Transcript Faraday isolation

Faraday isolator on the IB
E. Genin, P. La Penna
EGO
Amsterdam, 03/07/06
Commissioning Meeting
1
 It
was decided to put a Faraday isolator between the IMC
and the PR because the ITF reflected light (with PR
aligned) induced noise in the IMC (converted in frequency
noise)
 The
Faraday isolator had been tested in Nice with 20W
Yag beam in air: more than 40 db isolation (factor 10,000
in power), no significant thermal lensing effect
(http://wwwcascina.virgo.infn.it/collmeetings/presentations/2005/200503/DetectorMeeting/Cleva_28Feb05_Faraday.ppt)
 It
has been reassembled and tuned in clean room with
about 300 mW Yag beam: same isolation factor (about 40
db, surely better than 30 db)
Amsterdam, 03/07/06
Commissioning Meeting
2
Amsterdam, 03/07/06
Commissioning Meeting
3
How the IF was aligned
L/2
Amsterdam, 03/07/06
photodiode
Commissioning Meeting
4
Check of the Faraday isolation
L/2
Amsterdam, 03/07/06
Commissioning Meeting
5
Check of the Faraday isolation
L/2
Crystal: about 43.7° rotation
Amsterdam, 03/07/06
Commissioning Meeting
6
Check of the Faraday isolation
photodiode
L/2
Crystal: about 43.7° rotation
Amsterdam, 03/07/06
Commissioning Meeting
7
Check of the Faraday isolation
Photodiode < 10-3
L/2
2nd polarizer:
about 2×(45°-43.7°) rotation
Crystal: about 43.7° rotation
Total rotation back-and-forth: 90°
Amsterdam, 03/07/06
Commissioning Meeting
8
Faraday isolator
Amsterdam, 03/07/06
Commissioning Meeting
9
Faraday isolator
Amsterdam, 03/07/06
Commissioning Meeting
10
Faraday isolation
July 2004:
before beam attenuation
March 2006
New IB
9%
fringes
0.6%
fringes
More than factor 10 improvement
(corresponding to more than 100 attenuation in power)
Amsterdam, 03/07/06
Commissioning Meeting
11
Faraday isolation
Evidence of fringes in the
IMC when the PR is aligned
since the RFC was locked
(March 2006)
0.5% fringes
3% fringes
Amsterdam, 03/07/06
Commissioning Meeting
12
Faraday isolation
Light reflected towards the
laser and filtered by the last
LB Faraday (March 2006)
Power meter
PR aligned:
P change100 mW
Amsterdam, 03/07/06
Commissioning Meeting
13
Check of Faraday isolation in the tower
L/2
About 30 microwatts
12 – 6 mW
•6 mW entering  30 mW
reflected power (about a
factor 100 attenuation)
(100 times worse
than expected)
•In order to optimize the
attenuation the first
polarizer has to be rotated
by about 5 degrees
L/2
Amsterdam, 03/07/06
Commissioning Meeting
14
Check of Faraday isolation
1) The polarizers seem to be in the same position (rotator markers
are aligned), polarization in the ITF is vertical, dampers and
mirrors are centered;
2) In order to optimize the attenuation the first polarizer has to be
rotated by about 5 degrees: there is much more light reflected by
the second polarizer (more than 10 %), M9 and M10 (mirrors for
the Faraday reflection) have to be moved a lot (mounts to be
modified, etc.)
3) It’s not clear whether something has moved in the Faraday when
mounting the bench, damage occurred, external magnetic field
influence (50 mT are needed to rotate the polarization bay about
2.5 degrees).
Amsterdam, 03/07/06
Commissioning Meeting
15
Possible explanations for less isolation than expected:

Mechanical damage or movement of the Faraday polarizer during IB mounting
(unlikely, no evidence after visual inspection, to be checked more thoroughfully)

Misalignment of the isolator (doesn’t seem to explain the effect)

Thermal damages either of the crystal or of the polarizers

Action of external magnetic fields: already evaluated before, should be
negligible: a more accurate computation and simulation is being done

Possible rotation of the TGG crystal: this can be induced by an external
magnetic field

Thermal effects (thermal dependence of Verdet constant, photoelastic
birefringece, thermal lensing, …):
 the Faraday has been tuned in air with low power (0.5 W) whereas now
about 7W (in vacuum) are going through it
 test in tower have been performed with less than 50 mW: the Faraday
should be tuned again with effective 7 W (but it is impossible to tune in
vacuum)
Amsterdam, 03/07/06
Commissioning Meeting
16
Thermal effects

A Faraday isolator is influenced by parasitic effect since absorption
coefficient of magneto-optic media are relatively high

This property induces a non-uniform cross-section distribution of
temperature which can influence the laser beam in 3 ways:
• Thermal lensing (due to refractive index temperature dependence)
• A non-uniform distribution of rotation angle of the polarization
plane in the TGG is induced by Verdet constant temperature
dependence
• Simultaneous appearance of circular birefringence (Faraday
effect) and linear birefringence due to mechanical straints
(photoelastic effect) caused by temperature gradient.
Amsterdam, 03/07/06
Commissioning Meeting
17
Faraday Isolator normal behaviour
Polarizer
Polarizer
Rotation of 45° of the linear polarization
Reflected light from the ITF
Magneto optic rod
All the polarized light is stopped by the polarizer
 The isolation ratio is maximized.
Amsterdam, 03/07/06
Commissioning Meeting
18
Effects on Faraday isolation ratio


Verdet constant temperature dependence.
Photo-elastic effect
Due to depolarization effects we can decompose
the polarization in 2 orthogonal polarization
Polarizer
Polarizer
Reflected light from the ITF
Magneto optic rod
The « depolarization » of the light induces a decrease of the isolation ratio
because one part of the light is transmitted by the second polarizer.
Amsterdam, 03/07/06
Commissioning Meeting
19
Theoretical description

Self-induced depolarization in Terbium Gallium Garnet (magneto-optic
medium used in our Faraday rotator): studied by Khazanov et al.

