LCOIL13mH

1

2

icoil

Rdummy11

0

COIL (cabling included)

VOLTAGE AND CURRENT MONITORING

Page 6: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and Advanced Virgo

itself. We should in fact notice that voltage gain (from

output) is lower or equal to one and therefore input stage noise is not dominant in respect with

shaping filter noise.

For better understanding, we can comp

order filter.

Figure 5 Low Noise 1 Output Noise: Comparison between

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146AIssue: 1.0 Date: 05/02/2010 Page 6 of 15

itself. We should in fact notice that voltage gain (from differential stage input to shaping filter

and therefore input stage noise is not dominant in respect with

Figure 4 Input Referred Noise

For better understanding, we can compare output and input noise using a first order and a second

Low Noise 1 Output Noise: Comparison between 1st order and 2nd order shaping filter

0146A-10

differential stage input to shaping filter

and therefore input stage noise is not dominant in respect with

are output and input noise using a first order and a second

1st order and 2nd order shaping filter

Page 7: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and Advanced Virgo

Figure 6 Low Noise 1 Input Referred Noise: Comparison between 1st order and 2nd order shaping filter. High frequency (1 kHz) value is about 270 nV/sqrt(Hz). Completely removing shaping filters that value drops down to

about 12 nV/sqrt(Hz) while low frequency (below 1 Hz)

2.3 DAC Noise Once coil driver noise is known we can evaluate the contribution of DAC noise both at CD input

and output. In the following two plots we used the DAC noise measured as described in

and [RD3]. With filters in use during VSR2 (where only LOWNOISE1 was used) DAC

contribution equals CD contribution betw

Figure 7 Input Referred Noise superposed to DAC noise estimate

1 Removing the shaping filter block, we gain a factor ~10 due to filter transfer function. Total noise is furthermore reduced by the amount of noise introduced by the shaping filter operational amplifiers and resistors.

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146AIssue: 1.0 Date: 05/02/2010 Page 7 of 15

Noise: Comparison between 1st order and 2nd order shaping filter. High

frequency (1 kHz) value is about 270 nV/sqrt(Hz). Completely removing shaping filters that value drops down to about 12 nV/sqrt(Hz) while low frequency (below 1 Hz) behaviour remains unchanged

Once coil driver noise is known we can evaluate the contribution of DAC noise both at CD input

and output. In the following two plots we used the DAC noise measured as described in

. With filters in use during VSR2 (where only LOWNOISE1 was used) DAC

contribution equals CD contribution between 20 and 30 Hz.

Input Referred Noise superposed to DAC noise estimate

Removing the shaping filter block, we gain a factor ~10 due to filter transfer function. Total noise is furthermore reduced by the amount of noise introduced by the shaping filter operational amplifiers and resistors.

0146A-10

Noise: Comparison between 1st order and 2nd order shaping filter. High frequency (1 kHz) value is about 270 nV/sqrt(Hz). Completely removing shaping filters that value drops down to

nchanged1.

Once coil driver noise is known we can evaluate the contribution of DAC noise both at CD input

and output. In the following two plots we used the DAC noise measured as described in [RD2]A

. With filters in use during VSR2 (where only LOWNOISE1 was used) DAC

Removing the shaping filter block, we gain a factor ~10 due to filter transfer function. Total noise is furthermore

Page 8: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and Advanced Virgo

Figure 8 Output Noise: Coil Driver and DAC Contribution

2.3.1 Old Coil Driver It is worth reminding here that during VSR2 we used old coil drivers for Filter #7

actuators. Behaviour is different from the one described up to now and applicable to new coil

drivers (acting in VSR2 on mirror only).

nV/sqrt(Hz), almost flat above 1 Hz. At its input there is a first order shaping filter (pole at 0.9 Hz,

zero at 9 Hz). Output noise, including DAC noise contribution,

Figure 9 Old Coil Driver Output Noise (DAC contribution included)

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146AIssue: 1.0 Date: 05/02/2010 Page 8 of 15

Output Noise: Coil Driver and DAC Contribution

during VSR2 we used old coil drivers for Filter #7

Behaviour is different from the one described up to now and applicable to new coil

drivers (acting in VSR2 on mirror only). Voltage gain is 2 while input referred noise is about 10

nV/sqrt(Hz), almost flat above 1 Hz. At its input there is a first order shaping filter (pole at 0.9 Hz,

, including DAC noise contribution, is shown in the following picture.

