4.B The DESpec Fiber Positioner
The fiber positioner should be able to support ~4000 fibers at prime focus. The fiber positioner must move the tips of the optical fibers to predetermined positions for each exposure and then hold them in place for the length of the exposure. It must also gather the fibers into bundles that run to the spectrographs. Since DESpec will reutilize the corrector optics of DECam, the focal plane will be about the same size, i.e., a radius of 225.54 cm. This implies a separation between fibers (i.e., the pitch) of a bit less than 7 mm. Figure 4.4 shows the locations of 3781 fiber centers within the clear aperture radius using a 7 mm fiber pitch.
Figure 4.4: Array of DESpec fiber positioners. Here we show 3781 locations with 7 mm spacing (pitch) within the clear aperture. If the pitch were 6.3 mm, the clear aperture would accommodate 4675 fiber positioners. The x and y axes have units of mm. Again, we use a clear aperture radius of 225.54 mm.
There are two general classes of fiber positioners that could provide technical solutions for DESpec: the “Twirling Post” design and the “Tilting Spine” design. Figure 4.5 shows a schematic (Seiffert, 2009) of the “Cobra” fiber positioner, under development at Caltech/JPL, that was considered for WFMOS and is being considered for the Subaru Prime Focus Spectrograph (Sumire). The fiber is held at the tip of the “twirling post”. The post can rotate, and the tip can rotate with respect to the post, thus allowing the fiber to be positioned anywhere within a “patrol radius”. A variant of this (phi-phi) design has a rotatable post with a tip that moves along the radius (r-phi).
Figure 4.6 shows a schematic (Akiyama et al. 2008) of the second class, the “Echidna” fiber positioner developed at the Anglo Australian Observatory and currently used in the FMOS instrument on the Subaru Telescope in Hawaii. This “Tilting Spine” solution uses a unit positioner (Figure 4.7) that is tilted into place by a simple piezo-electric actuator at the base. The feasibility of an Echidna-type system for DESpec has been investigated by the AAO, and no major problems accommodating the required number of actuators, their pitch, or the curved focal surface, have been identified (Saunders et al. 2012, Proc. SPIE 8446-188).
At present, the operating “Tilting Spine” type positioner (Echidna) has a smaller spacing between fibers (7 mm pitch) than the smallest operating “Twirling Posts” (WFMOS-Cobra prototypes, 12 mm). The patrol radius for the Echidna is equal to the pitch, and all seven “nearest-neighbor” fibers can be positioned as close together as the size of the tip of the spine allows. The patrol radius for the prototype Twirling Posts is about 2/3 of the pitch and there are not so many nearest-neighbors. A scaled-up Echidna to DESpec size would have ~4000 fibers. The Cobra-type positioner would have 1000-1500 fibers with the present state-of-the-art; it is expected that they could be made with smaller pitch, which would increase the number of fibers, but there is more R&D required for this type of unit positioner than the other.
Figure 4.5: The “COBRA” Fiber Positioner designed for WFMOS (left). The tip of the fiber is held at the bottom end of the unit positioner (right). LAMOST uses a design similar to this.
Some infrastructure is required to operate the fiber positioner. Any design requires a system to measure the fiber position during configuration so that the position can be verified to be correct within about 10 microns. The fibers are moved into position and then back-illuminated. Next, a camera is used to check the new position of the fiber. This process is iterated up to 7 times. The total configuration time is expected to be 60 to 90 seconds. We will still need to have Guide and Focus capability, and this may be accomplished by having a small number of CCDs in the DESpec focal plane.
Figure 4.6: The 1600-spine Echidna Fiber Positioner proposed for WFMOS on the Subaru Telescope. The DESpec version will have the same physical diameter but with a smaller pitch and curved modules.
Figure 4.7: The Echidna Unit Fiber Positioner. The spines pivot from mounts at the base and are driven by piezo-electric actuators.
4.B.1 R&D for the Fiber Positioner
Both the tilting spine and twirling-post approaches produce viable designs for DESpec, with a varying degree of R&D. In either case, R&D would be aimed at decreasing the pitch, increasing the overall light throughput, and decreasing configuration time.