radiation-resistant optical fiber amplifiers for satellite ......optical fiber amplifiers are key...

Radiation-resistant optical fiber amplifiers for satellite

communications L. Stampoulidis, J. Edmunds, M. Kechagias, G. Stevens, J. Farzana, M. Welch

and E. Kehayas *

Gooch & Housego (Torquay), Broomhill Way, Torquay, Devon, TQ2 7QL, United Kingdom

ABSTRACT

Optical fiber amplifiers are key building blocks in laser communication terminals and telecom photonic payloads. In this

paper we present 1.55μm booster amplifiers and pre-amplifiers suitable for satellite to ground, inter-satellite links and

flexible photonic payloads. We validate the designs in the relevant space environment by characterizing the performance

against ionizing radiation and report on functional performance of the amplifiers over temperature, in thermal vacuum and

after vibration and mechanical shock.

Keywords: Fiber optics, Erbium-doped fiber amplifier (EDFA), radiation-hard optical amplifiers, communications, free

space optics, satellite communications, space photonics

1. INTRODUCTION

Satellite laser communications can enable high-data rate, secure free-space optical links with reduced mass, size, and

electrical power consumption requirements compared to traditional RF technology. Recent1 and planned2,3 missions have

focused on utilizing the eye-safe 1.55 μm wavelength window for short and long range laser-comms. On-board laser-

comm systems require optical fibre amplifiers (OFA) at the transmitter and receiver sides to boost signal power and amplify

optical signals before detection. Within the satellite, in a photonic payload scenario, amplifiers are required to compensate

for insertion losses between functional blocks of the payload. For these applications, the European Space Agency (ESA)

has identified requirements related to mid-power polarization maintaining (PM) fiber amplifiers as well as low noise, high

gain pre-amplifiers. In order to develop these units, ESA has launched the programme "Space validation of rad-hard erbium

optical fiber amplifier at 1.55 um" as part of the European Components Initiative (ECI) framework, dedicated to the

radiation validation and manufacturing of mid-power PM boosters and pre-amplifiers. In this fame, Gooch & Housego is

leading the activity and is performing a radiation evaluation as well as the assembly, integration and test of fiber amplifier

units to meet the requirements of space agencies and prime contractors.

The radiation validation campaign includes gamma radiation tests at 20 krad and 100 krad TID on Erbium doped fibers

and sub-assemblies to cover typical TID levels found in LEO and GEO satellite missions respectively. The G&H PM

booster Erbium-doped fiber amplifier (EDFA) developed (HYDRA series) deliver >19 dBm saturated optical power

demonstrating <2.7 dB radiation-induced absorption (RIA) gain drop at TIDs as high as 100 krad. The pre-amplifier

developed (PEGASUS series) can deliver >35 dB gain over the full C-band and >40 dB around 1550 nm with <1.87 dB

gain drop at 100 krad TID due to RIA. When noise management is employed the typical gain (1550 nm) increases to ~50

dB and the amplifier delivers more than 40 dB gain across the complete C-band. The post irradiation Noise Figure

measured is in the range of 4 dB at 20 krad and 4.5 dB at 100 krad within the entire C-band. In terms of unit development,

we present the assembly, integration and functional test of Engineering Model (EM) PM booster and pre-amplifier EDFAs.

The EM units employ G&H, high-rel fused components which are developed for the demanding submarine market and

have been qualified and deployed in satellite missions4. In addition, they include pump laser diodes manufactured in-house

by G&H. The functional test results demonstrate stable operation over the required module temperature range, validating

the unit thermal design. The heat rise of the pump laser diodes with respect to the case temperature indicates that the

amplifiers can operate at hot operating temperatures exceeding 40 degC. Mechanical tests and cycling in thermal vacuum

showed no degradation, validating the module applicability for space applications.

2. RADIATION TESTING

Radiation tests were performed in collaboration with ALTER Technology at the Centro Nacional de Aceleradores (CNA)

Gamma Facility in Seville, Spain. Gamma irradiation was performed at room temperature using 60Co source at a dose rate

Free-Space Laser Communication and Atmospheric Propagation XXIX, edited by Hamid Hemmati, Don M. Boroson, Proc. of SPIE Vol. 10096, 100960H · © 2017 SPIE · CCC code: 0277-786X/17/$18 · doi: 10.1117/12.2252086

Proc. of SPIE Vol. 10096 100960H-1

Downloaded From: http://proceedings.spiedigitallibrary.org/ on 03/16/2017 Terms of Use: http://spiedigitallibrary.org/ss/termsofuse.aspx

30 28

W.) s25 rn" 24

.4-s 747-11

z' 2020

BOL de-rated current 16 BOL de-rated current

15 EOL de-rated current - EOL de-rated current

- Max de-rated current 12 - Max de-rated current

10

8

54

Pre-rad 20kRAD TID

O o1520 1540 1560

Wavelength (run)

15 80 1520 1530 1540 1550 1560 1570

Wavelength (tun)

30

25

20 * .... .. -111- -1fr.(.1

15

10

5

BOL de-rated current

EOL de-rated current

- . Max de-rated current

100kRAD TIDo

1520 1530 1540 1550 1560 15 70

Wavelength (tun)

of 210 rad/h. The Erbium-doped fibers were passively irradiated, i.e. not pumped and hence no photo-beaching occurred

during irradiation. Hence, the radiation results correspond to the worst case gain drop, expected due to radiation induced

absorption (RIA).

