Test bed characterization

In accordance with the LRI configuration, our setup was designed to measure the round- trip fluctuations of the beams traveling through and between the TMA units in order to put an upper limit on the TMA length stability. This would verify that the path length fluctuations for a beam traveling through the TMA were less than the allocated noise needed to reach the desired science performance.

Our setup included two offset phase-locked lasers, two 2-inch beam splitters to inter- fere local and distant beams, various steering mirrors/beam splitters/apertures, and two quadrant photodetectors. Most of the hardware used in this experiment was the same as used in the acquisition experiment1_{. The optics and photodetectors were mounted on}

an aluminum breadboard in a large vacuum chamber. Light from both lasers entered the vacuum chamber via an optical quality window.

First tests of the round-trip path length stability were conducted using two commercial corner cubes made by PLX Inc. with dimensions and beam co-alignment properties similar to the TMA requirements (although not suitable for launch or a space environment). A diagram and photo of the setup are shown in Figures 5.6 & 5.7. We ordered these commercial units while the prototype CFRP TMA was being assembled so that the test bed could be developed in parallel and we would be ready to make performance measurements as soon as the assembled prototype was delivered to ANU. The PLX units had 2-inch diameter clear apertures, with parallelism between the incoming and outgoing beams stated to be within 1 arc second (_∼5µrad). Like the TMA design, two of the corner cube optics were on one side of the assembly (denoted as the roof prisms in Figure 5.6), while

1

We used the same two NPRO lasers, ADCs and FPGAs, and the quadrant photodetectors were of the same design.

BS 2 attenuation wheel BS 1 QPD 1 QPD 2 PLX unit 1 with beam splitter

PLX unit 2 RP RP LEGEND BS: Be am splitter RP: Roof prism

HR: High reflectivity mirror QPD: Quadrant photodetector

HR virtual corner cube

intersection point

BS 3

virtual corner cube intersection point L1 L2 a1 a2 b2 b1 iris

Figure 5.6: Test bed characterization with commercial corner cubes made by PLX Inc.

the third optic was on the other side of the unit 60 cm away.

Note that the two PLX units were not identical; one of the corner cube optics in unit 1 was an uncoated piece of glass (wedged at a few degrees), acting as a beam splitter with low reflectivity (96:4 splitting ratio between the reflected and transmitted beams, respectively). This beam splitter is labeled BS 1 in Figure 5.6. This modified design combined the functions of the retroreflector and beam splitter. Early in the GRACE Follow-On development, it was investigated as an alternative to the all-reflective design due to its immunity to path length errors introduced by a transmissible beam splitter in the presence of spacecraft rotation. On GRACE Follow-On, these errors are minimized by the introduction of a compensation plate, shown in the LRI layout in Figure 2.1.

Two additional apertures were built into this assembly to access all 4 ports of the beam splitter. The corner cube optics were mounted in a hollow glass enclosure made from fused silica (CTE: 0.55 ppm/◦C), while the assembly for unit 2 was a glass tube made of Pyrex (CTE: 3.25 ppm/◦C). Each glass optical assembly was mounted in an aluminum housing using flexible pads. The mirrors of the corner cubes were made of the same material as its glass tube, so the beam co-alignment would be largely insensitive to temperature changes. With the exception of the beam splitter optic of unit 1, all corner cube optics were coated with protected gold coatings for high reflectivity.

An attenuation wheel was installed in one of the sections of the round-trip loop to min- imize contributions in the phase read-out from successive round-trip passes. This problem is specific to our setup and is not an issue for GRACE Follow-On due to the inherent diffraction loss of the inter-spacecraft path (greater than 108_{). In the corresponding path}

§5.2 Static stability tests 69

(a) Black PLX units (60 cm long) installed in test bed housed in a large vacuum chamber (lid off).

(b) Overview showing proximity of vacuum chamber next to optical bench with lasers.

Figure 5.7: Test bed characterization with commercial corner cubes made by PLX Inc. Test bed built on an aluminum breadboard in the LRI-like configuration, sitting on a table inside a large vacuum chamber (the chamber lid is off in these photos). The viewport / optical window for coupling light into the chamber can be seen at the top of photograph (a). The optical bench where the lasers are mounted, sitting next to the vacuum chamber, is shown in photograph (b).

on the other side of the round-trip loop, we installed a high reflectivity beam splitter (BS 3) instead of an attenuation wheel. The beam splitter effectively reduced power for the multiple round-trip passes, but additionally acted as a compensation plate (as in the LRI setup) to reduce sensitivity to pointing fluctuations of beam 1. This is not a problem for beam 2 as the combination beam splitter is one of the TMA optics.