raising standards worldwide™NO COPYING WITHOUT BSI PERMISSION EXCEPT AS PERMITTED BY COPYRIGHT LAW BSI British Standards Optical amplifiers — Test methods — Part 10-1: Multichannel param
Calibration
Calibration of OSA power measurement
Calibrate the OSA power measurement by using a calibrated power meter
The calibrated optical power meter measures all optical power, including source spontaneous emission, while the Optical Spectrum Analyzer (OSA) only detects power within its resolution bandwidth To enhance calibration accuracy, it is advisable to use an optical filter with a full width at half maximum (FWHM) passband of 1 nm to 3 nm at the output of the optical pulse source.
Calibration of the pulse duty ratio
To calibrate the pulse duty ratio, first activate a channel of the optical pulse source at continuous wave (CW) mode with the specified power and wavelength Next, configure the pulse width (\$T_{source}\$) and pulse interval (\$T\$) according to the product specifications, ensuring that both \$T_{source}\$ and \$T\$ are significantly shorter than the gain-response time of the optical amplifier (OA) being tested For Erbium-Doped Fiber Amplifiers (EDFAs), typical values for \$T_{source}\$ and \$T\$ are 0.4 μs.
`,,```,,,,````-`-`,,`,,`,`,,` - and 1 μs (‡), respectively These values, however, depend on the amplifier saturation condition
Measurement accuracy in relation to pulse rates is detailed in Annex B, while Annex A provides EDFA output waveforms for various EDFAs To assess performance, measure the average output power, \$P_{\text{pulse-ave}}\$ using a power meter Next, drive the optical pulse source with a 100% duty pulse (DC drive) and measure the output power, \$P_{\text{DC}}\$ with a power meter Finally, calculate the equivalent duty ratio using Equation (1).
DR source DC ave - pulse
NOTE For the optical pulse source using an external optical switch, the calibration result is applicable to the other channels
For the optical pulse source using direct modulation, the calibration shall be repeated for all the channels, because the optical-pulse shape generated by each source can be different.
Calibration of the sampling module
Follow the steps below to calibrate the sampling module a) Arrange the optical pulse source, sampling SW, OSA and calibrated power meter as shown in Figure 6
To calibrate the sampling switch, first activate the optical pulse source to emit continuous wave (CW) light at the desired channel wavelength Next, configure the optical spectrum analyzer (OSA) bandwidth, \$B_o\$, to match the spectral bandwidth of the pulse signal, and adjust the OSA center wavelength accordingly Set the sampling pulse width, \$T_{sampler}\$, as per product specifications, ensuring that the combined duty ratios of the source and sampler remain below 100%, ideally between 80% and 90%, with \$T_{sampler}\$ being less than \$T_{source}\$ Measure the average output power, \$P_{OSA-pulse-ave}\$, using the OSA, then drive the sampling switch with a 100% duty pulse for DC drive and measure \$P_{OSA-DC}\$ Finally, calculate the equivalent sampling switch duty ratio using the appropriate equation.
DR sampler DC - OSA ave pulse
The DR sampler value obtained at one channel wavelength is applicable to other channel wavelengths First, measure the input power to the sampling switch, denoted as \$P_{\text{CW-calibd}}\$ using a calibrated power meter Next, activate the optical pulse source to emit continuous wave (CW) light at the subsequent channel wavelength for testing.
Repeat steps g) through i) for the next channel wavelength to be measured k) Calculate the calibration factor, CAL(λk), of the sampler including the OSA by using
Calibration of dynamic isolation
6.1.4.1 Timing adjustment of the sampling switch (SW)
To adjust the timing of the sampling switch, first connect the optical pulse source to the sampling switch and optical spectrum analyzer (OSA) using a fiber cord, as depicted in Figure 7, where the optical pulse source is labeled as source a.
Optical pulse source b is also applicable here
Narrow band optical source OSA
Figure 7 – Arrangement for timing adjustment b) Activate the optical pulse source to emit light at all channel wavelengths
The current testing procedure activates all channels to enhance the stabilization of the multichannel optical pulse source for subsequent measurement stages Additionally, it is essential to adjust the Optical Spectrum Analyzer (OSA) to a specific channel wavelength and configure the drive pulse timing for both the optical pulse source and the sampling switch as outlined.