2.1 PM booster amplifiers

The target key optical specifications for the PM booster amplifier are >18 dBm output power and a Noise Figure (NF) in

the range of 6 dB at 1550 nm. A total of 10 fiber amplifier engines were manufactured and measured pre- and post-

irradiation in an amplifier configuration. The figure below illustrates typical irradiation results including gain, NF and

polarization extinction ratio (PER) at different driving currents of the pump laser diodes.

Figure 1. Gain, noise figure and PER vs. wavelength for the PM amplifier with 0 dBm input: pre-irradiation (left), 20 krad

(middle) and 100krad TID (right). Results are shown for three different pump powers (driving currents).

Over the C band, the PM EDFA gain is between 20.5 dB to 21.5 dB at BOL pump power and the Noise figure was

measured between 5.39 to 7.14 dB. Over the narrower 1540 nm to 1565 nm band, performance improves with the gain

measured between 21 dB to 21.5 dB and the noise figure between 5.39 dB and 5.90 dB. The PER is consistently high and

measured to be > 23.3 dB. Figure 1 (middle) illustrates the post-radiation results of the best performing PM EDFA

configuration at 20 krad TID. Over the C band, the PM EDFA gain is between 19.9 and 20.8 dB at BOL pump power.

Noise figure is measured between 5.41 dB to 7.67 dB. Over the narrower 1540 nm to 1565 nm band, performance improves

and ranges from 20.35 dB to 20.8 dB and from 5.41 dB to 6.18 dB for the gain and NF respectively. The PER remains

unaffected from radiation and was measured to be ~ 24 dB for all pump powers. Fig. 1 (right) illustrates the post-radiation

results of the best performing fiber amplifier configuration at 100 krad TID. Over the C band, PM EDFA gain is between

18.46 to 19 dB at BOL pump power. Noise figure is between 10.1 to 6.46 dB. Over the 1540 nm to 1565 nm band,

performance improves to 18.65 < G < 19 dB and 6.46 < NF < 7.96 dB. The PER continues to remain at > 26 dB for all

pump powers. Figure 2 below shows the optical spectrum analyzer (OSA) traces within the 1530 nm to 1565 nm band

with a step of 5 nm for pre-irradiation (left) and following 100 krad TID (right) at BOL pump power. Graphs indicate a

negligible impact on the EDFA OSNR performance.



i

INPUT

(dBm)

BOL pump

0 1530

0 1535

o

o

o

o

o

o

WL

(nm)

1540

1545

1550

1555

1560

1565

Gain 0

(dB)

Gain 20

(dB)

Gain 100

(dB)

DG 0-20 DGO-100 NP UT

dBm)

WL

(nm)

NFO

dB

NF20

dB

NF 100

dB

DN F0-20

dB

DN F0-100

dB

20.5

20.8

19.9

20.16

18.46

18.6

0.6

0.64

2.04 BOL pump

7.141 7.668 10.119 0.527 2.9782.2 0 153021 20.35 18.65 0.65 2.35

0 1540 5.895 6.18 7.958 0.285 2.06321.2 20.5 18.82 0.7 2.38

21.3 20.6 19 0.7 2.3 0 1550 5.364 5.775 7.248 0.411 1.884

21.5 20.75 19.03 0.75 2.47 0 1560 5.482 5.587 7.034 0.105 1.55221.5 20.8 18.99 0.7 2.51

21.5 20.8 18.8 0.7 2.7 0 1565 5.385 5.411 6.459 0.026 1.074

Figure 2. Amplified signal OSA traces pre-irradiation (left) and post-irradiation after 100 krad TID (right)

The tables below summarize the gain and NF degradation due to RIA experienced at 20 and 100 krad TID. With 1550 nm

wavelength as reference, measurements reveal a gain drop of 0.7 dB and NF increase of 0.41 dB at 20 krad TID. The

corresponding values at 100 krad TID are 2.3 dB and 1.88 dB.

Table 1. Gain degradation at 0 dBm input and BOL pump power (left), NF degradation at the same conditions (right)

The table below illustrates the gain recovery as a function of pump driving current at 1550 nm reference wavelength.

Table 2. Gain recovery of irradiated PM EDFA after increase of pump power

The results indicate that the increase of pump power from BOL to EOL value is sufficient to recover fully the gain drop

at 20 krad. At 100 krad, the residual gain drop that is not recovered by a BOL to EOL pump power increase is 1.43 dB,

but still the amplifier is capable to deliver a gain just below 20 dB. The gain values are also presented for the case where

the pump lasers are driven at the maximum de-rated current. The results indicate that a further recovery of the gain

degradation is possible.