Old Coil Driver Output Noise (DAC contribution included)

0146A-10

during VSR2 we used old coil drivers for Filter #7 – Marionette

Behaviour is different from the one described up to now and applicable to new coil

Voltage gain is 2 while input referred noise is about 100

nV/sqrt(Hz), almost flat above 1 Hz. At its input there is a first order shaping filter (pole at 0.9 Hz,

is shown in the following picture.

Page 9: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146A-10 Issue: 1.0 Date: 05/02/2010 Page 9 of 15

3. SIGNALS DYNAMIC Once the noise is known we have to decide what should be the equivalent displacement. Adding a

series resistor to the coil we can reduce the noise equivalent displacement but of course in this way

we also limit the maximum displacement we can achieve. A correct evaluation of required

dynamical range is therefore important for selecting a proper operational mode. For this purpose we

analyzed the first 231 science mode segments during VSR2 (almost all segments). A Science Mode

segment is the time interval in which Virgo interferometer is acquiring scientific data and

corresponds to more than 80% of the total time. Each Science Mode segment was than split into 30

minutes sub-segments and for each segment we computed mean value, standard deviation and

absolute maximum. Analysis was performed only for end mirrors since input mirrors are not driven

along the beam direction and other mirrors have a negligible contribution to total noise.

3.1 Reference Mass – Mirror Actuators

Page 10: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146A-10 Issue: 1.0 Date: 05/02/2010 Page 10 of 15

As we can clearly see from last two plots, we used only a small fraction of available dynamical

range. We could in fact operate in LOWNOISE2 mode (4x series resistor) for about 95% of science

mode time and in LOWNOISE3 mode (8x series resistor) for more than 80% of science mode time.

Figure 10 Reference Mass - Mirror Actuators. The two plotted lines correspond to two different period of time

Page 11: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146A-10 Issue: 1.0 Date: 05/02/2010 Page 11 of 15

3.2 Filter #7 - Marionette

In this case we can see that we could operate for 99% of science mode time inserting a 5x series

resistor.

Page 12: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146A-10 Issue: 1.0 Date: 05/02/2010 Page 12 of 15

Figure 11 Filter#7-Marionette Actuators

Looking at DAC output spectra and comparing with the spectra of reference mass coils (Figure 10)

we notice that at F7 coils we can have a further improvement increasing low pass filter order (from

1st to 2nd)

4. PAYLOAD TRANSFER FUNCTIONS For a correct evaluation of actuation noise contribution we use a model of the payload transfer

function.

4.1 Reference Mass – Mirror DC displacement is about 5.5 um/A per coil. For example, using two coils as we did in VSR1 and

VSR2 we have 11 um/A DC displacement. Mechanical transfer function is the classical single

pendulum response with 0.6 Hz as resonant factor and a quality factor large enough to be assumed

equal to infinity.

Above resonant frequency (0.6 Hz), mechanical transfer function is therefore

|����� � ��| 5.5 10�� ���

��

� ����

�� � �⁄ (acting with 1 coil)

Page 13: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146A-10 Issue: 1.0 Date: 05/02/2010 Page 13 of 15

4.2 Filter #7 – Marionette Acting from filter #7 onto Marionette, DC displacement is about 1.7 um/VDAC (using a single coil

and old coil drivers) corresponding to 0.85 um/Vcoil or, assuming the coil impedance to have a real

part equal to 16 Ohm, 7 um/A. Mechanical transfer function is equivalent to a double pendulum

response having f1=0.46 Hz and f2=0.98 Hz as resonant frequencies:

|���� � ��| 0.85 10�� �����

��

�. ! ���"