The DR sampler must be smaller than the DR source To optimize the received optical power with the Optical Spectrum Analyzer (OSA), determine the minimum delay time, T d-min, by adjusting the CH2 delay time, T d Additionally, calculate the maximum delay time, T d-max, that enhances the received optical power using Equation (4).
NOTE The delay time thus obtained at one channel wavelength is applicable to the other channel wavelengths
Leakage signal from source module
Figure 8 – Timing adjustment of the sampling switch
Follow the steps below to calculate the dynamic isolation a) Keep activating the optical pulse source to emit light pulses at all channel wavelengths
To accurately measure dynamic isolation, it is essential that all channels remain active, as the Optical Spectrum Analyzer (OSA) must capture optical powers from both the targeted and adjacent channels Begin by connecting the optical pulse source to the sampling switch and OSA using a fiber cord, as illustrated in Figure 7 Next, configure the sampling switch timing according to Figure 9a and measure the average signal power, P Sig OSA-ave, with the OSA tuned to the desired channel Then, adjust the sampling switch timing as per Figure 9b and measure the average leakage power, P Leak OSA-ave, while keeping the OSA tuned to the same channel Repeat these measurements for each channel under test Finally, calculate the average dynamic isolation for each channel, denoted as ISO(λ k ) dyna-ave, using the appropriate formula.
ISO(λ k ) dyna-ave ave OSA ave OSA Sig
Leakage signal from source module
Figure 9a – Measurements of P OSA-ave Sig Figure 9b – Measurements of P OSA-ave Leak
Figure 9 – Timing chart for dynamic isolation calibration
OA measurement
Timing adjustment for ASE and amplified signal power measurement
Follow the steps below to adjust the timing for ASE and amplified signal power measurement a) Keep activating the optical pulse source to emit pulsed light at all channel wavelengths
To ensure stability in the multichannel optical pulse source, all channels remain activated, even if timing adjustments are made using a single channel Connect the optical pulse source, the optical amplifier (OA) under test, the sampling switch, and the optical spectrum analyzer (OSA) as depicted in Figure 10, with the option to use either optical pulse source a or b Activate the OA according to the detailed specifications while preventing surge generation Tune the OSA to a selected channel wavelength and configure the drive pulse timing for both the optical pulse source and the sampling switch as indicated.
In Figure 10, the sampling switch is operated out of phase with the optical pulse source for the purpose of measuring Amplified Spontaneous Emission (ASE) To optimize the measurement, it is essential to determine the delay time \( T_{d-ASE} \) that minimizes the average ASE power \( P_{ASE\ OSA-ave} \) by adjusting the CH2 delay time \( T_d \) Additionally, the delay time \( T_{d-sig} \) should be calculated to maximize the output signal power \( P_{Sig-OA-out\ OSA-ave} \) using Equation (6).
NOTE The delay time thus obtained at one channel wavelength is applicable to the other channel wavelengths
For gain measurement For ASE measurement
Switch time chart Typically at 1 MHz
Narrow-band dB optical source λ1 ∼ λN
Figure 10 – Arrangement for OA measurement
ASE measurement
To measure the ASE, first activate the optical pulse source to emit pulsed light across all channels Next, adjust the average signal power for each channel entering the optical amplifier (OA), denoted as \$P_{OA-in-ave}\$, according to the detailed specifications This adjustment can be accomplished using an optical spectrum analyzer (OSA).
1) Connect the optical pulse source and the sampling switch with a fibre cord
2) Set the sampling switch timing: T d-max, as given in Equation (4)
3) Measure P Sig OSA-ave with the OSA at the wavelength under test
4) P OA-in-ave is given in Equation (7)
P(λk)OA-in-ave sampler k source
For single-channel applications, the average output power, P OA-in-ave, can be adjusted using a calibrated power meter instead of following the previous steps Additionally, the timing for the sampling module should be set according to item e) of section 6.2.1 to accurately measure the ASE power, as illustrated in the timing chart in Figure 11 Finally, the average ASE power, P ASE OSA-ave, should be measured with the OSA at the channel being tested.