-60

-40

-20

0

20

1520 1540 1560 1580

Att

.P

ow

er (

dB

m)

Wavelength (nm)

-60

-40

-20

0

20

1520 1540 1560 1580

Wavelength (nm)

Gain (@ 1550nm) (dB)

Pump Powers G0 G20 G100

BOL 21.3 20.6 19

EOL 22.3 21.5 19.87

@ Max. De-rated

current22.55 21.8 20.12



60 60

7 50 50

40 40

30 30BOL current

EOL current20 20

10 10

. . . . _

0. . -

0

1520 1530 1540 1550 1560 1570 1520 1530 1540 1550 1560 1570Wavelength (tun)

Wavelength (nm)

60

c7

50

40BOL current

EOL current30

20

10

0

1520 1530 1540 1550 1560 1570 1520 1530 1540 1550 1560 1570

Wavelength (inn) Wavelength (Inn)

c7

60 -50

:

40

30

20

10

0

2.2 Pre-amplifiers

The target optical specification for the pre-amplifiers is to achieve >45 dB gain and noise figure (NF) in the range of 4 dB

in a wavelength range of 1540 to 1560 nm. A total of 8 fiber amplifier engines were manufactured with irradiated fibers

and subsequently tested in several multi-stage co- and counter-propagating amplifier configurations.

The figures below illustrate typical pre-irradiation results including gain and NF at BOL pump power and EOL pump

power. Figure 3 (left) shows the gain and NF performance without any noise management and Fig. 3 (right) shows the

gain and NF performance over wavelength with a 0.8 nm band-pass filter.

Figure 3. Pre-irradiation gain and noise figure vs. wavelength for the pre-amplifier at -40 dBm input optical power. Results

are shown for BOL and EOL pump powers with no noise management (left) and with noise management (right)

Over the C band, the gain of the pre-amplifier varies significantly between 37 to 52 dB at BOL pump powers as the

amplifier is operated in the small signal gain regime. The noise figure is relatively flat over the C-band and varies between

4 dB to 4.25 dB. Over the 1540 nm to 1560 nm band, the gain is approximately 43 dB and the NF is less than 4.2 dB. The

incorporation of noise management increases the gain to a maximum of 61 dB, wheras the gain is > 52 dB for the complete

1540 - 1560 nm range. The figures below illustrate the post-radiation results of the best-performing fiber amplifier

configuration at 20 krad TID. Over the C band, the gain of the pre-amplifier is between 36.75 < G < 52.15 dB at BOL

pump power. The noise figure is between 3.95 < NF < 4.3. Over the 1540 nm to 1565 nm band, gain is G > 42.25 dB.

Figure 4. Post-irradiation gain and noise figure vs. wavelength for the pre-amplifier at -40 dBm input at 20 kRAD TID.

Results are shown for BOL and EOL pump powers with no noise management (left) and with noise management (right).



60

050

40

30

20

10

z0

BOL de -rated current

EOL de -rated current

-a- Max de -rated current

.

60 60 -50

:. 050

..40 40

BOL de -rated current

30 BOL de -rated current 30 EOL de-rated currenttEOL de -rated current - - Max de -rated current

20 - Max de -rated current 20

z

10

wz

10

0 0

1520 1540 1560 1580 1520 1540 1560 1580 1520 1540 1560 1580

Wavelength (mn) Wavelength (tun) Wavelength (nm)

INPUT

(dBm)

WL

(nm)

Gain 0

(dB)

Gain 20

(dB)

DG0-20 INPUT WL

(nm)

NFO NF 20 DN FO-20

Pump BOL(dBm)

-40 1530 52.1 52.15 -0.05 Pump BOL

-40 1535 48.35 48.15 0.2 -40 1530 4.254 4.298 0.044-40 1540 42.8 42.25 0.55-40 1545 44.7 44.1 0.6 -40 1540 4.17 4.099 -0.071

-40 1550 45.1 44.4 0.7 -40 1550 4.059 3.998 -0.061-40 1555 44.9 44 0.9 -40 1560 4.02 3.945 -0.075-40 1560 43.1 42.3 0.8

-40 1565 37.65 36.75 0.9 -40 1565 4.151 4.119 -0.032

INPUT WL Gain Gain 100 DGO-100 INPU I WL NF NF 100 DNFO-100(dBm) (nm) (dB) (dB)

(nm)Pump BOL

(dBm)

-40 1530 51.7 51.1 0.6 Pump BOL

-40 1535 48.45 47.42 1.03 -40 1530 4.485 4.677 0.192-40 1540 42.9 41.44 1.46-40 1545 45 43.56 1.44 -40 1540 4.278 4.58 0.302

-40 1550 45.2 43.82 1.38 -40 1550 4.096 4.574 0.478-40 1555 44.9 43.39 1.51

-40 1560 4.064 4.572 0.508-40 1560 43.15 41.54 1.61

-40 1565 37.65 35.78 1.87 -40 1565 4.289 4.665 0.376

The figure below illustrates the post-radiation results of the best performing fiber amplifier configuration at 100 krad TID.

