NAOJ GW Elog Logbook 3.2
Tuesday, March 14th, 2017
Morning
We close the MC vacuum chamber.
We move the clean-booth around the PRM vacuum chamber.
We opened the PRM vacuum chamber. Then we installed the clean-booth ceiling and run the clean air flux.
We clean the vacuum chamber as well as the floor around the PRM vacuum chamber.
Afternoon
We installed the Faraday-Isolator inside the PRM vacuum chamber.
While doing that, we moved the two steering mirrors equipped with picomotors from
the Faraday Isolator platform on two independent posts so to place them at the right positions
(drawing and pictures will be added later).
The mirrors are mounted on 7" tall posts so to have the mirrors center at the
correct height i.e. 120 cm from the ground and 21.3 mm from the optical bench which
is the same height as the suspended mirror.
Monday, March 13th, 2017
Morning
We moved the PRM clean-booth around the MC chamber.
We open the top of the MC vacuum chamber and protected the inside of the
vacuum chamber with a plastic sheet.
We then install the clean-booth ceiling and operate the air flux.
Afternoon
We remove the Faraday-Isolator as well as the two steering mirrors from the MC chamber.
The two steering mirrors are equipped with picomotors.
Together with the Faraday-Isolator block we removed the corresponding four
beam dampers (pictures will be added later).
We left the cable used to drive the picomotors inside the MC vacuum chamber.
The second one is the green path on the bench now.
The third one is the infrared path, we increase the beam height.
1.There is some high frequency oscillation when the alignment is better, efficiency reach around 40%(Output power of the laser is 500mW). Even if we reduce the power to only 50mW output from the laser, the oscillation still there. We try to change the temperature and increase the power, find out the oscillation disappear, but under this situation, the alignment is not very good, so we need to find out which is the reason that make the oscillation disappear.
2. When we change the PZT output with hand to get near the resonance from one side, the high frequency oscillation will increase first, but then it starts to reduce, when we reach the resonance, we are not at the maximum of the oscillation.
2.We find out there are two TEM00 mode, one is more powerful then the other.
Now we can lock the cavity with 500 gain, and it is stable.
We installed the simple version of the green path, without the EOM, AOM and MZ, just to send the green beam to the chamber to check the position, and for the infrared beam which should go together into the chamber with the green, we increase it to the right height( the height we will reach after we increase the table leg) and align it to let it go through the Faraday Isolator and reflected by a mirror, goes to the direction to the BS chamber.
Today we checked the picotmotors and the cabling of the 2 steering mirrors in the PR chamber. They both work fine.
The connections has been done in order to have two cables exiting from the flange, labelled SM PITCHS and SM YAWS. (See first attached picutre)
According to the hand-pad convention (A, B, C), the connections are the following:
SM PITCHES:
- A: Pitch SM1
- B: Pitch SM2
- C:
SM YAWS
- A: Yaw SM1
- B: Yaw SM2
- C:
Where SM1 is the steering mirror closer to the suspension (on the path of the green beam) and SM2 is the mirrror further from the suspension (on the path of the IR beam). See second attached picture.
NB: cables for the steering mirror picomotors are connected to the flange closer to the optical table( which has also the coils cables), those for the control of the suspension are connected to the opposite one.
I used the SolidSpec3700 spectrophotometer at ATC to measure the transmission spectrum of the Crystalline coating from CMS.
We have 2 samples, one is AlGaAs transferred on a silica substrate and the other one is transferred on a sapphire substrate.
The settings of the instrument are:
- Slit width = 2nm
- Sampling = 0.5nm
- Range 1400nm - 600nm
- Lamp switch point: 840nm
There are some little differences between the two samples:
- they are shifted by about 3nm
- silica substrate sample has larger side oscillations
Today we checked the picomotors of the suspension in the PR chamber (a part of the filter cavity telescope).
The suspesion is equipped with 5 picomotors and 2 piezo.
Starting from the top (see attached picture):
1° stage:
- YAW (orizontal) + piezo
- PITCH (vertical) + piezo
2° stage:
- Y (defined as the vertical translation of the mirror)
3° stage:
- Z (defined as traslation along the beam axis ( perpendicular to the mirror))
- X (the left one)
In order to control the picomotor we have used as out-of vacuum cable the original cable used for the double pendulum suspension in NM1.
