NAOJ GW Elog Logbook 3.2
I have measured transfer functions and spectra for the PR mirror.
- Figure 1: TF when injecting noise in YAW
- Figure 2: TF when injecting noise in PITCH
- Figure 3: spectrum YAW (comparison open and closed loop)
- Figure 4: spectrum PITCH (comparison open and closed loop)
NB. It seems that there is some acoustic noise: I did the measurement with the light switched off but the chamber was open and I didn't put any boxex around the laser and the PSD
While looking at the error signals in the past days I have noticed that often some strong oscillations at 1 Hz and 3 Hz were present (see figure 5). We observed it also in the past and we were suspecting the stack vibratiion introduced by the steering mirrors used for the optical lever ( see entry 276: the attached spectra taken last july are very similar to those of fig 5)
Since this oscillation is not present when the intermadiate mass is well centered on the magnet older plate (as it was just after the mirror installation or today after that I performed the check on the yaw motion) and it appeared after we tried to move as much as possible the yaw picomotor yesterday, I think it is caused by the intermediate mass touching the magnet holder plate.
In order to better quantify the yaw motion of the PR suspended mirror I made the optical lever reflected beam imping on a ruler put at a distance of about 88 cm from the mirror.(See attached picture 1)
I started from one side of the picomotor range (left column of the second attached picture) and I kept turing the screw as far as I was able to see a beam movements on the ruler.
I had to stop well before that the end of the range of the picomotor (right column of the second attached picture) since the beam was not moving anymore (and the mirror was oscillating a lot)
Even if it difficult to clearly see it, I think that the intermiate mass was touching the magnet holder plate preventing a further motion.
The total motion of the beam on the ruler is about 1.5 cm which corresponds to 8.5e-3 rad.
After the problem reported yesterday here. I have replaced the picomotors GPIB driver with a similar one that I have found in the south end room.
At the beginning it seems to work fine and I was able to move the 3 degree of freedom of the 2'' telescope mirror in both direction both with the pad that using the PC.
Then, I have noticed that while connecting more than one cable to the driver output ports, it starts showing some problems.
Its behaviour becomes very strange and it is no loger possibile to move properly the selected channel neither from remote nor with the pad.
For exemple: if I send a remote command to move a specific channel the driver display shows correctly the name of the selected channel but it cannot move it and will instead make other picomotor ( connected to a different channel) to produce some noise.
I could not find any repetable pattern in such kind of strange behaviour while trying to test different combination of slots and cables.
I only noticed that everything is working fine when only one cable is connected.
Since Raffaele suggested that it can be a cabling problem, I spent some times to check the connections but I haven't find anything wrong so far.
In the attachement there is a picture of the connection we made between the 2" telescope picomotor and the invacuum cable.
Participants: Matteo Tacca, Eleonora
Today we installed the 2" telescope mirror in the BS chamber. The infrared and green beams have been superposed. They impinge the first telescope mirror and are reflected on the second one (PR suspended mirror). Beams are both hitting the PR mirror almost in the center. We observed that the green beam is slightly elliptical (also before hitting the first telescope mirror)
Then we tried to send the beams back on the BS mirror but we didn't manage to move the yaw of the PR mirror enough to do that.
We spent some time investigating this issue.
Some remarks:
1) The PR mirror is very sensitive to pich motion: few turns of the pitch picomotors screw are enough to remarkably change the vertical angle of the beam.
2) The yaw picomotor screw works fine as we can observe the relative slide to move (in both directions) when turning the screw.
3) We checked that the mirror is free to move and it not touching the earthquake stops
4) It was possible to observe the pitch motion using the error signal of the optical lever. The displacement is big enough to make it sweep the all the signal range (from -5V to +5V). Assuming a calibration of 2.2e-4 rad/V this means that the yaw motion achievable is al least larger than 2.2e-3 rad.