They proposed a model describing self-induced depolarization for TGG
without magnetic-field and verified it experimentally.

Moreover they studied the self-induced depolarization phenomenon (isolation
ratio decrease) in the Faraday isolator experimentally.

Model: estimate:
• depolarization ratio due to Verdet constant temperature dependence
• photo-elastic effect
 = f(P0, Q, , , , V)
Amsterdam, 03/07/06
P0 is the incident power
Q, TGG thermo-optical constant,
, the wavelength
, TGG thermal conductivity
, optical losses in TGG
V, Verdet constant
Commissioning Meeting
20
Theoretical results

Estimated the depolarization with respect to the incident laser power P0 (without
magnetic field):

between 0.5 W and 7W, about 20 dB of isolation lost
Amsterdam, 03/07/06
Commissioning Meeting
21
Theoretical results

Khazanov et al. , “Investigation of self-induced dpolarization fo laser radiation in
Terbium Gallium Garnet”, IEEE journal of quantum electronics, Vol. 35, No. 8, August
1999.
Dependence of depolarization parameter  on P0
without magnetic field
▪ experimental result
- Theoretical model
- with magnetic field
 experimental result (FI tuning at 0.5 W)
□ experimental result (FI tuning at 8 W)
Amsterdam, 03/07/06
Dependence of depolarization parameter  on θ in a magnetic field
for diffrent transverse location of the beam (experimental results).
 Faraday
Isolation depends on the power used to
perform the tuning:
 Should be retuned with 7-8 W incident power.
Moreover, it depends also of the beam transverse
location in the TGG.
Commissioning Meeting
22
Check of Faraday isolation
1) Modelization of the Faraday to be developed (we have the optical
model:
• magnetic field mapping and simulation could be useful.
• EOT refuses to give the description of the magnets, it’s
proprietary)
2) Measurements in the tower (low power, lack of space, short time) are
not very reliable;
3) Better measurements should be made if the attenuation is not
sufficient:
• more time (days),
• preparation (send part of the 20 W beam directly inside the tower
without passing through the dihedron).
4) Open the tower, check and retune the Faraday at full power (proper
setup and optical simulation being prepared)
Amsterdam, 03/07/06
Commissioning Meeting
23
Faraday isolation measurement
Isolation
7W
Amsterdam, 03/07/06
Commissioning Meeting
24
Faraday isolator measurement
Amsterdam, 03/07/06
Commissioning Meeting
25
Faraday isolation measurement
Isolation
7W
Amsterdam, 03/07/06
Commissioning Meeting
26
Faraday isolator measurement
Amsterdam, 03/07/06
Commissioning Meeting
27
Zemax simulation : Consequences of first brewster
rotation on beam position in the ITF.
M6
M4
M5
Brewster #2
Faraday
y
z
Input Waist = 2.65 mm
x
Brewster #1
Turning the first polarizer (Brewster #1) of several
degrees improves the isolation ratio
Input beam
Amsterdam, 03/07/06
 We computed the horizontal and vertical shift of
the beam in order to check if rotation induces
alignment problems in the ITF.
Commissioning Meeting
28
Vertical and horizontal beam shifts
Amsterdam, 03/07/06
Commissioning Meeting
29
Mounting modification
Amsterdam, 03/07/06
Commissioning Meeting
30
Possible beam dump modification
Amsterdam, 03/07/06
Commissioning Meeting
31
Future upgrades
Future actions:
 Thermal
problems could be always there and could change with the time
 Different
problems will arise with higher power laser (Virgo +)
 It
would be interesting to have the possibility to tune the Faraday
isolation
 Design
change of the Faraday setup (CRE in preparation):
possibility to remotely rotate one polarizer
or
place a remotely controlled waveplate between the crystal and the polarizer
(this will allow isolation tuning at the expense of power losses).
Amsterdam, 03/07/06
Commissioning Meeting
32
Remotely controlled waveplate
L/2
L/2
Amsterdam, 03/07/06
Commissioning Meeting
33
Remotely controlled polarizer
L/2
Amsterdam, 03/07/06
L/2
Commissioning Meeting
34
Conclusion
• The Faraday isolation (about 100, 20 dB) is sufficient at the moment
• It could and should be better (10,000, 40 dB)
• Unless major mechanical problems, it’s probably due to the tuning
with too low power
• An intervention in tower could improve the situation
Rotation of the Faraday
Modification of some mount
• A system for remotely control the isolation could be designed
Amsterdam, 03/07/06
Commissioning Meeting
35