Over the C band, the gain of the pre-amplifier is between 35.78 < G < 51.1 dB at BOL pump power. Noise figure is

between 4.57 < NF < 4.68 dB. Over the 1540 nm to 1560 nm band, gain is G > 41.44 dB and NF < 4.57dB. The inclusion

of noise management increases the gain to between 42.74 < G < 56.95 dB for the entire C-band and the gain is G > 49.88

dB for the 1540 - 1560 nm range.

Figure 5. Post-irradiation gain and NF vs. wavelength for the pre-amplifier at 100 krad TID. Results are shown for BOL and

EOL pump power and pump power at maximum de-rated driving current of the pump diode with no noise management at -

40 dBm input (left), with no noise management at -45 dBm input (mid) and with noise management at -40 dBm input.

The tables below illustrate the gain and NF degradation for the best performing LEO (20 krad) and GEO (100 krad) pre-

amplifier. With 1550 nm wavelength as reference the gain drop of the amplifier is limited to 0.7 dB at 20 krad TID with

no impact on NF. At 100 krad the gain degradation (at 1550 nm) is 1.38 dB and NF degrades by 0.47 dB. The results

indicate that the fiber amplifier can sustain a gain of >41.54 dB in the 1540 to 1560 nm range even at 100 krad incident

on the fiber and without assuming: 1) any increase of pump power from BOL to EOL and 2) any shielding from spacecraft

and the amplifier unit that would reduce the TID level incident on the fiber.

Table 3. Gain and NF degradation recorded at -40 dBm input and BOL pump power for 20 krad TID (top), Gain and NF

degradation at the same conditions for 100 krad TID (bottom).



Table 4 shows how the gain drop recovers as a function of pump driving current at 1550 nm reference wavelength for the

LEO (left) and GEO (right) pre-amplifier.

Table 4. Gain recovery of irradiated PM EDFA after increase of pump power.

The results indicate that the increase of pump power from BOL to EOL value is sufficient to fully recover the gain drop

at 20 krad. At 100 krad, the residual gain drop that is not recovered by a BOL to EOL pump power increase is 0.64 dB,

but the amplifier is still capable to deliver a gain of ~44.5 dB. The gain values are also presented for the case where the

pump lasers are driven at maximum de-rated current for 100 krad. The result indicates that almost a full recovery of the

gain degradation is possible.

3. UNIT ASSEMBLY, INTEGRATION AND TEST

3.1 Engineering model development and testing

Two engineering model (EM) amplifiers were assembled and tested. The opto-electronic components in the EM units are

representative to the components to be used in the EQM or FM unit, which means that they are Form-Fit-Function (FFF)

equivalent and manufactured with identical fabrication processes and materials as the EQM/FM parts. The same applies

for the fused fiber optic components. As such, the optical performance of the EM units is expected to be representative to

the performance of the higher TRL units. The fused components are manufactured in-house by Gooch & Housego

according to high-rel processes. The pump diodes used are also manufactured in-house using Gooch & Housego laser

welding assembly processes, which delivers fully hermetic and high-reliability opto-electronic modules. These

components have been evaluated successfully against mechanical, thermal vacuum and radiation5.

The figure below illustrates the CAD drawing and the finished unit which is identical for both the PM-booster and pre-

amplifier. The mechanical layout is representative of the EQM/FM unit. The unit mass is 560 grams and the dimensions

are (L) 159.5 mm x (W) 146 mm x (H) 27.5 mm (length and width excluding the mounting flanges). The optical interface

is Mini-AVIM and the electrical interface is 44-pin high density D-sub. The unit design accommodates pump laser diode

redundancy and can host 2 fiber amplifiers in the same housing. In terms of monitor and control functions, the unit includes

input and output optical power monitors, pump driving current and temperature monitors as well as pump enable / disable

function and loss-of-input functions.

In terms of electronics interface, the unit has a digital interface with LVDS transceivers and ADC/DAC converters, as well

as an on board DC-DC converter that enables operation from a single voltage supply. Using the digital interface, the unit

offers various advantages including lowest external and internal I/O requirements, known standard at the interface

requiring no additional qualification, lowest cost on the system level and least amount of real estate required within the

pump diode circuit. Finally, the unit includes an internal thermistor mounted on the unit case reference point to provide

monitoring of the unit temperature. In the EM model shown below the electronic circuitry is developed with COTS

components. An EQM/PFM EDFA with space-grade EEE parts has been manufactured and is ready to undergo

qualification.

Gain (@ 1550nm)

(dB)

Pump Powers G0 G20

BOL 45.1 44.4

EOL 45.95 44.99

@ max de-rated

current- -

Gain (@1550nm)

(dB)

Pump Powers G0 G100

BOL 45.2 43.82

EOL 45.95 44.56

@ max de-rated

current- 45.06



34

32

30

28

0 26

24

22- --10degCt 21degC

40degC20

1520 1540 1560 158C

Signal Wavelength (nm)

40

35

30

25

20

á15

10

5

0

.--.--+ -- 10degC-s- 21degC

40degC

8

7

6

5

4

Ñ

z32

1

o

is-lOdegC21degC

-. -40degC

1520 1540 1560 1580 1520 1540 1560 1580

Signal Wavelength (nm) Signal Wavelength (nm)

Figure 6. EM units: optical pre-amplifier (left) and PM booster amplifier (right)