�� m/V (with 1 horizontal coil)

or, equivalently

|���� � ��| 7 10�� �����

��

�.$ ����

�� m/A (with 1 horizontal coil)

5. NOISE BUDGET

5.1 VSR2 Noise Budget In VSR2 we used the following configuration: new coil drivers in LOWNOISE1 mode acting on

two coils at NE and WE. NI and WI mirrors, due to the extremely large value used for series

resistors (6 kOhm), give a negligible contribution to long arms actuator noise.

Coil driver used for Filter #7-Marionette actuators is the old one with a first order shaping filter.

Figure 12 VSR2 Actuators Noise Budget (Long Arms only)

101

102

10-24

10-23

10-22

10-21

10-20

10-19

10-18

VSR2 Actuators Noise Budget

Frequency (Hz)

h [1

/sqr

t(H

z)]

F7-Mar Actuation NoiseRM-Mirr Actuation Noise

Virgo Design

VSR2 Sensitivity

Virgo+MS DesignAdvanced Virgo Design

Page 14: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146A-10 Issue: 1.0 Date: 05/02/2010 Page 14 of 15

5.2 Virgo+ MS Noise Budget After monolithic payloads installation, thermal noise contribution to Virgo sensitivity will drop

down. Meeting requirement for actuation noise is possible using the LOWNOISE2 mode at

reference mass level and adding a 5x series resistor to marionette actuators.

Figure 13 Virgo+ Monolithic Payloads Actuators Noise (Long Arms only)

To be noticed that further improvements are possible implanting second order filter at Filter#7-

Marionette level and switching to LOWNOISE3 mode at reference mass – marionette level.

Furthermore acting on more coils (at present we use only 2 of the 4 coils available) and/or more

mirrors (at present we act only on end mirrors and we do not use input mirrors for cavities

longitudinal lock).

5.3 Advanced Virgo With Advanced Virgo we will be able to take advantage from new DAC. Recent measurements on

evaluation boards show that new DAC performances are much better than old converters (see

[RD2]) with a gain of a factor about 20 at 10 Hz.

101

102

10-24

10-23

10-22

10-21

10-20

10-19

10-18

V+MS Actuators Noise Budget

Frequency (Hz)

h [1

/sqr

t(H

z)]

F7-Mar Actuation NoiseRM-Mirr Actuation Noise

Virgo Design

VSR2 Sensitivity

Virgo+MS DesignAdvanced Virgo Design

Page 15: VIRGO Collaboration - ego-gw.it · 0.1 0 r12 560 r3 560 0 u7b opa2604e/bb + 5-6 v+ 8 v-4 out 7 r4 33k +15 r5 3.6k-15 0 c2 4.7u +15-15 v+u2 7 v-4 out 6 n1 1 n2 8-15 r16 560 r25 560

Actuation Noise in Virgo+ and

Advanced Virgo

Doc Nr: VIR-0146A-10 Issue: 1.0 Date: 05/02/2010 Page 15 of 15

Figure 14 Advanced Virgo Actuators Noise (Long Arms only)

Moving from old DAC to new one we will reduce DAC noise by about a factor 20 at 10 Hz. As it

can be noticed comparing Figure 13 and Figure 14, sensitivity will not improve by the same amount

since DAC noise is not the limiting noise (see Chapter 2.3). The actual gain is at low-pass filter

level where we could guarantee the same performances using a 1st order filter

Increased linearity will allow achieving required performances without dithering (either ‘natural’ or

‘artificial’)

oOo

101

102

10-24

10-23

10-22

10-21

10-20

10-19

10-18

AdV Actuators Noise Budget

Frequency (Hz)

h [1

/sqr

t(H

z)]

F7-Mar Actuation NoiseRM-Mirr Actuation Noise

Virgo Design

VSR2 Sensitivity

Virgo+MS DesignAdvanced Virgo Design