Experiments on time-variation due to polarization and MMF shaking and initial results

1. Experiments on time-variation due to polarization and MMF shakingand initial results jonathan king 27-Oct-04

2. Quasi static time variation measurements • Slow speed testing on FDDI fibres: 300m link with 2 worst case connector offsets • Impulse response measurements • Gain switched DFB laser source, 65um active area 8GHz receiver • Measuring: • time variation 'envelopes' • quasi-static impulse response variation • Separate polarization controls : • manual controller 'ears' • automatic polarization scrambler (HP11896A) • Figure-of-8 'shaker' in single-mode fibre (9ply, 120mm diameter) • to be able to emulate single mode patchcord perturbation • Separate MMF fibre shaker : • Figure-of-8 'shaker' multi-mode loop (9ply, 120mm diameter) • Actuator has up to 5" travel, from 0.01 to 1 Hz cycle rate • Max acceleration at higher frequencies is ~ 0.5g

3. polarization controls (3 kinds) main fibre 2m MM patchcord TP2 TP3 Tx Rx figure-of-8 fibre shaker(s): 7um MM-MM offset (MSL patchcord) 'Worst case link' test bed: two 7um offset connections • 2 lossy connections emulated with offset spliced patchcords • one 7um offset before, one 7um after main fibre • 3 additional MMF-MMF connections at input to main fibre (with nominal offset) • Polarization controls and mechanical fibre 'shakers'

4. figure of 8 shaker linear actuator >25mm movement at up to >1Hz TIA-F04 recommends a mechanical fibre shaker made up of 3 independent figure-of-8 elements with N=3 for full fibre exploration. This test bed aims to use similar total fibre length with 1 synchronous actuator to effect full fibre exploration twist make N ply fibre loop fix

5. Proposedmeasurements Capture worst case impulse response variation for several fibres using 1 of 3 options • worst case polarization or synchronous polarisation shaker + quasi continuous step through excursion of single figure-of-8 fibre shaker (N measurements x M fibres) • quasi continuous step through excursion of two fig 8 loop shakers (N2 measurements x M fibres) • quasi continuous step through excursion of two fig 8 loop shakers (N3 measurements x M fibres) • practically, need exploration to be shrunk to a linear problem ! Purpose: • to support a generalized model of fibre dynamic impulse response, • find worst case channel response to perturbation • use GR-63-CORE to predict channel dynamic response as function of frequency

6. Initial Results: IPR envelopes 1 1 Polarization vs mechanical perturbation effect on impulse response Envelopes of impulse response with 1 axis mechanical fibre shaker on and: • input polarization 1 • input polarization 2 • input polarization scrambler on • polarization scrambler off, and second mechanical shaker added notes • 1 horizontal axis fibre mechanical shaker may not be fully exploring mode power distribution • adding 2nd scrambler in vertical axis has similar impact as polarization scrambler 2 3 4

7. Initial Results: IPR envelopes 1 Mechanical fibre shaker on and • input polarization manual ears operated • input polarization scramber on • synchronous input polarization shaker on Similar exploration of input conditions with all 3 methods

8. fibre 'shaker' test bed LF signal generator Linear actuator Driven synchronously SMF Fo8 shaker to Polarization Scrambler and manual 'ears' MMF Fo8 shaker xG DFB MM Rx

9. Quasi static IPR sequence 175mm of Fo8 movement

10. Best to worst IPR in 5mm 5mm ~ 10 Hz 5mm

11. Discussion • Synchronous SMF plus MMF shaker provides full IPR exploration • shrinks quasi static measurements to linear measurement domain • representative of a perturbed patchcord ? • Selected fibre for study: centre launch (2 dominant temporal modes) and range of causal to symmetric to anti-causal IPR variation possible • a worst case ? • Causal to anti-causal ('best to worst') IPR occurs in 5mm shaker movement • equivalent to 10Hz vibration rate at 1g