Both EM units have been placed in a thermal chamber and tested at 21 degC and the qualification cold and hot temperatures

of -10 degC and +40 degC. The figure below illustrates the gain, PER and NF performance of the PM booster EM-EDFA

unit over temperature. The PM booster EDFA has been optimized to provide peak gain around 1560 nm region. The results

show that at room temperature and -10 dBm input power the amplifier would deliver typically (1550 nm) >20 dBm output

power with >25 dB PER and NF ~5 dB. The results indicate that performance is sustained at 40 degC. At -10 degC, there

is a drop of gain and NF increase which is attributed to the decrease of PER down to ~12 dB. The decrease in PER is

attributed to the performance of the passive fiber optic components which are specified for PER performance >20 dB at

>-5 degC. As such a >20 dB PER performance cannot be guaranteed at cold temperatures >-5 degC.

The temperature of the pumps was monitored during test through the internal thermistor elements. At the unit case hot

temperature of 40 degC, the pump temperature heat rise was ~5.5 degC when the pumps were driven at BOL pump driving

current and ~7.5 degC when pumps were driven at maximum de-rated driving current. As such the fiber amplifier can

operate at hot temperatures beyond 40 degC, given that the pump laser diodes have a maximum operating case temperature

of 75 degC.

Figure 7. PM-booster EM-EDFA performance over qualification temperature range: (left) gain, (middle) PER, (right) NF.

The figure below illustrates the performance of the pre-amplifier EM-EDFA unit over temperature. Results are shown with

noise management. The results show that at room temperature and -40 dBm input power the amplifier would deliver

typically (1550 nm) >55 dB gain with NF ~<4 dB. The results indicate that performance is sustained at both cold and hot

temperatures of -10 and 40 degC respectively.



65

60

55

.-.50

45

40

35

5

4

z3.5

30 3

1520 1530 1540 1550 1560 1570 1520 1530 1540 1550 1560 1570

Signal Wavelength (nm) Signal Wavelength (nm)

- --lOdegC --21degC 40degC - --lOdegC --21degC 40degC

Mechanical Shock/20

ShockY -axis

ShockX-axis

ShockZ-axis

DYNAMICSEQM / OFA / 300

Sine & Random Vibration/10

VibY-axis

VibX-axis

VibZ-axis

TVACInitial

Ambiento

THERMALVACUUM

EQM / OFA/ 400/10

TVAC@ Tq-max

TVAC@ Tq-m in o

TVACFinal

Ambient

Figure 8. Pre-amplifier EM-EDFA performance over qualification temperature range: (left) gain, (middle) PER, (right) NF.

3.2 Qualification Model gain block development

In order to assess the integration methodologies and unit-level assembly processes followed, a gain block was

manufactured in order to go through ECSS-based thermo-mechanical testing. The gain block contained one fully functional

PM optical fiber amplifier including hi-rel fused and opto-electronic components. The gain block excluded driving and

interface electronics for programmatic reasons associated with the high cost and long lead times of space-grade EEE parts.

The pump lasers were de-golded and pre-tined, whereas electro-optic measurements and seal tests were performed before

and after to verify hermetic sealing. Flight-level connectors were used and the Printed Circuit Board Assembly (PCBA)

was performed in an ESA-approved assembly house using space-grade processes. Mass dummies were used to replace the

electronic ICs and other components in order to maintain similarity to an optical fiber amplifier module. The test flow is

shown in the figure below and follows the rationale provided by ECSS-E-10-03.

Figure 9. Initial unit-level test flow based on ECSS-E-10-03

The test sequence is as follows:

FFT and PT: Following the Gain Block integration a Full Function Test (FFT) and a Performance Test (PT) are

performed to verify compliance to specifications.

Mechanical test: The mechanical testing comprises random vibration, sine vibration and mechanical shock.

Resonance test and Reduced Function Test (RFT) are performed before and after each axis of vibration / shock

testing is completed. Additionally, at the end of the Mechanical Testing sequence an RFT is performed.

FFT

FFT

Mass

PM-BOFA-GB / 200

PM-BOFA-GB / 300

PM-BOFA-GB / 400 / 10

Mechanical Shock /30

DYNAMICS (2)PM-BOFA-GB / 300



bration ST Description RESULTS

Low Level Sine Results LPT Results

MainRasone

nce

_knits (Hz) (+/-S%Main

Resonance

Imits (gn) (+1-20% Gain

(dBm)

Min

Umh(dem)

NFLD1 -

ThermistorL02-

ThermistorLD3 -

ThermistorLD1 -PD LD2 -PD L03 -PD External TC

Max Min Max Min

Z-12 LPT and Low Level Sine PASS 923 969.2 876.9 8 9.6 6.4 21.08 19.53 4.19 34.61 34.25 35.51 1.86E -03 1.76E -03 1.67E -03 25.94Z-13 Random -5d8 (15.28gRMS) - 20s PASS

Z-14 LPT and Low Level Sine PASS 923 969.2 I 876.9 7.7 9.6 6.4 20.94 19.53 4.41 32.52 32.14 33.38 1.86E-03 1.76E-03 3.67E-03 24.12Z-15 Random -4d8 (17.14gRMS) - 20s PASS

