Differences

This shows you the differences between two versions of the page.

--- fiber_run [2024/09/11 19:46] – [Fiber cladding and termination] jasonj
+++ fiber_run [2026/05/26 15:45] (current) – 24.4.246.2
@@ Line 3: / Line 3: @@
 ===== Fiber cladding and termination =====
-This is being carried out at FiberTech Optica in Canada. The 50 micron spool left over from SONG Australia was sent there in July 2024.
+We inherited fiber from the Australia SONG project. This was about 500 meters of octagonal (50/70/95) fiber. We signed a contract with FiberTech Optica (FTO) in Canada to construct 2 separate 83m cables (for redundancy), each with 3 fibers in it (for science fiber, cal fiber, and extra fiber), with a ferrule on one end and a FC connector on the other, and sent them the fiber in July 2024.
+However, FTO has had issues with that fiber:
+- for a short (2m) segment, they measured 25% lower throughput than the same length of circular fiber
+- for an 83m segment, they measured 80% throughput at 632nm (not so far from the 86% that we were anticipating)
+- most troubling, they say the fiber came off the spool very twisted and stressed, so that it coils itself back up, and is very difficult to work with. Injecting a slow beam, they estimate 40% loss in an output f/6 beam, and thus, likely worse for an input f/6 beam (which they will test).  Since they used
+the same spool for the Australia SONG node a few years ago, they looked up the notes for that, and found that the technician who handled that fiber described it as "devil fiber", so it perhaps had the same issues back then. They did not recall doing FRD testing at the time. If FRD was not tested, this perhaps might have implications for performance at the Australia node.
+Because of this, we decided not to proceed with the octagonal fiber for the full run. However, Fiber Tech Optica has developed a method of fusing fiber together with apparently good success. Based on their test results, we decided to proceed by fusing a short section of the octagonal fiber to a long section of circular fiber, in order to maintain scrambling. Unfortunately, 50 micron circular fiber was not readily available in the lengths required at the time. As a result, we decided to proceed with a long run of 53 micron circular fiber. At the telescope end, FTO fused a short section of the 50 micron octagonal fiber, and in front of that,  short section of 50/125 circular fiber for a good match in the FC connector at the telescope.
+The cable has three fibers that are terminated in a ferrule at the spectrograph end, and are separated into three individual fibers with FC connectors at the telescope end. Two full fibers were constructed by FTO. FTO provided a {{::fiber:final_report_pf_02_compressed_1_.pdf|test report}} on these two cables; the first one that the constructed had poorer performance than the second one. At APO, we pulled the better one through the conduit, and the other one has remained on its spool.