untitled BRITISH STANDARD BS EN 60793 1 49 2006 Optical fibres — Part 1 49 Measurement methods and test procedures — Differential mode delay The European Standard EN 60793 1 49 2006 has the status of[.]
Optical source
Use an optical source that introduces short duration, narrow spectral width pulses into the probe fibre
To accurately measure the differential delay time, the optical pulse must have a sufficiently short duration The maximum allowable duration, defined as the full width at 25% of maximum amplitude, is influenced by the target differential mode delay (DMD) and the sample length For instance, achieving a length-normalized DMD limit of 0.20 ps/m over a 500 m sample necessitates a measurement of 100 ps, requiring a pulse duration of less than approximately 110 ps Conversely, testing the same DMD limit over a 10,000 m fiber necessitates measuring a DMD of 2,000 ps, allowing for a pulse duration of around 2,200 ps Detailed limits are specified in section 6.1 and may vary based on the source's spectral width.
Chromatic dispersion-induced broadening due to source spectral width must adhere to the limits specified in Annex A This spectral width requirement can be satisfied by employing a spectrally narrow source or by utilizing suitable optical filtering at the source or detection end.
The centre wavelength shall be within ±10 nm of the nominal specified wavelength
A mode locked titanium-sapphire laser is an example of a source usable for this application.
Stability
Devices shall be available to position the input and output ends of the test specimen with sufficient stability and reproducibility to meet the conditions of 4.3 and 4.4.
Launch system
The probe fibre connecting the light source and test sample must support only a single mode at the measurement wavelength According to IEC 60793-1-45, the mode field diameter of the probe fibre at wavelength \$\lambda\$ is given by the formula \$(8.7\lambda - 2.39) \pm 0.5\text{ µm}\$, where \$\lambda\$ is measured in micrometers This results in a mode field diameter of 5 µm at 850 nm and 9 µm at 1,310 nm, aligning with commercially available single-mode fibres.
To achieve single-mode output from the probe fibre, one effective technique is to wrap the fibre three times around a 25-mm diameter mandrel, which helps to strip higher order modes.
The output spot of the probe fibre shall be scanned across the endface of the test sample with a positional accuracy less than or equal to ±0,5 àm
The output beam from the probe fibre shall be perpendicular to the endface of the test sample to within an angular tolerance of less than or equal to 1,0 degree
The launch system shall be capable of reproducibly centring the output spot of the probe fibre to within ±1,0 àm
When directly connected to the test sample, the distance between the output end of the probe fiber and the end face of the test sample must not exceed 10 µm.
A free space optics system utilizing lenses or mirrors can effectively image the output spot of a probe fiber onto the endface of a test sample It is crucial to ensure that the same modes are excited in the test fiber as would occur if the beam were directly coupled from the output of the single-mode probe fiber This imaging system allows for accurate coupling by projecting the output of a single-mode fiber onto the test sample's end face.
To effectively eliminate cladding light from the test sample, the fibre coating is typically adequate If additional removal is necessary, cladding mode strippers should be employed at both ends of the sample When using small weights to secure the fibre on the cladding mode strippers, it is crucial to prevent microbending at these locations.
Detection system
To ensure accurate testing, utilize an optical detection apparatus that is compatible with the test wavelength This apparatus must effectively couple all guided modes from the test sample to the detector's active area, ensuring that detection sensitivity remains consistent across modes Additionally, the detector and any associated signal preamplifier should exhibit a linear response within ±5% across the detected power range.
The temporal response of the detector system, including any optional optical attenuator, shall not be significantly mode dependent
A specific test for mode dependence is outlined in section 6.1 Additionally, the detector's temporal response can vary with offset, provided it remains stable throughout the measurement, ensuring that the requirement of ±5% for ∆T PULSE (r) as stated in section 6.1 is met.
The ringing of the detector system must be controlled to ensure that the maximum overshoot or undershoot is kept below 5% of the peak amplitude of the detected optical signal, as referenced in measurements.
The detected optical signal's waveform must be recorded and displayed using a high-speed sampling oscilloscope with a calibrated time sweep Additionally, the recording system should have the capability to average the detected waveform across multiple optical pulses.
