Expand this Topic clickable element to expand a topic
Skip to content
Optica Publishing Group

Simple-structure low-loss multi-core fiber LC connector using an align-by-contact method

Open Access Open Access

Abstract

We present a novel single-MCF connector without any additional or high-precision parts not found on a standard single-SMF connector. The proposed connector realizes rotational fiber alignment and ferrule floating simultaneously by employing a standard MU ferrule with a straight flange edge and a modified LC housing with a tapered hole that can make contact with the ferrule flange. Fabricated connectors achieved an average loss of 0.07 dB in a random connection test and passed the Telcordia GR-326-CORE mechanical and environmental reliability test. Furthermore, we conducted numerical simulations and confirmed these positive results were due to suppression of ferrule rotation from external forces.

© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement

1. Introduction

The rapid growth of optical network traffic has dramatically increased the demands for high bandwidth optical interconnects in large-scale data centers [1]. Today multi-fiber MPO connectors are widely deployed in data centers to splice multiple optical channels and create higher density. While an MPO connector [1] has been popular for multi-fiber connection in data center optical networks, larger pressing force and/or stricter control of fiber protrusion would be a challenge for physical-contact (PC) in case of higher fiber count [2]. Besides, installation and reconnection of MPO connectors are troublesome due to the difficulty of end face cleaning due to debris created by the insertion of connector alignment guide pins and the high mating force for physical contact of all optical fiber channels. A single-fiber connector with a multi-core fiber (MCF) is a possible solution to realize high-density multi-channel splicing with easy end-face cleaning and low mating force similar to today’s conventional single-fiber connectors [2]. However, since MCF has outer cores located in the non-center part of the cladding, precise rotational fiber alignment and ferrule floating from the connector housing needs to be implemented simultaneously in order to realize low insertion loss (IL) and a stable single-MCF connection. To address these challenges, various custom connectors with additional and/or higher-precision parts have been reported [37], as summarized in Table 1. However, such additional and/or high-precision parts would make the volume manufacturing of single-MCF connectors much more difficult compared to that of conventional single-fiber connectors with single-mode fibers (SMFs).

Tables Icon

Table 1. Comparison of previously-reported and present single-MCF connectors.

In this paper, we present in details a simple-structure single-MCF connector that does not require any additional or higher-precision components, whose preliminary fabrication results were presented in [8]. The present connector simultaneously realizes low insertion loss and ferrule floating, thanks to an Align-by-Contact method where the ferrule is aligned by its direct contact to the inner surface of the connector housing in an unmated state and the ferrule is floated from the housing in a mated state. Except for fiber rotation alignment in the ferrules, every part of the present connector can be manufactured with the volume-manufacturing facilities for conventional single-fiber connectors. We conducted Telcordia GR-326-CORE mechanical tests and environmental tests with the fabricated connectors, and the results demonstrate the reliability of the present connectors under external stresses. A concern with the Align-by-Contact method is that the floated ferrule could be rotated by external stress like a cable twist, but our simulation result shows such a ferrule rotation requires strong torque to the ferrule which cannot be transferred to the ferrule. From the results, we successfully confirmed low loss connection performance and the reliability of the proposed simple Align-by-Contact structure.

2. Connector structure for rotational alignment with floating

To realize an IL less than 0.5 dB in a single-MCF connector, the rotational misalignment between connectors has to be suppressed and a structure that allows ferrule floating have to be achieved simultaneously. The misalignment has to be less than 1 degree if using a 125-µm-cladding 8-core fiber where the cores are circularly located at the points distant from the cladding center by 40.5 µm and have a mode field diameter of 8.3 to 8.5 µm at a wavelength λ of 1310 nm [9]. Figure 1 shows a conventional LC connector structure. The clearance between the LC front housing and ferrule flange allows ferrule floating which isolates fiber from external force, but induces 10-degree ferrule rotation at maximum. Thus, the structures of the front housing and the ferrule flange need to be optimized to realize a single-MCF connector with small rotational misalignment. To realize an LC-interface MCF connector without improving dimensional accuracies of components and without any additional parts for rotational alignment, we employed the Align-by-Contact method (ABC method) whereby the MCF ferrule flange is rotationally aligned by contact to the housing in an unmated state, and floated from the housing by being pushed back by the opposing connector in a mated state. Figure 2 shows the difference between standard LC connectors and the proposed MCF-LC connector, and the conceptual connector structure that we designed based on the described alignment–floating scheme. We utilized an MU ferrule flange (Fig. 2(b)) instead of an LC ferrule flange (Fig. 2(a)) because the longer straight edge of the rectangular MU flange is suitable to suppress ferrule rotation, compared to an LC flange with conical and hexagonal portions. We modified the interior of the LC front housing such that the hole for the flange has a tapered interface for the MU flange. The straight edge of the MU flange can make contact with the tapered interior of the hole, and thus the ferrule rotation angle is fixed in the unmated state, as shown in Fig. 2(c). The ferrule flange is floated from the housing when the ferrule is pushed by the opposing ferrule during connector mating, as shown in Fig. 2(d). The only non-standard component of the proposed connector is the front housing with the tapered hole that does not require higher dimensional precision.

 figure: Fig. 1.