Z-16 LPT and Low Level Sine PASS 923 I 969.2 I 876.9 1 7.65 I 9.6I

6.4 I 20.82 I 19.53 I 4.41 I 33.06 I 32.69 I 33.93 I 1.87E-03 I 1.76E-03 I 1.67E-03 I 24.60

Z -17 Random -3dB (19.23gRMS) - 2min PASS

Z -18 LPT and Low Level Sine PASS 921 I 969.2 I 876.9 1 7.82 I 9.6 I 6.4 I 20.81 I 19.53 I 4.42 1 33.25 32.90 I 34.15 I 1.87E-03 1 1.76E-03 I 1.67E-03 1 24.73Change to X axis

X -1 LPT and Low Level Sine PASS 1805 I 1895.3 I 1714.8 I 1.1 I 1.32I

0.88I 20.92

I19.53

I4.31 I 33.59 33.21 I 34.47

I1.87E-03 I 1.76E-03

I1.67E-03 I 25.06

X -2 Random -5d6 (15.28gRMS) - 20s PASS

X-3 LPT and Low Level Sine PASS 1810 I 1895.3 I 1714.8 I 1.1 9.6 6.4 20.97 19.53 4.28 I 35.64 1 35.25 I 36.55 1.86E-03 I 1.76E-03 1.67E-03 I 26.87X-4 Random -4d8 (17.14gRMS) - 20s PASS

xs LPT and Low Level Sine PASS 1815 I 1895.3 I 1714.8 I 1.1 I 9.6 I 6.4 I 20.95 19.53 4.22 I 37.64 T- 37.25 I 38.59 1.86E-03 I 1.75E-03 1.67E-03 I 28.63

X-6 Random -3d8 (19.23gRMS) - 2min PASS

X-7 LPT and Low Level Sine PASS 1816 I 1895.3 1 1714.8 1 1.1 9.6 6.4 20.93 19.53 4.29 I 38.32 1 37.94 I 39.30 1.85E-03 1 1.75E-03 1.67E-03 1 29.30Change to Y axis

Y-1 LPT and Low Level Sine PASS NA 1 NA I NA I NA I NA I NA I 20.911 19.53 I 4.25 1 39.93 39.50 1 40.72 I 1.85E-03 1 1.75E-03 11.66E-03 1 30.28

Y-2 Random -SdB (1S.28gRMS) - 20s PASS

Y-3 LPT and Low Level Sine PASS NA I NA I NA I NA I NA I NA 20.85 119.53 I 4.33 I 41.12 I 40.66 I 41.90 I 1.85E-03 I 1.74E-03 1.66E-03 I 31.18Y-4 Random -4d8 (17.14gRMS) - 20s PASS

Y.S LPT and Low Level Sine PASS NA I NA I NA I NA NA NA 20.85 19.53 4.31 I 41.791

41.33 I 42.55 1.85E-03 I 1.74E-03 1.66E-03 I 31.65

Y-6 Random -3d6 (19.23gRMS) - 2min PASS

Y-7 LPT and Low Level Sine PASS NA I NA 1 NA1

NA NA NA 20.83 19.53 4.35 I 42.25 I 41.81 I 43.04 125E-03 I 1.74E-03 1.66E-03 I 32.16

Thermal vacuum test: Temperature cycling in vacuum is performed with an RFT performed at the lowest and

highest temperature plateaus.

Final FFT and PT: Establishes no performance degradation following environmental testing and handling.

Figure 10. Photos of thermo-mechanical test setups at ALTER Technology: thermal vacuum chamber (top left), unit

mounted in the chamber (top center), out-of-plane vibration (bottom left), in-plane vibration (bottom center), shock (right)

3.3 Mechanical testing

As per the test flow, the unit underwent low-/high-level sine, random vibration in all three axes before being subjected to

shock levels initially up to 1,000 g and following cycling in thermal vacuum tests up to 2,500 g were performed. The unit

passed 25 g sine and 19.2 g RMS following a failure and corresponding repair at 17 g. The failure was at component-level

and was due to a unit-level assembly error.

Table 5. Low-/High-level sine and random vibration up to 19.2g RMS showing no optical amplifier gain drop

Following vibration testing, the unit was subjected to 1,000 g shock in three axes, with 3 shocks per axis. As the structural

strength of the unit was unknown, and in order not to jeopardize the evaluation test campaign, the shock testing was split

into two parts, one before and one after thermal vacuum testing. The initial shock testing was limited to 1,000 g.