Pinout: http://tamago.mtk.nao.ac.jp/tama/ifo/www_suspension_install_status/burndy.jpg (we are not sure that our definitions of X, Y, Z agree with that reported in this link)
There are two cables (with 6 pins) coming out from the connector, repectively named EF ALIGNMENT and EF LENGTH. We have connected them to the picomotor driver and checked the degree of freedom controlled by each of them.
NB: each cable can move 3 picomotors, which using the hand-pad convention are named A, B, C.
Here what we found:
EF ALIGNEMENT
- A: PITCH
- B: YAW
- C:
EF LENGTH
- A: Z
- B: X
- C: Y
We checked that all the picomotors work in both directions.
- The pressure was 1.2 Torr before vent.
- The GV in the mid point is opened.
- The TMP and the Drypump can work.
I checked the gluing test with the uncoated half ball (elog entry http://www2.nao.ac.jp/~gw-elog/osl/?r=390).
I pushed a bit the half ball and I couldn't remove it, so the gluing is successful.
Therefore I could proceed with the gluing of the coated one.
I made a hole in a plastic cap to hold the coated half ball.
Put some glue on the border oh the central hole of a 1" metal plate
Took the glued plate and attached on the half ball.
Used the plastic cap to adjust the position.
Now I wait for the glue to dry and to cure.
PICOMOTORS CONTROL
Today we completed the test on picomotor motion succeding in moving a test picomotor using "old" picomotor Multiaxis driver 8732 (see picture1) controlled by pc.
A recap of the different option we tested to move the picomotrs:
1) by using "new" newfocus 8753 picomotor controller, connected to the pc by ethernet connection and controlled using labview (MCL commands). See picture 2.
2) by using Multiaxis driver 8732 controlled by hand-pad.
3) by using Multiaxis driver 8732 conected to the computer by GPIB (using GPIB-USB -HS cable) and custom labview VI.
PICOMOTORS CABLING
But if everything works well, the Faraday Isolator should stop the reflect beams. So we checked the polarization again, try to get pure p or s polarization at the place we want. After that, the situation gets better, the peak looks normal, and the green stopped flashing. But the error signal sometimes still have some high frequency oscillation.(To further check tomorrow)
Before the two steering mirror, one of them is dichoric mirror and the other is normal mirror, for get rid of the green reflection, we changed the normal mirror with a second dichoric mirror.
Then we check the power everywhere, we have about 490mW at the output of the laser, 350 after the EOM and 230mW enter the cavity. So now nearly half of the power reach the cavity. With not very good alignment, we got 95mW in green finally, so now the green efficiency is 35%-40%. We try to measure the power reflect back after one of the dichoric mirror, 5% of green power is reflected.
I have 3 sets of roughness measurement data from LMA. It is a sample of silica substrate used to transfer crystalline coating on it. The sets are taken at different points and surfaces of the same sample. I plot the measurement after tilt and curvature removal. I plot the PSDs of the roughness. Result is that the roughness is uniform on the sample. RMS=1nm
Participants: Eleonora, Yuefan, Manuel, Matteo Tacca, Raffaele, Tatsumi-san
Today Tatsumi-san gave us a practice class about how to glue magnets and stand-off ( a.k.a wire braker or spacers) to the mirrors.
Pictures showing the procedure can be found in the following link
https://www.dropbox.com/sh/ggpodec805a30k3/AAB0Bd2UBSd00-coYK7Wp5QYa?dl=0
In the attached file you can find the instruction to make the glue. The glue can be used up to 1 hour after its preparation and the glued parts has to wait for at least 24 hours.
The procedure for Kagra mirror (which is not so different) can be found here
http://gwdoc.icrr.u-tokyo.ac.jp/cgi-bin/private/DocDB/ShowDocument?docid=4248
There are some problems we need to fixed,
1. The crystal is not fixed, when we try to get the mode matching, sometimes cannot avoid touch the cavity, then we lost everything. Matteo did the alignment many times in few hours working.
2.There are too many green reflected back, the maximum efficiency we got is about 17%, with a input infrared power of 670mW, and output green of 119mW, when the cavity is locked. But there are at least four reflect green beam on the mirror before the dichoric mirror, we are able to measure two of them, one has 1-2 mW power, the other has 4mW.
3. The green beam has some fringe on it, so this maybe means there is not only one green beam come out from the cavity, but we are not sure.