5) Another critical point is the intermidiate mass. During the mirror istallation we noticed that one side of it was almost touching the magnet holder plate. We cannot rule out the possibility that the yaw motion is limited by this.
Participants: Matteo Tacca, Eleonora
Today we moved the BS suspension in order to have the reflecting surface of the mirror 10 cm south with respect to the center of the chamber, according to the telescope design.
We have also installed the fixed mirror support and connected the in-vacuum and out-of-vacuum cable to move the picomotors. We had some issues in moving the pitch and the translation picomotors in one direction (the other direction works fine).
The picomotors seems to receve signals from the drivers since they make a noise but most of the time they don't move and sometimes they move very slowly.
NB: Ony few days ago I I have checked that the 3 picomotors can work in both directions (entry 420)
We made many checks in order to find out the cause problem:
1) We checked carefully all the connections in and out of the vacuum chamber. They were all fine.
2) We changed the picomotor with a spare one but it didn't solved the problem. So it is not a problem of the picomotor itself.
3) We have tried to control them by PC using labview: nothing changed.
4) We have changed the pad of the driver: nothing changed.
5) We have changed the slot at which the cable was connected in the driver: nothing changed.
6) We have inverted the pins of pitch and yaw in the in vacuum connector and this time the pitch was working fine in both directions while the yaw was now showing the issue.
This last test made as to suspect the driver which, by the way, showed some other problems ( i.e the knob to change the controlled slot gets often stuck and it is necessary to switch it on and off).
I found a similar driver in the instrument shelf at the first flor but it didn't switch on.
Tomorrow we will:
1) Check the driver on other picomors, as those of the PR suspension which were also succesfully tested before.
2) Look for other GPIB picomotor drivers in TAMA. (NB according to their label these drivers have been bought in 1999)
Today I restored the optical lever of the telescope suspended mirror in PR chamber. Error signals look fine even if, since the chamber is open, they are a bit noisy than usual due to the light.
In order to close the loops I had to increase a bit the gain of the yaw loop and change both the loop signs. I have also tweaked the driving matrix to reduce the exicitation of yaw when injecting noise on pitch. In the previous configuration I have added a rotation of the sensing of 0.04 rad (as reported here), now I have removed it since it seems to be not needed anymore.
I will upload soon spectra and transfer functions.
Today while tiding up the area close to the BS chamber I noticed that there was same air leaking from a pipe between BS and NM2 chamber (see attached picture).
As soon as I touched the pipe tying to tight it better, it came off, loosing a lot of air. With the help of Matteo, Fujii and Ishizaki-san we managed to reconnect and tight it. Now the pipe seems well attached but there is still some leakege even if it is smaller than it was before.
It has to be checked by an expert as soon as possible.
Then install the reflect mirror just after the dichoric mirror temporarily to adjust the reflect beam of infrared to goes the position it should be at the optical bench as the new optical scheme.
We also check the power loss of the beam goes through the Faraday and comes back to the optical bench, it is about 6% and 10%.
The we put back the middle part of the PR chamber, since then the beam will first go through the thick window mirror, if it is not exactly perpendicular to the mirror, there is at least some shift of the beam. And also it seems the mirror also is a wedge mirror, now the beam in the chamber moves a lot. One of the idea is to move the last mirror before the beam goes into the chamber on the bench, but then there is the risk that the beam will be move out from the mirror. Matteo will work on this in these days.
And on Friday, we plan to measure the green beam, and do some calculation to decided the lens we are to use, and we should buy them, since now in the lab the shortest focus length is 100mm.
I placed a lamp at a distance D from the convex surface and I looked
for the image reflected by the second face of the mirror (which is flat).
The image is real and placed at a distance D' from the convex surface.
I changed D until D'=D. In this situation the focal length f is
f = D = D' = 1.305 +/- 0.01 m
I can deduce the ROC from the expression ROC = f * (n-1) where n is the
fused silica index of refraction in the visible (green) i.e n=1.4607 at
lambda = 532 nm. I found ROC=0.601 +/- 0.005 m.