Shock STEP Description RESULTS

LPT Results

Gain(dBm)

Min Limit(dBm)

NPLD1-

ThermistoLD2 -

ThermistoLD3-

ThermistoLD1- PD LD2 - PD LD3 - PD

ExternalTC

SRS -Z -1 LPT PASS 20.38 19.38 5.03 31.27 30.88 32.03 1.87E -03 1.76E -03 1.68E -03 22.77

SRS -Z -2 SRS Z AXIS 3 shocks (1000gn) PASS

SRS -Z -3 LPT PASS 20.39 19.38 5 31.46 31.07 32.23 1.87E -03 1.76E -03 1.68E -03 22.89

SRS -X -1 LPT PASS 20.47 I 19.38 I 4.99 I 31.89 I 31.50 I 32.71 I 1.87E-03 I 1.76E-03 I 1.67E-03 I 23.41SRS -X -2 SRS X AXIS 3 shocks (1000gn) PASS

SRS -X -3 LPT PASS 20.46 I 19.38 I 5 I 32.00 I 31.61 I 32.83 I 1.87E-03 I 1.76E-03 I 1.67E-03 I 23.51

SRS -Y -1 LPT PASS 20.41 I 19.38 I 5.01 I 32.20 I 31.77 I 32.97 I 1.87E-03 I 1.76E-03 I 1.67E-03 I 23.50SRS -Y -2 SRS Y AXIS 3 shocks (1000gn) PASS

SRS -Y -3 LPT PASS 20.42 I 19.38 I 5.01. I 32.33 31.89 I 33.09 I 1.87E-03 I 1.76E-03 + 1.67E-03 I 23.63

Pressure

P mm)A

T,.weaR

tie paws)

Ix 10^ CeMBINEOCCLINGANO VACUUM TEST yi

e Initial and final RFT

Internmediate RFT

tE Dwell time

Switch ON

o Switch OFF

Table 6. Shock results at 1,000g

3.4 Thermal vacuum testing

The gain block was subjected to a total of four thermal cycles over the temperature range with 2-hour dwell time, as

described below. The table illustrates the operating and non-operating as well as the switch ON hot and cold temperatures

that were used during testing.

Parameter Annotation Value

Max non-op hot temp (degC) TNO-max +60

Min non-op cold temp (degC) TNO-min -40

Max op hot temp (degC) TQ-max +40

Max switch ON temp (degC) TSU-high +40

Min op cold temp (degC) TQ-min -10

Min switch ON temp (degC) TSU-low -5

Figure 11. Thermal vacuum test profile (left) and temperature points (right)

The figure below shows the behavior of the amplifier gain under thermal cycling in vacuum. Three thermocouples were

used to monitor the temperature at the baseplate holding the unit (T1, T2, T3) and one thermocouple positioned on the unit

was used as the control (TCtrl). The results show a maximum gain variation of 0.6 dB at the lower temperatures, primarily

due to the performance of polarization-maintaining fused components at temperatures below 0 degC.



Shock STEP1 Description RESULTS

LPT Results

GainfdBml

Min LimitfdBm1

LD1-Thermisto

LD2 -Thermisto

LD3

ThermistoLD1 - PD LD2 - PD W3 - PD External

TC

SRS -Z-4 LPT PASS 20.03 J 19.03 5.38 [ 30.19 29.71 30.90 J 1.86E -03 J 1.76E -03 1 1.68E -03 I 21.35

SRS -Z -5 SRS Z AXIS 3 shocks (1500gn) PASS

SRS -Z -6 LPT PASS 20.04 I 19.03 I 5.39 I 30.27 I 29.79 I 30.98 1 1.86E -03 I 1.76E-03 1 1.68E -03 I 21.50

SRS -X -4 LPT PASS 20.05 I 19.03 I 5.29 30.74 30.28 I 31.46 1 1.85E -03 I 1.76E-03 1 1.69E -03 I 22.27SRS -X -5 SRS X AXIS 3 shocks (1500gn) PASS

SRS -X -6 LPT PASS 20.06 I 19.03 I 5.33 30.54 30.08 31.27 1 1.86E -03 I 1.76E-03 1 1.68E -03 1 22.50

SRS -Y -4 LPT PASS 20.03 J 19.03 I 5.38 J 30.87 30.41 I 31.59 1 1.85E -03 1 1.76E -03 1 1.68E -03 1 22.78SRS -Y -5 SRS Y AXIS 3 shocks (1500gn) PASS

SRS -Y -6 LPT PASS 20.02 I 19.03 I 5.44 30.53 f 30.07 31.27 J 1.86E -03 I 1.76E-03 1 1.69E -03 1 23.05

SRS -Z -7 LPT PASS 20.87 I 19.87 4.56 30.27 I 29.78 I 30.96 1 1.86E -03 I 1.76E-03 I 1.69E-03 1 21.56SRS -Z -8 SRS Z AXIS 3 shocks (2000gn) PASS

SRS -Z -9 IPT PASS 20.63 J 19.87 I 4.68 J 31.66 I 31.2.5 I 32.36 1 1.85E -03 I 1.76E-03 1 1.69E -03 1 22.48

SRS-X-7 IPT PASS 20.73 I 19.87 I 4.62 I 30.83 I 30.37 I 31.57 1 1.85E -03 I 1.76E-03 1 1.69E -03 1 21.99SRS -X -8 SRS X AXIS 3 shocks (2000gn) PASS