The power measurement is influenced by the resolution bandwidth of the Optical Spectrum Analyzer (OSA) To measure the average amplified spontaneous emission (P ASE OSA-ave), position the OSA at the subsequent channel for testing while maintaining consistent conditions.
Amplified signal power measurement
a) Keep activating the optical pulse source to emit pulsed light at all channels b) Set the sampling switch timing as determined by step g) of 6.2.1 to measure the signal power
The timing chart is illustrated in Figure 12 It is essential to maintain the average power level of P OA-in-ave across all channels consistent with the ASE measurement Additionally, the average output power P sig-OA-out OSA-ave should be measured using the OSA at the specific wavelength being tested.
`,,```,,,,````-`-`,,`,,`,`,,` - e) Measure P sig-OA-out OSA-ave with the OSA at the next channel to be tested while keeping other conditions unchanged
Figure 11 – Timing chart for ASE measurement
Figure 12 – Timing chart for amplified signal power measurement
General
The calculation must be performed for each channel individually, as the parameter values vary depending on the channel being tested.
P(λk) OA-in-ave Average input signal power, mW
P(λk) ASE OSA-ave Average ASE power measured with the OSA, mW
P(λk) sig-OA-out OSA-ave Average output signal power from OA measured with the OSA, mW
ASE(λk, B 0 ) ASE power within the optical bandwidth of the OSA, mW
CAL(λk) Calibration factor of the sampler plus OSA
ISO(λk) dyna-ave Average dynamic isolation, dB
F(λk) sig-sp Signal-spontaneous noise factor (expressed in linear form)
NF(λk) sig-sp Signal-spontaneous noise figure, dB
Noise factor calculation
Noise factor, F sig-sp , at each channel at a wavelength, λ, is given by using the following equations: source 0
0 ave in OA ave dyna sampler
ASE ave sp OSA sig hv B DR
DR GhvB CAL× (P ASE OSA-ave−ISO dyna-ave × P sig-OA-outOSA-ave) (9) where
B 0 is the OSA resolution bandwidth, in Hz, h is Planck's constant, ν is the optical signal frequency, in Hz
NOTE The second terms in Equations (8) and (9) are used to cancel the effect of the signal leakage in ASE measurement
The ASE power distribution around the signal wavelength allows for the estimation of ASE power, excluding signal leakage, through an interpolation technique This estimation can be represented using Equation (10).
ASE power
ASE power at the OA output is given by using Equation (11) or (12)
DR ISO G × P OA-in-ave (11) or
ASE(B o ) = sig - OFA - out OSA - ave ave - ave dyna
Gain calculation
Signal linear gain is given by using the following equations;
- OFA - sig ave - ASE OSA ave
Average output signal power
Average output signal power is given by using Equation (15)
P OA-out-ave = { } sampler source ave
Noise figure calculation
Noise figure NF is obtained from noise factor F by using Equation (16)
Each channel will present key details including the wavelength range of measurement, the spectral linewidth (FWHM) of the optical source, and the input signal wavelength (\$λ_k\$) Additionally, the OSA optical bandwidth (\$B_o\$), optical pump power indication (if applicable), and ambient temperature will be provided Important timing parameters such as pulse interval (\$T\$), signal pulse width (\$T_{source}\$), and sampler width (\$T_{sample}\$) will also be included The average input signal power (\$P_{OA-in-ave}\$) and average output signal power (\$P_{OA-out-ave}\$) will be reported, along with the linear gain (\$G\$), ASE power (\$ASE(B_o)\$), and noise factor (\$F_{SIG-SP}\$) or noise figure (\$NF_{SIG-SP}\$).
Output waveforms for various EDFAs at 25 kHz and 500 kHz pulse rates
Figure A.1 illustrates the output waveforms of different types of EDFAs, highlighting that the gain of the EDFA fluctuates within a single pulse waveform at a pulse rate of 25 kHz Additionally, the gain varies among the different EDFA types A, B, and C.
