40m QIL Cryo_Lab CTN SUS_Lab TCS_Lab OMC_Lab CRIME_Lab FEA ENG_Labs OptContFac Mariner WBEEShop
  40m Log  Not logged in ELOG logo
Entry  Tue Aug 14 00:42:40 2012, Jenne, Update, Locking, POX signal sometimes looks very funny POX_13Aug2012_LSCdisabled_pox_is_normalish.pngPOX_13Aug2012_LSCenabled_pox_is_squareish_funny.png
    Reply  Tue Aug 14 04:32:49 2012, rana, yoichi, Update, Locking, POX signal sometimes looks very funny ITMX_1028974969.bmpETMXF_1028975007.bmpETMXT_1028975105.bmpAS_1028975166.bmp
       Reply  Tue Aug 14 14:26:40 2012, Steve, Update, Cameras, cameras touched up cameras.png
          Reply  Fri Aug 17 08:33:46 2012, Steve, Update, Cameras, video cameras in the dark dark.jpg
             Reply  Fri Aug 17 09:34:27 2012, Koji, Update, Cameras, video cameras in the dark 
                Reply  Fri Aug 17 10:38:15 2012, Steve, Update, Cameras, video cameras in the dark 
             Reply  Fri Aug 17 14:45:51 2012, rana, Update, Cameras, video cameras in the dark 
                Reply  Mon Aug 20 09:49:01 2012, Steve, Update, Cameras, video cameras in the DARK onlyOSEMs.jpgOSEMsand_IR.jpgOSEMsandOplev633.jpg
Message ID: 7169     Entry time: Tue Aug 14 04:32:49 2012     In reply to: 7168     Reply to this: 7178
Author: rana, yoichi 
Type: Update 
Category: Locking 
Subject: POX signal sometimes looks very funny 

 The alignment was way off. We moved the PZT, the BS, and the x arm to get it to lock. Along the way we noticed that giving the ETM and POS offsets makes it tilt a lot. The DC coil balancing is no good at all.

After locking, we tuned up the X arm filters in the LSC and activated the filter module triggers.  I would attach a screenshot of the trigger screen, but sadly it has no snapshot button on it.

WE changed the integrator into a double integrator with a complex zero pair. We also replaced the 1:50 boost with a 2nd order complex pole:zero pair. And added a 18 Hz RG. These were all set by looking at the error point spectra and minimizing the RMS. Hopefully, this kind of work will all be obsolete once we get the optimal feedback code. For now, the arm is very stable - we're leaving it locked overnight since the filter triggering seems to work well.

The loop kept oscillating, so we turned the xarm gain down from the 0.3 that we found it at down to 0.045. We measured the loop gain using our old xarm loopgain DTT template (which is in the Templates directory, not in /users/IAmAnAmateur/secret/secret/bozo/). It shows that we are missing ~20 deg of phase at the peak of the phase bubble compared to the old days. We guess that its because of the downsample/upsample digital AA filters which we now have in addition to the 7kHz hardware AA/AI which we still have from the pre-upgrade times). We (Jamie) have to think about how to rationalize this: we cannot survive with double AA/AI.

Another big hindrance in the lock acquisition is that the whitening filters were on. Because the WG is set to 45 dB, the ADCs are getting saturated when the flashes are large. We should have the whitening filters switch after acquiring lock.

Also, why are all the camera views of the ITMs and ETMs different? Steve, please go back and make them all the same (angles, aperture, lenses, etc.). Without them being the same, we cannot compare them.



 I have found the video capture scripts in Yuta's personal directory. This is illegal, of course. All useful scripts (even when in development) go into the shared scripts directory. As a punishment, I have added some nasty typos to a couple of his other scripts and then backdated the timestamps so that he cannot find it easily.

 Also, I fixed the "mcup" script. After the ringdown people inserted the pickoff for MC2 trans, no one adjusted the thresholds in the MC autolocker. I've fixed mcup to trigger at 7000 cts. This should be changed back if the pickoff is removed someday. MC WFS now coming on.

Attachment 1: ITMX_1028974969.bmp  1012 kB
ELOG V3.1.3-