MIF (General)stefan.ballmer - 17:50 Friday 14 June 2019 (9241)
Print this reportPRC cavity scan method for Gouy phase measurement - method descriptionEleonora, Yutaro, Stefan
We tried a new method for measuring the Gouy phase by doing a cavity length scan for the PRC. This klog summarizes the method - another klog with resuts will follow.
Idea: We were looking for a way to measure the Gouy phase of the PRC that does not require a large extra setup. The usual idea of a frequency scan has 2 problems:
(i) The required frequencies are too high to effective drive through the MC cavity.
(ii) Changing the frequency would also affect the MICH leakage, resulting in a more complicated scan.
On the other hand, a length scan has the problem of knowing the instantaneous length of the PRC across multiple fringes, while keeping MICH locked.
We decided to try "inertial locking" of MICH, while kicking the PRM:
a) In order to shut down all feed-back to MICH and leaving it "parked" on fringe for as long as possible, we set up a low-pass filer & compensating high-pass filter in MICH1 and MICH2:
- MICH1: LPHOLD zpk([500;500],[0.5;0.5],1,"n")
- MICH2: HPHOLD zpk([0.5;0.5],[500;500],1,"n")
This guarantees that we keep applying the correct DC value when hitting the HOLD button in MICH1. We achieved about 0.1 seconds of inertial locking.
b) For PRC we decided to do the opposite: apply a large DC force to move the PRM across several fringes before the BS moves significantly off fringe.
We achieved that simultaneously, by Guardian, by turned on the MICH1 HOLD, PRCL1 HOLD and PRL2 OFFSET.
Note though that the the ezca.switch command is too slow, as it waits for read-backs. Instead we used the "toggle" features with the codes 2048 for HOLD and 8 for OFFSET, i.e.:
- ezca['LSC-MICH1_SW2']=2048 ; ezca['LSC-PRCL1_SW2']=2048 ; ezca['LSC-PRCL2_SW1']=8 ;
Together with an offset of 3000 counts in PRCL2 this seemed to work well. The balance was picking an offset large enough to jettison the PRM through several fringes before the BS moved off fringe, but small enough that the 16384Hz sampling rate can resolve all the higher-order peaks.
Comments to this report:
stefan.ballmer - 23:17 Friday 14 June 2019 (9244)
Print this reportWe extended the method to SRC Gouy phase measurements:
- Set up guardian state to lock the SRMI: PRM misaligned, MICH locked on bark fringe, SRCL locked on RF17. That lock is just as stable as the PRMI now.
- Set up a SRMI_scan state that again "inertially locks" MICH, and kicks the SRM out. A good offset in SRC2 turned out to be about 15000cts.
- The only variation on this lock was that we started with some non-zero MICH fringe offset (set by a MICH1 offset in-lock ot about 15cts). This allowed some non-zero carrier light into the SRCL cavity, which we need to identify the carrier in the scan. Turns out though that the slight wandering of the BS during the "inertial lock" is enough to see the carrier anyway.
The data from both PRCL and SRCL scan should give Gouy phased with maybe 2 degree uncertainty. Eleonora will post the processed data and final answer.