Utilize a delay device, like a digital delay generator, to accurately time the triggering of detection electronics This device can either trigger the optical source or be activated by it, and it can be integrated within the recording instrument or exist as an external component.
To ensure accurate measurements of optical delay times, the combined impact of timing jitter and noise in the detection system must be minimized, keeping the difference between successive measurements below 5% of the measured DMD value Utilizing averaging for multiple optical pulse waveforms can effectively mitigate the effects of timing jitter and noise When employing averaging, it is essential to record each waveform with a minimum number of averages consistent with the determination of ∆T PULSE as specified in section 6.1 The system must maintain this stability throughout the measurement process.
Computational equipment
This test method generally requires a computer to store the intermediate data and calculate the test results
Test sample
The test sample shall be graded-index glass-core (category A1) multimode fibre.
Specimen endfaces
Prepare flat endfaces at the input and output ends of the specimen.
Specimen length
The length of the fibre shall be measured using a suitably accurate method such as that of IEC 60793-1-22.
Specimen packaging
Support the test fibre in a manner that relieves tension and minimizes microbending.
Specimen positioning
Position the input end of the test sample such that it is aligned to the output end of the launch system as described in 4.3
Position the output end of the test sample such that it is aligned with the detection system, as described in 4.4
Adjust and measure system response
To connect the probe fibre to the detection apparatus, you can either mount the probe directly within the apparatus or use a short fibre (less than 10 meters) that matches the type of the test fibre Alternatively, you can couple the probe output to the detector using a system of lenses and mirrors.
To ensure accurate measurements, adjust the optical pulse amplitude to align with the smallest peak amplitude anticipated from the test fiber Typically, this minimum peak amplitude is observed at the largest radial offset.
To ensure the complete capture of the pulse, adjust the detection system's time scale to align with the time scale utilized in acquiring data from the test sample (refer to section 6.2).
Measure the optical pulse waveform and determine its temporal width at 25% of the peak amplitude This measurement will be utilized to calculate the test results, referred to as the temporal width.
∆T PULSE Linear interpolation may be used between successive time points to calculate
– Repeated measurements of ∆T PULSE shall differ by no more than 5 % of the value of DMD being measured
– If using either a short length of fibre, or a system of lenses and mirrors, the values of
∆T PULSE shall differ by no more than 5 % from the values obtained by coupling the probe fibre directly into the detection apparatus
To ensure that the detector apparatus is not significantly mode dependent, prepare a short-length test sample identical to the fiber being tested Measure the value of ∆T PULSE for each radial offset used in the measurement, ensuring it meets the requirement specified in section 6.1.
Use Annex A to calculate a value of ∆T REF appropriate for the values of ∆T PULSE , source spectral width, and fibre chromatic dispersion.
Adjust detection system
To conduct the experiment, launch light from the probe fibre into the test fibre and adjust the detection system's time scale and trigger delay to display a complete optical pulse for all relevant probe spot offsets This includes capturing all leading and trailing edges with an amplitude of at least 1% of the peak amplitude Ensure that all data from the test fibre is collected without any additional adjustments to the delay and time scale.
To locate the center of the core of a test fiber, one effective method involves scanning the probe spot across the fiber's face Begin by identifying both edges of the fiber core along an arbitrary "x" axis, where the edge is defined as the point at which the total received power reaches approximately 15% of the maximum Center the probe spot along the "x" axis, then proceed to scan along the orthogonal "y" axis to find the fiber core edges and center the probe spot accordingly This process may require iteration to meet the desired positional tolerance When the probe spot is properly centered, the DMD will exhibit symmetry between positive and negative offsets along both axes Additionally, IEC 61280-1-4 outlines an alternative method for determining the optical center of the fiber.
Measure the test sample
Measure the response of the test sample, U(r,t), for radial offsets, r, of the probe spot For measurement of DMD, r ranges from R INNER ≤ r ≤ R OUTER at intervals of ≤2 àm R INNER and
R OUTER shall be provided in the specification (see item 3 in clause 9) Depending on the values specified for R INNER and R OUTER , intervals less than 2 àm may be required
If the specifications require R INNER to be 0 and R OUTER to be 17 àm, the minimum number of radial offsets needed is ten Acceptable sets of offsets include (0, 2, …, 16, 17) àm or (0, 1, …, 15, 17) àm Additionally, it is possible to utilize 18 offsets to meet the requirements.