After first time we got a efficiency of about 16%, we break the lock and wait for twenty minutes to see if the system is stable or not. When we try to lock it again 20min later, the output of green is almost the same with some tiny re-alignment.
I finished the configuration of the optical setup for our hemispherical scatterometer Yesterday.
The beam of the new laser is now focussed to a beam diameter of ca. 0.32 mm at the sample's surface.
However, I think there are still things that need to be done:
A diagram of measurements of the beam's profile is shown in the attachment. It shows the measured values of the profile without a focussing lens (circles) and with a lens at 430 mm distance to the collimator (stars). The actual distance of the lens in the setup is, however, 490 mm. Due to a small mistake, I underestimated the distance to the sample a little bit and had to readjust the positionof the lens.
Anyway, as can be seen, the measured values are in consistence with the calculated ones (dotted curves) only until the waist has been reached. With increasing distance again, there is a considerable difference between measured and theoretical value of the beam's radius*.
I think I have to measure again and try to find out the reason of the difference.
*The values are related to a Gaussian fit of the beam's cross-sectional intensity and represent the diameter of a Gaussian curve.
https://www.thorlabschina.cn/newgrouppage9.cfm?objectgroup_id=4808&pn=BSX17
And we test the cable which can send signal from end room to the central room, we followed the cable to the end room, found a same box as in the central room, send a sine signal from there and got the signal with the oscilloscope in the central.But in the west end, we could not find the box, but I think we will not use that arm.
On Friday, Manuel and Takahashi-san went to TAMA and sucessfully tested picomotor controller connection. Here a summary of what I have understood about this:
In order to make the computer be able to comunicate with the picomotor controller they must have the same IP address (except for the last digits).
Controller should have a static iP address and for this reason cannot be connected directly to NAOJ network (since it doesn't allow a static IP address). We have to choose a port in the getaway which are isolated from NAOJ network (and disconnetected from the internet).
The network used for the optical lever control is one if the and can be a good choise. The relative IP adreddes are of the form 133.140.121.XX.
The ports relative to this network in the gataway that can be found up-left corner in tama central room are the number 1 and 2 (see pictures 1 and 2).
The steps to establish the connection :
1) Pick a gateaway port which is isolated from the naoj network (disconnected from the internet) and find the relative IP addreses
2) Using serial port assign to the controller a proper IP address ( command IPADRR
3) The proper IP address has to be the same of the network except for the last digits which has to be not yet assigned ( make a ping to check..)
5) Also the pc IP adress should be check and eventually changed (with the " assign manually option") to a proper one. (Same of the network with different last digits)
6) Connect both the PC and the controller to the same network and try the ping
[NB if we are not interested in being connected to a network we can simply connect the pc and the controller to the same switch and manually assign them a compatible IP address. In order to connect the PC directy to the controller a crossover cable is needed ]
following Takahashi-san's advice I connected PC, controller and network cable to the same Ethernet hub. The ping still fails. BUT if I unplug the nework cable, which means I isolate my system from NAOJ network, then I get this ping returns:
Pinging 10.68.10.241 with 32 bytes of data:
Reply from 133.40.117.65: Destination host unreachable.
Reply from 133.40.117.65: Destination host unreachable.
Reply from 133.40.117.65: Destination host unreachable.
Reply from 133.40.117.65: Destination host unreachable.
Ping statistics for 10.68.10.241:
Packets: Sent = 4, Recieved = 4, Lost = 0 (0% loss),
In the case of 192.168.11.121-IP controller the ping still fails, even in the isolated case.
I found that NAOJ network uses DNS, so I follow the instructions of "Ethernet Configuration Example 2" at page 50 of the manual. (last time we tried Example 1 and, apparently, it didn't work because of DNS).
Through serial port, the IPADDR command returns 192.168.11.121 for one controller and 10.68.10.241 for the controller labeled NM2.
I connect the controller with IP address 192.168.11.121.
HOSTNAME command returns: nf8752-295333
I change the hostname with the command HOSTNAME NF8752-001
Save the configuration with the SAV command
RESET
Try to ping NF8752-001, failed (it says: "Ping request could not find host NF8752-001")
In the second figure, the y-axis is the green power over infrared power, I just want to show the efficiency of infrared change into green at different input power.
I already changed the dichoric mirror from Thorlab two inch one into CVI one inch mirror has a anti-reflect coating on the back side can reduce the amount of power reflect by it. Still doing the alignment with the new mirror.