In conclusion, within the error of the measurement the mirror ROC
is equal to the required value (which is 0.6 m).
This morning I have removed the supports used to glue the standoffs on the BS telescope mirror that I glued last wednesday evening. For the moment, they look well attached to the mirror.
The magnets were glued on tuesday and their supports were removed on wednesday in order to rotate the mirror and allow for the glueing on the standoffs.
Since BS mirror has a different disposition of magnets with respect to test masses, the glueing has to be done in 2 times.
http://www2.nao.ac.jp/~gw-elog/osl/uploads/260_20160627054235_mirrormagnet.jpg
Both magnets and standoff have been glued using a new set of MasterBond EP30-2, stored in the fridge of TAMA.
In the optical lever scheme, a beam is sent to the center of the the mirror with an angle α.
A rotation of the mirror around its vertical axis (a yaw) of an angle β, corresponds to an angular displacement of the reflected beam of 2β.
This produces a shift of the beam's position on a PSD placed at a distance L from the mirror of
X = L tan (2β) ≈ 2Lβ
Now, if the beam impinges (with an angle α) at a distance d from the center, the shift on the PSD becomes
Xd = L tan (2β) + d (sinβ/cos(α+β)) ≈ 2Lβ + d β (1/(cosα - βsinα))
In the case of the beamsplitter, the optical lever beam is currently sent with a very small incidence angle (α ≈ 0) and impinges at a distance d (in the orizontal direction) of about 3.5 cm from the mirror center. The distance between the mirror and the PSD (arm of the optical lever) is about 60 cm.
The relative errore due to the mis-centering is (Xd-X) / X = d/(2L) = 3%
which is small. So it seems not so crucial to hit the center of the mirror.
For what concers the pitch, I would say that as long as the beam is on the orizontal axis (as in our case), its mesurement is not affected by the mis-centering.
I disassembled one of the two HEPA filter of the cleanbooth for the absorption measurement system ( Model MAC-103 ) in order to check the conditions and to figure out whether and how to change it. The circular prefilter is very easy to replace and very dirty, I think we should change it. The real HEPA filter inside is quite dirty but is not easy to replace because the filter is kind of glued at the frame, so maybe we can contact the company. I hope there is the possibility to change only the filter, and not the whole fan.
The filter's fan part was fine, so it was enough to buy only the filter part. New HEPA filters were delivered to Tama last friday. I washed the prefilters with water, I cleaned the fans, replaced the filters and placed them back to the top of the absorption bench clean booth.
And we could not kill one higher order mode to get better mode matching, and also there is one mode jump out from time to time when we try to align the cavity. The final efficiency we got is about 20%(260mW goes into the cavity and 55mW green received at the transmission of the dichoric, 12mW received by the reflection of the dichoric)
Then we increased the power a little bit, do the alignment again, then the error signal and the infrared signal shows in the third picture. There is a flashing spot in the infrared signal.
But the changing part is when we lock the cavity, the 500mW situation has the oscillation also(picture 4), but the second seams not.(pic 5)
According to this circumstance, we at least got one conclusion is that the high frequency oscillation has nothing to do with the green.
Then we rethought the polarization again, found out no matter s or p polarization reflect back, it is should stopped by the Faraday Isolator, but now it seems not.
We decided to check the polarization again from the beginning. Firstly one thing for sure is that, the input infrared beam to the cavity is in p polarization, and the reflect infrared beam is also p polarization.
Then we try to put another FI and half wave plate after the BS mirror and before the telescope to cancel the reflect infrared beam from the cavity. We use the same FI as we used before, but found out the two polarization cube point to different direction of the original one. Then the new put FI did reflect out the reflection beam, and get rid of the oscillation.