Fig. 1. (a) Appearance, (b) cross section, and (c) clearance between the ferrule flange and the housing of a conventional LC connector [8].

Download Full Size | PDF

 figure: Fig. 2.

Fig. 2. Schematics of (a) standard LC connector, and (b) the proposed MCF-LC connector in (c) unmated state and (d) mated state [8]. The differences between (a) and (b) are ferrule type and the front housing which has a tapered hole for the standard MU flange. When the connectors are unmated, the ferrule flange is precisely rotationally aligned by the taper. When the connectors are connected, the clearance makes it possible for the ferule to float.

Download Full Size | PDF

3. Fabrication results

We fabricated LC-interface MCF connectors with the alignment–floating scheme proposed above, using a 125-µm-cladding 8-core fiber [9]. The front housing with LC interface was modified to have a tapered hole for the flange–housing contact and floating feature, and fabricated using an ordinary injection molding method without any high-precision machining. A standard MU flange was utilized as mentioned above. The other components are standard LC connector parts. Thus, the present connector is compatible with high-volume production processes for conventional single-fiber connectors and the external dimensions of the connectors are compliant with the IEC 61754-20 standard for LC connectors. We rotationally aligned an MCF to an MU ferrule with a misalignment of less than 0.20 degrees by monitoring the MCF end face.

To suppress Return Loss (RL), PC has to be achieved between each mated pair of the MCF cores. However, it becomes difficult for the non-center cores of the MCF to achieve PC under the standard requirements for the connector end face dimensions [10,11]. By using a finite element method, we studied the PC condition of an 8-core fiber with a 125-µm cladding at a mating force of 5 to 6 N which is the IEC standard mating force. The PC condition of the 8-core fiber is more severe because the cores are located 40.5-µm apart from the fiber center. We modeled the ferrule end shape with the following parameters: fiber withdrawal U, mating force F, curvature radius R, apex offset of MCF end faces d, as shown in Fig. 3(a). We set U and F as 50 nm and 5 N, respectively, which are typical values in standard LC SMF connectors. The calculation results are shown in Fig. 3(b). The results show that a small apex offset enables PC in the 8-core fiber when a standard connector spring is used. Based on the results, we fabricated the proposed connector with a small apex offset by optimizing ferrule polishing conditions.

 figure: Fig. 3.

Fig. 3. (a) Parameters relating to the single-MCF connector’s PC (b) Calculated PC area of 125µm-8core fiber.

Download Full Size | PDF

To confirm that the fabricated connectors simultaneously have a ferrule floating function and precise rotational alignment, we conducted Telcordia GR-326-CORE mechanical tests with one pair of the fabricated connectors. Figure 4(a) shows the appearance of the fabricated connector and axis definition, and Fig. 4(b) shows the jumper test apparatus to apply controlled bend θ, tension T, and twist φ. Table 2 summarizes the test results. Figure 5 shows the results of “Transmission with applied tensile load” tests, which indicate an IL change when the tensile load is applied to a boot with a bend θ of 90°. The IL variation was less than 0.37 dB, which indicates the lateral and/or tilted misalignments caused by external force were suppressed by ferrule floating. Figure 6 shows the IL and RL variations during the durability test of 200 mating cycles. We confirmed an IL of less than 0.40 dB and an IL variation of less than 0.40 dB under all the mechanical tests. We also confirmed that the RLs were more than 40 dB for all the cores.

 figure: Fig. 4.

Fig. 4. (a) The appearance of the fabricated MCF-LC connector and axis definition (b) Jumper test facility for the mechanical tests.

Download Full Size | PDF

 figure: Fig. 5.

Fig. 5. The IL variation during the transmission with applied tensile load test: (a) x-axis, θ = 90°, (b) y-axis, θ = 90°.

Download Full Size | PDF

 figure: Fig. 6.

Fig. 6. Results of the durability test: (a) the IL variation and (b) the RL variation.

Download Full Size | PDF

Tables Icon

Table 2. Results of the mechanical tests of the fabricated MCF connectors at λ of 1310 nm.