Type A EDFA operates with a constant pump power in a saturated regime, while Type B EDFA features a slower automatic power control (APC) In contrast, Type C EDFA is equipped with a rapid APC, functioning within an operating band greater than 25 kHz.
When the pulse rate is increased to 500 kHz, the gain change in type C EDFA is eliminated, as illustrated in Figure A.1 (c) and (d) Therefore, gain and noise figure measurements are reliable for frequencies exceeding 500 kHz.
`,,```,,,,````-`-`,,`,,`,`,,` - a) EDFA type A at 25 kHz b) EDFA type B at 25 kHz c) EDFA type C at 25 kHz d) EDFA type C at 500 kHz Figure A.1 – EDFA output waveforms for various EDFAs
Measurement accuracy versus pulse rate
Figure B.1 illustrates the accuracy of NF measurements in relation to pulse rate, utilizing optical pulse source a (refer to Clause 4, Figure 2) The AOM switches facilitated both source pulsation and sampling under the specified measurement conditions.
1 MHz; pulse duty ratios: 0,4 for pulsation and 0,2 for sampling; wavelength-division multiplexed channels: 1 550,4 nm, 1 551,2 nm, 1 552,0 nm and 1 552,8 nm; Total OA input power: 0 dBm;
OA gain: 9 dB to 17 dB
Figure B.1 – NF measurement accuracy versus pulse rate
The NF value remains stable for pulse rates exceeding approximately 250 kHz, eliminating the impact of waveform distortion caused by the slow gain dynamics of EDFAs Additionally, high measurement accuracy is attained at these elevated pulse rates.
The measurements outlined in this annex are feasible due to the slow gain response of rare-earth doped fibre amplifiers, specifically greater than 100 μs for Er-doped fibre amplifiers This slow gain recovery enables pulse repetition rates between 25 kHz and 100 kHz A straightforward setup for assessing the optical amplifier gain response in relation to modulation frequency is illustrated in Figure C.1.
An optical source with a variable modulation frequency is utilized in the optical amplifier (OA), where the average output power is measured using an optical power meter As the modulation frequency increases, the power meter readings asymptotically approach a final value However, at lower modulation frequencies, there is a noticeable increase in error attributed to the non-linear gain recovery of the OA.
Figure C.1 – Set-up to evaluate gain recovery error versus modulation rate
Figure C.2 illustrates measurements from a 980 nm pumped Er-doped fiber amplifier at three different pump current levels As the pump power rises, the gain recovery time constant decreases, leading to a greater deviation from the high-frequency gain value For optimal performance, this amplifier necessitates a modulation frequency exceeding 20 kHz to achieve less than 0.1 dB error in the measured gain.
Gai n rec ov e ry error (dB )
Figure C.2 – Gain recovery error versus modulation frequency with pump current as a parameter
When considering modulation frequency, two key factors must be addressed First, higher pump currents can significantly reduce recovery time, as shown in Figure C.2 Second, testing operational amplifiers (OAs) is essential when automatic gain control (AGC) or automatic level control (ALC) circuitry is active, as the bandwidths of these control loops can limit the modulation rate Therefore, it is advisable to conduct this test to determine the suitable modulation rate for a specific amplifier design.
NOTE In performing the above test, modulation rates below about 10 kHz should not be used A large output power transient could destroy OA or test system components
IEC 60793-1 (all parts), Optical fibres – Part 1: Measurement methods and test procedures
IEC 60825-1, Safety of laser products – Part 1: Equipment classification and requirements
IEC 60825-2, Safety of laser products – Part 2: Safety of optical fibre communication systems
IEC 60874-1, Connectors for optical fibres and cables – Part 1: Generic specification
IEC 61290-1-1, Optical amplifiers – Test methods – Part 1-1: Power and gain parameters –
IEC 61290-3, Optical fibre amplifiers – Basic specification – Part 3: Test methods for noise figure parameters