DriptaB, TonyS, RickS
Responding to a request from Jeff Kissel, we modified the Pcal measurement setup and used the standard Pcal analysis scripts to measure the reflectivity of two types of beamsplitters that have been procured for the SPI upgrade: 50/50 and 85/15 beamsplitters (two of each). Usually we are interested in the ratios of the responsivities of the two power sensors, but the same measurement results can be combined differently to yield the beamsplitter ratio.
NOTE: The Pcal system wavelength is 1047 nm, not 1064 nm.
The boxes containing the optics were labeled:
SPI PWR BS, 1064nm
50R:50T 45 DEG-Ppos
Batch 16075HA1
and
SPI PWR BS, 1064nm
85R:15T 45 DEG-Ppos
Batch 16075HA1
We measured two of each of the two beamsplitter types (50/50 and 85/15).
The measurement results are shown in the plots attached below. Note that HOP stands for "hundredth of a percent."
The measured reflectivities are:
50R/50T #1, two measuremetns: R = 0.4830 ± 0.002 %; R = 0.4830 ± 0.002 %
50R/50T #2: R = 0.4827. ± 0.002%
85R/15T #1 two measurements: R = 0.8624 ± 0.0005%; R = 0.8625 ± 0.0005 %
85R/15T #2: R = 0. ± 0.8599 ± 0.0004 %
Given the small specified wedges in the beamsplitter substrates, the ghost beams from the AR surface would likely pass through the entrance apertures on the power sensors (one inch diameter) and add to the received power in both reflection and transmission. However, the AR surface reflectivty appeared to be below the specified 0.1 %. We were not able to identify and measure the power in the beams reflecting from the AR surface. We will have another look when we break down the measurement setup.
Measurement setup and details:
We increased the Pcal Optical Follower Servo offset to give about 600 mW in each of the two output beams. We blocked one of the output beams and used the other beam for the beamsplitter measurements.
We aligned the laser steering mirror inside the transmitter module such that this beam propagates parallel to the table surface and along a row of holes in the optical table.
We found that the Pcal module output beam had some stray light of unknown origin outside the main beam (mostly three spots - see photo below) so we installed an aperture upstream at the Tx module output to block this stray light.
We installed a Karl Lambrect TFPC-12-1047nm thin film polarizing beamsplitter cube and aligned it such that the plane of the cube is parallel to the optical table surface. The Pcal Tx module output beams are nominally polarized paralllel to the surface of the optical table, but we installed this BS cube and carefully aligned it's plane of incidence to ensure that the polarization would be parallel to the plane of incidence of the beamsplitter to be tested.
We then installed the beamsplitter to be tested in the transmitted (polarized parallel to the surface of the table) beam and aligned it such that the plane of incidence is parallel to the optical table surface and the reflected beam is parallel to a row of holes on the optical table (45 deg. AOI, p-pol) and centered on the entrance aperture of one of the Pcal integrating spheres. The power incident on the beamsplitter was measured to be 570 mW.
The beams transmitted by the beamsplitter to be tested was reflected by a Pcal HR mirror (Precision Photonics / ATFilms Y1S-1025-45, Lot# WO22401, ITM100509, Run# 213472) and aligned to propagate along a row of holes and be centered on the entrance aperture of the second Pcal integrating sphere. The fractional power transmission of this HR mirror was measured twice to be: 3.2e-4 and 3.9 e-4. Average = 3.6e-4. We correct for this loss of power in the transmitted beam in the BS reflectivity calculations.
