The purpose of the optical system is to reimage AST/RO's focal plane onto a compact 2x2 array of lenses located in the array cryostat and to efficiently inject local oscillator power into each mixer. Focal plane reimaging is achieved by a pair of parabolic mirrors and two flats (Figure 4). Each AST/RO beam expands from the focal plane located ~900 mm above the receiver pallets on to the first offset parabola (P1). P1 collimates the beams and conveys them via two flats to the second parabola (P2). P2 focuses and re-parallelizes the beams so that they arrive at the mixer lenses with the correct size and placement.

Figure 4: Optical layout for Pole Star Click on thumbnail for full-size image!

The local oscillator for the array is a single oscillator and multiplier. A quasi-optical LO power divider (Figure 5) is used to split the focused LO beam into 4 equal parts. The power dividing is accomplished by two, low-loss, ~50% crystalline quartz beamsplitters. Each beam splitter is paired with a flat mirror to give the emerging beams the proper horizontal and vertical offsets. After including all loss mechanisms, we estimate that 18% of the incident LO power is in EACH LO beam at the output of the power divider. The LO polarization is orthogonal to the signal beams. The four beams are then injected into the `sky' signal path by a high quality wire grid. A Martin-Puplett interferometer then rotates the LO beam polarization to match that of the sky and directs them into the cryostat. A 2x2 array of HDPE lenses then focus the beams into the mixer feedhorns. The lenses are rigidly mounted to the mixers and are designed for operation at 4K (Figure 6).

Figure 5: LO power divider Click on thumbnail for full-size image! Figure 5b: COOL image of laser beam being diplexed by our LO power divider. The LO beams have different optical intensities because of the attenuation of visible light in the quartz beamsplitters. Click on thumbnail for full-size image! Figure 6: Mixer foreoptics Click on thumbnail for full-size image!