For accurate EMBc measurements, scan from the optical center to within 1 µm of the nominal core radius, with the option to use additional radial offsets For A1a.2 multimode fiber with a 50 µm core diameter, measure U(r,t) across the range of 0 ≤ r ≤ 24 µm, ensuring intervals of 2 µm or less.
At each radial offset, measure the optical pulse waveform and identify the temporal positions of the leading and trailing edges at 25% of the maximum amplitude Utilize linear interpolation between successive time points to enhance the accuracy of the leading and trailing edge time estimates Document the leading and trailing edge times for every radial offset position.
7 Calculations and interpretation of results
The minimum effective modal bandwidth (EMB) of a fiber is defined as the least bandwidth corresponding to excitation from transmitters under specified launch conditions, as outlined in IEC 60793-2-10 This minimum EMB is determined by calculating either the differential mode delay (DMD) or the minimum calculated EMB (EMBc), ensuring that the fiber's EMB surpasses the requirements for any mode power distribution compatible with conforming transmitters The conformance of these transmitters is often defined by encircled flux requirements, which are measured according to IEC 61280-1-4.
Differential mode delay (DMD)
Find T FAST , the minimum of the leading edge times for excitation between R INNER and
R OUTER from among the output pulses recorded in 6.3
Find T SLOW , the maximum of the trailing edge times for excitation between R INNER and
R OUTER from among the output pulses recorded in 6.3
Using the value of ∆T REF from 6.1, DMD = (T SLOW – T FAST ) – ∆T REF
The minimum reporting limit for DMD, as determined by the equation, is 0.9(∆T REF) due to practical measurement challenges outlined in Annex B Therefore, if the calculated DMD value falls below 0.9(∆T REF), it should be reported as "less than 0.9(∆T REF)."
DMD can be calculated by deconvolving the reference pulse from the pulses collected from the test fiber It is crucial that the deconvolution algorithm minimizes significant errors in the pulse shapes during measurement, particularly those caused by the selection of a high-frequency noise filter.
A fiber can be defined by various DMD values, each assessed for distinct ranges of R INNER and R OUTER All DMD values can be derived from the output pulses recorded in 6.3, as long as the radial offset criteria of 6.3 are satisfied for each range of R INNER and R OUTER.
Minimum calculated effective modal bandwidth
The minimum EMBc represents the lowest value of EMBc for a specific fiber, calculated using the complete set of weightings that correspond to various mode power distributions, as outlined in sections 7.2.1 to 7.2.4.
The DMD weightings reflect the mode power distributions that align with the launch condition specifications of the optical transmitters used in the application These weightings are defined by the user's detailed specifications, which may also include an additional multiplier to adjust the effective modal bandwidth (EMBc) to meet the theoretical requirements of the application A default set of weightings is available, such as those applicable to IEEE 802.3 10GBASE.
The S and INCITS 364 10GFC standards are outlined in IEC 60793-2-10, with an example provided in Annex D Additionally, Annex C details the procedure for generating Differential Mode Delay (DMD) weights using encircled flux data.
The calculations utilize weight functions based on near-field encircled flux data from laser sources relevant to specific applications For each fiber, applying various weight functions produces multiple EMBc values, with the minimum value representing the fiber's minimum EMBc.
NOTE When DMD data are collected at offsets separated by 2 à m, the U(r,t) values at the intervening 1 à m offsets may be interpolated for the purpose of these calculations
Calculate a resultant output temporal response, P o (t) utilizing the fibre output pulse information and a weighting function
U is the sample output pulse measured at each radial offset r as a function of time t
Each output pulse is raw (un-normalized in amplitude), after an appropriate subtraction of baseline noise;
The DMD weighting function, denoted as W, is associated with the transmitter utilized in the application For detailed calculations of W(r), refer to Annex C, and for specific W(r) values related to a particular launch specification, consult Annex D.