SRS -X -9 LPT PASS 20.87 I 19.87 4.48 31.39 30.94 32.14 1 1.85E -03 I 1.76E-03 1 1.69E -03 I 22.55

SRS-Y-7 LPT PASS 20.66 J 19.87 4.68 30.98 30.51 31.67 1 1.85E -03 I 1.76E -03 1 1.69E -03 I 22.03

SRS -Y -8 SRS Y AXIS 3 shocks (2000gn) PASS

SRS -Y -9 LPT PASS 20.68 J 19.87 4.61 J 31.01 J 30.54 31.70 J 1.85E -03 1 1.76E -03 1 1.69E -03 1 22.06

SRS -Z -10 LPT PASS 20.77 I 19.77 j 5.18 j 39.01 I 38.57 j 39.75 1 1.83E -03 I 1.74E-03 1 1.68E -03 1 28.14SRS -Z -11 SRS Z AXIS 3 shocks (2500gn) PASS

SRS -Z -12 LPT PASS 20.91 I 19.77 5.11 j 31.00 I 30.54 31.70 1 1.85E -03 I 1.76E-03 1 1.69E -03 1 22.25

5RS -X -10 LPT PASS 20.8 I 19.77 j 5.18 j 31.05 I 30.59 j 31.74 1 1.85E -03 I 1.76E-03 1 1.69E -03 I 22.31SRS -X -11 SRS X AXIS 3 shocks (2500gn) PASS

SRS -X -12 LPT PASS 20.86 I 19.77 5.18 I 31.40 J 30.96 j 32.12 I 1.85E-03 I 1.76E -03 1 1.69E -03 I 22.71

Figure 12. Amplifier gain performance in thermal vacuum

3.5 Extended shock testing

Following performance evaluation in vacuum, the unit was subjected to higher levels of shock in order to investigate any

failure modes. The shock levels were increased sequentially from 1,500 g to 2,500 g. Three shocks per axis was performed

at every 500 g increment. After each shock, the values of the thermistor, back facet photodiodes, module temperature as

well as the optical gain of the amplifier were recorded. No degradation was observed at 2,500 g.

Table 7. Shock results from 1,500 g up to 2,500 g

18.0

18.5

19.0

19.5

20.0

20.5

21.0

-60

-40

-20

0

20

40

60

80

Gain

(d

Bm

)

Tem

per

atu

re (

ºC)

Time

TCtrl T1T2 T3Gain



4. CONCLUSIONS

We have reported on the design and development of radiation-resistant PM mid-power booster and pre-amplifiers for

satellite communications. The radiation results demonstrate the PM booster is robust against ionizing radiation with <0.75

dB and <2.7 dB gain drop at 20 krad and 100 krad respectively. The degradation is recoverable through increase of pump

power without exceeding the maximum de-rated operational limits of the lasers. In terms of optical performance, the

amplifier is expected to deliver 21.5 dBm saturated output power at 1550 nm for 20 krad TID at EOL pump power. The

corresponding EOL performance at 100 krad TID is >19.5 dBm. Both performance values are in-line with target

requirements. Similarly, the pre-amplifier exhibits a gain drop of 0.7 dB at 1550 nm and 1.38 dB respectively due to RIA.

In terms of optical performance, ~45 dB of gain is delivered at 20 krad and 100 krad TID at EOL pump powers. We have

also reported on the assembly of Engineering Model units using high-rel opto-parts and COTS electronics. The results

validate the optical, electronics and thermal designs for both types of amplifiers. Finally, an amplifier gain block was

subjected to thermo-mechanical tests and passed 25g RMS random vibration, 2,500g shock and successfully operated in

thermal vacuum over the required temperature ranges.

5. ACKNOWLEDGMENT

This work has been supported by ESA through the ECI program "Space validation of Rad-Hard Erbium Optical Fiber

Amplifier at 1.55 um" (ESA contract number 4000110438). We gratefully acknowledge ESA/ESTEC project officer Ms

Charlotte Bringer for the support and technical guidance during the contract execution.

REFERENCES

[1] B. L. Edwards, et al, “The Laser Communications Relay Demonstration,” Proc. International Conference on Space

Optical Systems and Applications (ICSOS) 2012, 1-1, Ajaccio, Corsica, France, October 9-12 (2012)

[2] Toshihiro Kubo-oka et al, “Optical Communication Experiment Using Very Small Optical Transponder Component

on a Small Satellite RISESAT,” Proc. International Conference on Space Optical Systems and Applications (ICSOS)

2012, 11-4, Ajaccio, Corsica, France, October 9-12 (2012)

[3] M. Bacher et al, “A modular solution for routine optical satellite-to-ground communications on small spacecrafts, ”

Proc. International Conference on Space Optical Systems and Applications (ICSOS) 2012, 11-2, Ajaccio, Corsica,

France, October 9-12 (2012)

[4] Soil Moisture Ocean Salinity (SMOS) Earth Explorer mission: https://earth.esa.int/web/guest/missions/esa-

operational-eo-missions/smos

[5] J. MacDougall et al, “Transmission and pump laser modules for space applications,” paper 10096-16, Free-Space

Laser Communication and Atmospheric Propagation XXIX, SPIE Photonics West, Jan 30 2017



radiation-resistant optical fiber amplifiers for satellite ......optical fiber amplifiers are key...

Documents