We confirmed that ILs were suppressed even after various stresses were applied in the mechanical tests, although the connector structure does not have a ferrule rotation aligning mechanism during ferrule floating. One might be concerned that the twisting to the cable induces torque to the floated ferrule hence results in rotation misalignment, but we consider that ferrule rotation due to torque is suppressed by the friction between the ferrule and other parts making contact with each other, even in the mated state. To estimate the torque required to rotate the ferrule, we performed a numerical simulation. Figure 7(a) shows the schematics of the simulation model. We assumed that the resistances to the rotation is the friction between the ferrule flange and the spring, between the ferrule and the split sleeve in the adapter, and between the mated ferrule end faces. In the simulation, constantly increasing torque was applied to one of the mated ferrules, and the rotation angle difference between the mated ferrules was evaluated. The result is shown in Fig. 7(b). The rotation of the ferrule starts when the applied torque exceeds the static friction force. Once the torque exceeds the threshold of the static friction force, the friction force changes to dynamic friction force, which is lower than the static one and the ferrule rotates rapidly. We found that the threshold is around 0.7 N·mm. Such a torque is unlikely to be applied to the ferrule, since the torque is transmitted to the ferrule from the cable through the rear housing and the spring but the rear housing cannot rotate so much as it is fixed to the front housing.

 figure: Fig. 7.

Fig. 7. (a) Schematic model and (b) results of the simulation of the torque-induced ferrule rotation.

Download Full Size | PDF

We also conducted Telcordia GR-326-CORE environmental tests with two pairs of the fabricated connectors. We monitored the IL and RL for 8 cores of each connector pair. The results are summarized in Table 3, and Fig. 8 shows the IL and RL variations during Humidity Condensation Tests. We confirmed an IL of 0.44 dB at maximum and IL variations of 0.12 dB at maximum under all the tests. We also confirmed the RL were more than 40 dB and the RL variations were less than 5 dB under the all the tests.

 figure: Fig. 8.

Fig. 8. Result of Humidity condensation cycle test (a) the IL variation, (b) the RL variation.

Download Full Size | PDF

Tables Icon

Table 3. Results of the environmental tests of the fabricated MCF connectors at λ of 1310 nm.

Finally, we conducted a random mating test to evaluate the IL characteristics of the fabricated MCF connectors. Figure 9 shows the IL histogram for 112 random connections with 3 mating cycles of the 8 cores of the MCFs, wherein 7 connectors were randomly selected from 20 connectors, and mated to each other and to the unselected 13 connectors (7C2 + 7×13 = 112 connections).These connectors had same MCF direction, and we connected the same core IDs. The IL was 0.07 dB in average, ≤0.21 dB for >97% of the samples, and 0.39 dB at maximum. The result is compatible to IEC 61753-1 Grade B for low-loss SMF connectors [8].

 figure: Fig. 9.

Fig. 9. The IL histogram for the random mating test (Sample size 2688 = 112 random connection pairs × 3 mating cycle × 8 cores). [8]

Download Full Size | PDF

4. Conclusion

We proposed and demonstrated a novel simple-structure single-MCF connector. The present connector has a standard LC-interface and does not require any additional or higher-precision components and simultaneously realizes low insertion loss and reflection. Rotational fiber alignment and ferrule floating are realized by employing a standard MU ferrule with a straight flange edge and a modified LC housing with a tapered hole that can make contact with the ferrule flange. Every component of the present MCF connector is compatible to volume-manufacturing using conventional manufacturing facilities for single-fiber connectors. We conducted Telcordia GR-326-CORE mechanical and environmental tests. We confirmed that the IL and the RL were suppressed under the criteria of GR-326-CORE in all the tests. We also conducted a random connection test and confirmed that the IL distribution of the present connector is compatible to IEC 61753-1 Grade B for low-loss SMF connectors. These results assured that the proposed design has precise rotational fiber alignment and a structure that allows ferrule floating. The present connector structure can provide a volume-manufacturable and economically-viable solution for low-loss and high-density multi-channel optical connections with easy handling.

Funding

National Institute of Information and Communications Technology.

Acknowledgments

This research is supported in part by the National Institute of Information and Communications Technology (NICT), Japan.

Disclosures

The authors declare no conflicts of interest.

Data availability

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.

References

1. C. Kachris and I. Tomkos, “A Survey on Optical Interconnects for Data Centers,” IEEE Commun. Surv. Tutorials 14(4), 1021–1036 (2012). [CrossRef]  

2. T. Hayashi, T. Nagashima, T. Morishima, Y. Saito, and T. Nakanishi, “Multi-core fibers for data center applications,” in Eur. Conf. Opt. Commun. (ECOC) (2019), paper M.1.D.6.