Deconvolve the reference temporal response, R(t), from the resultant output response, P o (t), in a similar fashion to that done in bandwidth measurements described in IEC 60793-1-41
This gives the fibre frequency response, H Fib (f) , also called the fibre transfer function
P o (t) is the resultant output pulse from 7.2.1;
R (t) is the resultant reference pulse from 6.1;
FT is the Fourier transform function
NOTE These calculations yield an array of complex numbers
7.2.3 Calculated effective modal bandwidth (EMBc)
To calculate the -1.5 dB optical bandwidth, identify the lowest frequency at which the magnitude of the transfer function is 1.5 dB below the zero frequency value This -1.5 dB optical value is then extrapolated to -3 dB using Gaussian assumptions by multiplying it accordingly.
The bandwidth is traditionally defined by the 3 dB point, where the transfer function, H Fib (f), reaches 50% or 3 dB However, real fiber and sources can produce highly non-Gaussian responses, making the measured 3 dB value less indicative of system performance To address the limitations of a wavy transfer function and its impact on the –3 dB value, the 1.5 dB metric offers a more accurate assessment.
7.2.4 System stability frequency limit (SSFL)
R A and R B are two independent reference pulses;
SSFL is the minimum frequency where |G(f)| exceeds 1,0 ± 0,05 (see 60793-1-41)
If the EMBc calculated for a fibre/laser combination exceeds the SSFL, report the normalized bandwidth value as greater than SSFL × length.
Length normalization
Normalizing the values of DMD or EMBc to a unit length, such as ps/m or MHz⋅km, can be beneficial When this normalization is applied, it is essential to report the length dependence formula.
Report the following information for each test
– length normalization formula, if used;
– source wavelength (nominal or actual);
– minimum and maximum radial offsets, R INNER , R OUTER;
– test result: DMD(R INNER , R OUTER ) and/or minimum EMBc.
The following information shall be available upon request
– description of the test equipment, including: source type and actual source centre wavelength, maximum specified or actual spectral width (r.m.s.);
– for DMD measurement, documentation of method used to calculate ∆T REF For minimum
EMBc, the transfer function features that are used to determine bandwidth, and the set of weightings used;
– detector type and operating conditions;
– mode field diameter of probe fibre at measurement wavelength (nominal or actual);
– method of stripping cladding light;
– date of latest calibration of test equipment
When specifying fibre performance using this test method, the following information shall be specified:
– number and type of samples to be tested;
The DMD requirements specify the necessary DMD value for a defined range of minimum and maximum radial offsets, denoted as DMD(R INNER, R OUTER) It may be necessary to evaluate various DMD values across different specified ranges for R INNER and R OUTER.
– for DMD measurements, reporting method option from 7.1;
– for EMBc requirements: Required minimum EMBc value;
– for EMBc requirements: Required set of weights per Annex C
A.1 Limiting the effect of chromatic dispersion on the value of DMD
The effect of errors introduced by chromatic dispersion on the value of DMD shall be less than
To meet the 10% requirement, one can either utilize a source with a sufficiently narrow spectral width to disregard chromatic dispersion or precisely assess the spectral shape of the source to calculate the correct value of ∆T REF.
The chromatic dispersion D(λ) may be estimated using the data given in Clause A.2
One can utilize D(λ) from IEC 60793-1-42 for the specific fibre type under examination To satisfy the spectral width requirement, a spectrally narrow source can be employed, or an optical filter can be applied at the source or detection end.
Several examples of methods for meeting the requirement of this annex are now given
A.1.1 Use a source with sufficiently narrow spectral width such that the value of
∆chrom 4 ln( 2 ) δλ λ (A.1) is less than 10 % of the DMD to be measured This gives a constraint on r.m.s spectral width δλ,
The minimum value of DMD, denoted as DMD min, is determined by the chromatic dispersion D(λ) and the sample length L Assuming that the modal delays in a fiber scale linearly with length, this constraint remains independent of length.