Then we guess that the original FI did not do its job maybe because the two wave plate before it did not clean the polarization as they should do.So we put a PBS after the first quarter wave plate and second wave plate, found out exactly what we suspected, so we adjusted them firstly then the Faraday and the other two half wave plate, got pure p polarization inject into the cavity.
Now the oscillation seems not there but the efficiency of green beam is only about 10%, so we tried to increased the alignment, but it seems we should do it from the start. After we make a progress in the alignment, we can check again if the oscillation really disappear. Then we can be sure that it is caused by the polarization problem.
After doing this, we remove all the tall mounts for infrared and do the alignment again both in the infrared and green path. During the alignment, we found out maybe the cavity itself is a little bit tilt up, with one mirror we recover the beam to the right height.
Then we inject both the green and infrared beam into the chamber.
Firstly, adjust the position of G1 mirror in the picture try to get the green beam in the middle of it in horizontal direction. Then use adjust it slightly to get the reflection beam goes to the GR mirror.
Then we adjust the infrared beam, use the last two mirror on the bench in infrared path to adjust the infrared go through the Faraday Isolator, reflect by the R2 mirror and also goes to the GR mirror.
Then after the GR mirror we try to do our best to overlap the infrared and green beam and get a good result, next step is let these two beams goes into the BS chamber and try to overlap them there.
Then we installed the RP1 mirror, this is a mirror with high reflection with 0 zero degree. Then also install the RP2 mirror, this is the mirror has high reflection in P polarization. With this two mirror we got the p polarization reflection of the infrared beam and it arrives about the right position in the bench.
This morning I have removed the support of the standoff glued on saturday. This time it has remained attached to the mirror and it looks ok.
This morning I went to check the status of the magnets and stand-off we glued on thursday on the telescope mirror.
While removing the jigs, a stand-off came off. We probably put to much glue on it and it overflowed a bit on the support causing the stand-off to remain attached to the support.
After removing it, I have cleaned the mirror and the support and glued it again paying more attention to avoid overflow.
Matteo gently touched the other 3 stand-offs and they seem to be well glued. We did not touch the magnets but they also look good.
Some pictures taken by Matteo are attached.
A special thank to Yuefan who woke up at 10 a.m. to prepare the glue!
This morning I have removed the support of the standoff glued on saturday. This time it has remained attached to the mirror and it looks ok.
Today we checked the picomotors in the BS chamber.
BS suspension picomotors
The out-of-vacuum connector is connected to the flange named PO 4 (see first attached picture)
There are two cables (with 6 pins) coming out from the connector, that I labelled BS ALIGNMENT and BS LENGTH.
Here what we found:
BS ALIGNEMENT
- A: PITCH
- B: YAW
- C:
BS LENGTH
- A: Z
- B: X
- C: should be Y but doesn't work!
We checked the connections and they seem to work, so it is probably the picomotor itself which doesn't work although the screw can be rotate by hand.
2'' telescope mirror picomotors
The not-suspended 2" telescope mirror will be equipped with 3 picomotors (pitch, yaw and a translation). See picture 2.
I have checked that the 3 picomotors can work and I prepared a cable for the out-of-vacuum connection.
I re-used a former cable used for PO1 in the PR chamber adding 2 pins for controlling the translation picomotor. The pinout of the modified, cable labelled TELESCOPE 2", is shown in figure 3.
We checkd the ROC of the 4" telescope mirror. To do that we used the mirror to focus the
image of a lamp placed at distance D1 and measured the distance D2 of the image from the mirror.
When the two distances are made equal D1 = D2 = 6.05 m = ROC of mirror.
The error of the measurement is about +/- 2 cm.
The specification was 6 m +/- 0.3 m. Looks good.
Wednesday, March 15th, 2017
We opened the top of the BS vacuum chamber.
Apart from the BS suspension and the corresponding cables for the coils and the
picomotors, there is no additional cable in the vacuum chamber.
There are two masses that were used as counter masses to balance the PO suspensions.