3. R. Nagase, K. Sakaime, K. Watanabe, and T. Saito, “MU-Type Multicore Fiber Connector,” IEICE Trans. Electron. E96.C(9), 1173–1177 (2013). [CrossRef]  

4. T. Morishima, O. Shimakawa, J. Ito, T. Shimazu, H. Arao, T. Yokochi, F. Uehara, M. Ohmura, T. Nakanishi, T. Sano, and T. Hayashi, “Ultra-High-Density MCF Connector Technology,” in Opt. Fiber Commun. Conf. (OFC) (2018), paper W1A.5.

5. T. Kobayashi, M. Watnabe, and Y. Minagawa, “Multi-core fiber free space coupling type Connecter,” in Proc. IEICE General Conference (2013), paper C-3-34 (in Japanese).

6. K. Saito, T. Matsui, K. Nakajima, and T. Kurashima, “Multi-Core Fiber Connector with Precise Rotational Angle Alignment,” in OECC/ACOFT (2014), paper TH10B-3.

7. Y. Lee, K. Tanaka, E. Nomoto, H. Arimoto, and T. Sugawara, “Multi-core fiber technology for optical-access and short-range links,” in 12th Int. Conf. Optical Internet (COIN) (2014), paper TB2-4.

8. T. Morishima, K. Manabe, S. Toyokawa, T. Nakanishi, T. Sano, and T. Hayashi, “Simple-Structure LC-Type Multi-Core Fiber Connector with Low Insertion Loss,” in OFC (2020), paper Th3I.2.

9. T. Hayashi, T. Nakanishi, K. Hirashima, O. Shimakawa, F. Sato, K. Koyama, A. Furuya, Y. Murakami, and T. Sasaki, “125-µm-cladding eight-core multi-core fiber realizing ultra-high-density cable suitable for O-Band short-reach optical interconnects,” J. Lightwave Technol. 34(1), 85–92 (2016). [CrossRef]  

10. O. Shimakawa, M. Shiozaki, T. Sano, and A. Inoue, “Pluggable Fan-out realizing Physical-contact and low coupling loss for Multi-core fiber,” in OFC (2013), paper OM3I.2.

11. K. Shikama, Y. Abe, S. Awakawa, S. Yanagi, and T. Takahashi, “Multicore Fiber Connector with Physical-Contact Connection,” IEICE Trans. Electron. E99.C(2), 242–249 (2016). [CrossRef]  

Data availability

Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.

Cited By

Optica participates in Crossref's Cited-By Linking service. Citing articles from Optica Publishing Group journals and other participating publishers are listed here.

Alert me when this article is cited.


Figures (9)

Fig. 1.
Fig. 1. (a) Appearance, (b) cross section, and (c) clearance between the ferrule flange and the housing of a conventional LC connector [8].
Fig. 2.
Fig. 2. Schematics of (a) standard LC connector, and (b) the proposed MCF-LC connector in (c) unmated state and (d) mated state [8]. The differences between (a) and (b) are ferrule type and the front housing which has a tapered hole for the standard MU flange. When the connectors are unmated, the ferrule flange is precisely rotationally aligned by the taper. When the connectors are connected, the clearance makes it possible for the ferule to float.
Fig. 3.
Fig. 3. (a) Parameters relating to the single-MCF connector’s PC (b) Calculated PC area of 125µm-8core fiber.
Fig. 4.
Fig. 4. (a) The appearance of the fabricated MCF-LC connector and axis definition (b) Jumper test facility for the mechanical tests.
Fig. 5.
Fig. 5. The IL variation during the transmission with applied tensile load test: (a) x-axis, θ = 90°, (b) y-axis, θ = 90°.
Fig. 6.
Fig. 6. Results of the durability test: (a) the IL variation and (b) the RL variation.
Fig. 7.
Fig. 7. (a) Schematic model and (b) results of the simulation of the torque-induced ferrule rotation.
Fig. 8.
Fig. 8. Result of Humidity condensation cycle test (a) the IL variation, (b) the RL variation.
Fig. 9.
Fig. 9. The IL histogram for the random mating test (Sample size 2688 = 112 random connection pairs × 3 mating cycle × 8 cores). [8]

Tables (3)

Tables Icon

Table 1. Comparison of previously-reported and present single-MCF connectors.

Tables Icon

Table 2. Results of the mechanical tests of the fabricated MCF connectors at λ of 1310 nm.

Tables Icon

Table 3. Results of the environmental tests of the fabricated MCF connectors at λ of 1310 nm.

Select as filters


Select Topics Cancel
© Copyright 2024 | Optica Publishing Group. All rights reserved, including rights for text and data mining and training of artificial technologies or similar technologies.