Use ∆T REF = ∆T PULSE in 6.1 and for calculating the value of DMD
Example: DMD values as small as 100 ps are to be tested on fibre lengths of 0,5 km at a wavelength of 850 nm From Table A.1 in A.2, the value of D(λ) at 850 nm is
The source's r.m.s spectral width, δλ, must be ≤ 0.056 nm when substituting the values into equation (A.2), calculated as (0.03 × 100 ps)/(107 ps/nm-km × 0.5 km) Additionally, this source is capable of functioning over a test length of 10 km with DMD values as low as 2,000 ps.
A.1.2 Use a source with sufficiently narrow spectral width that ignoring ∆t chrom in relation to
∆T PULSE changes the value of ∆T REF by less than 10 % This gives a constraint on r.m.s spectral width δλ,
Use ∆T REF = ∆T PULSE in 6.1, and for calculating the value of DMD
The minimum measurable value of DMD is determined directly, without any explicit dependence on the source spectral width δλ.
The maximum allowable spectral width is explicitly dependent on the sample length When the spectral width is fixed, chromatic broadening becomes significant and cannot be overlooked beyond a certain sample length.
A laser source and optical detector with a pulse duration of ∆T PULSE = 60 ps are utilized to measure samples over a distance of 0.5 km at a wavelength of 850 nm By applying this data to equation (A.3), the root mean square (r.m.s.) spectral width of the source, denoted as δλ, must be less than or equal to 0.138 times a specific value.
60 ps)/(107 ps/nm-km × 0,5 km) = 0,15 nm
To determine the suitable value of ∆T REF for the utilized source, calculate the full width at 25% of each mode at the output of the fiber under examination For near-Gaussian pulse and spectral shapes, apply the appropriate methods to obtain accurate results.
The upper limit on the spectral width of the source is indirectly determined by the requirement in section 7.1, which specifies that the minimum value of DMD reported by the measurement must be met.
When dealing with sources that exhibit multiple spectral peaks or significant non-Gaussian characteristics, the accuracy of the formula may be compromised To ensure that the calculation of DMD remains reliable, it is essential that the error introduced by using equation (A.4) to determine the value of ∆T REF does not exceed 10%.
A.2 Chromatic dispersion in multimode fibres
Table A.1 displays the highest expected dispersion for commercially available Category A1 fibres, determined by nominal dispersion performance and numerical aperture (NA) For wavelengths below 1,200 nm, the highest dispersion occurs in fibres with a maximum zero-dispersion wavelength (\(\lambda_0\)) of 0.29 NA Conversely, for wavelengths exceeding 1,400 nm, the greatest dispersion is found in fibres with a minimum \(\lambda_0\) of 0.20 NA.
Table A.1 is not used for wavelengths between 1 200 and 1 400 nm Instead, use D 16,6 ps/nm-km
Table A.1 –Highest expected dispersion for any of the commercially available Category A1 fibres λ
– for λ < 1 200 nm: S 0 =0,09562 ps/(nm 2 ãkm); λ 0 = 1 344,5 nm for a nominal multimode fibre with 0,29 NA;
– for λ > 1 400 nm: S 0 =0,101 ps/( nm 2 ãkm); λ 0 = 1 310 nm for a nominal multimode fibre with 0,20 NA
This standard aims to assess the variation in delay times between the fastest and slowest mode groups under specific excitation conditions, as shown in Figure B.1 The resulting DMD data from these measurements can be compared to established DMD specifications, which have been derived from both modeling and experimental results to ensure a minimum EMB across various transmitters.
Leading and trailing edge times (25% threshold) are identified with “+” Traces are offset for different excitation positions Inset shows ∆T PULSE
At any offset position, a single-mode probe excites multiple mode groups, resulting in a complex output waveform, U(r,t), characterized by multiple peaks and potential overlap of modes To effectively detect these modes, the detection level is set at 25% of the peak amplitude of the waveform This approach ensures that even when a maximally excited mode group is temporally separated from others, it can still be identified, despite the possibility of overlapping amplitudes from modes with the same delay time.
The disparity between T SLOW and T FAST exceeds the DMD, influenced by the temporal width of the optical pulse, the limited bandwidth of the optical detector, and the broadening of each mode caused by the source's spectral width and the chromatic dispersion of the tested fiber.
The temporal width of the optical pulse and the finite detector bandwidth are characterized as