C-Band Optical Amplifiers (BOAs and SOAs), 1550 nm

Overview
Specs
Graphs
Pin Diagram
Optical Amplifiers
Feedback

Optical Amplifier Selection Guide
785 nm BOA
930 nm BOAs
980 nm BOAs
1050 nm BOAs
O-Band (1285 - 1350 nm) BOAs
E-Band (1410 nm) BOAs
C-Band (1550 nm) BOAs and SOAs
L-Band (1590 - 1625 nm) BOAs
1685 nm BOAs

The center wavelength of a BOA can be readily tailored for specific applications. It is quite common to adjust the BOA wavelength spectrum to match the specific laser source. Please contact us if you have custom wavelength requirements for pilot-projects or OEM applications.

Click to Enlarge
When current is applied across the ridge waveguide, excited state electrons are stimulated by input light, leading to photon replication and signal gain.

Features

Semiconductor Optical Amplifier (SOA)
- Polarization Independent: Amplifies All Polarization States
- SM or PM Fiber Pigtails (1.5 m) with FC/APC Connectors
- Typical Applications: Inline Amplifier, Detector Pre-Amp, Fast Optical Switch
  (~1 ns Switching Speed)
Booster Optical Amplifier (BOA)
- Polarization Maintaining: Amplifies Only One Polarization State
- SM or PM Fiber Pigtails (1.5 m) with FC/APC Connectors
- Typical Applications: Boosting Laser Transmitters, Compensating for Transmit MUX/DeMUX Insertion Loss, Optical Shutter

BOAs and SOAs are single-pass, traveling-wave amplifiers that perform well with both monochromatic and multi-wavelength signals. Since BOAs only amplify one state of polarization, they are best suited for applications where the input polarization of the light is known. For applications where the input polarization is unknown or fluctuates, a Semiconductor Optical Amplifier (SOA) is required. However, the gain, noise, bandwidth, and saturation power specifications of a BOA are superior to that of a SOA because of the design features that make the SOA polarization insensitive.

Thorlabs offers both free space and fiber-coupled versions of its BOAs and SOAs. Free space versions are offered either as a chip on submount (C) or chip on heatsink (H). Fiber-coupled optical amplifiers are offered and exhibit low coupling losses. Losses typically range from 1.5 to 2.5 dB for the fiber-to-chip and chip-to-fiber coupling (each). This affects the total gain, noise figure (NF), and saturation power (P_sat). While the gain produced by the amplifier exceeds that of the losses, these losses remain an important factor in determining the device's performance. For instance a 1 dB drop in input coupling efficiency increases the noise figure by 1 dB. Alternatively, a 1 dB drop in output coupling decreases the saturation power by 1 dB.

Driver Option
The CLD1015 butterfly LD/TEC controller is one possible controller for these amplifiers. The LD/TEC controller and mount combination offers full control via a touch screen. When operating the optical amplifiers on this page with the CLD1015, the orientation for type 1 pin configurations should be used.

Center Wavelength Note
The center wavelength (CWL) of the ASE spectrum in broadband semiconductor devices such as optical amplifiers may show variation between lots. Please refer to the Specs tab for the CWL tolerances of each particular model. For applications in which a specific ASE center wavelength is a critical concern, please contact Tech Support for information on the CWL of currently available lots.

Booster Optical Amplifiers

The BOA consists of a highly efficient InP/InGaAsP Multiple Quantum Well (MQW) layer structure. As seen in the schematic to the right, the input and output of the amplifier is coupled to the reliable ridge waveguide on the optical amplifier chip. C-Band BOAs are available in a standard 14-pin butterfly package with either SMF or PMF pigtails that are terminated with FC/APC connectors. The connector key is aligned to the slow axis on all PMF pigtailed models. Optional polarization-maintaining isolators at the input, output, or both input/output are also available (specifications may vary with different configurations). Please contact Tech Support to order such a device. Alternatively, an unpackaged C-Band BOA chip is available on a submount or heatsink.

Click to Enlarge
Our SOA1117S and BOA1004P optical amplifiers are also available in the S7FC1013S and S9FC1004P benchtop optical amplifiers, respectively.

Semiconductor Optical Amplifiers

The Semiconductor Optical Amplifier (SOA) is a polarization insensitive optical amplifier; therefore, all polarization states are amplified. These devices are an ideal in-line amplifier. Advanced epitaxial wafer growth and opto-electronic packaging techniques enable a high output saturation power, low noise figure, and large gain across a broad spectral bandwidth. These devices come in an industry-standard 14-pin butterfly package with either SMF or PMF pigtails that are terminated with FC/APC connectors. The connector key is aligned to the slow axis on all PMF pigtailed models. These come without isolators, but we are able to provide units with polarization-insensitive isolators at the input, output, or both. Please contact Tech Support for help in ordering such a device.

Item #^a	Center Wavelength	3 dB Bandwidth	Saturated Output Power (@ -3 dB)	Small Signal Gain (@ Pin = -20 dBm)	Noise Figure
BOA1004S and BOA1004P	1550 nm Typical	85 nm Typical	15 dBm Typical	27 dB Typical	7.5 dB Typical
SOA1013S	1500 nm Typical^b	-	14 dBm Typical	13 dB Typical	8.0 dB Typical
SOA1117S and SOA1117P	1550 nm Typical	-	9 dBm Typical	20 dB Typical	10.0 dB Typical
BOA1007C and BOA1007H	1550 nm Typical	85 nm Typical	18 dBm Typical	30 dB Typical	6.0 dB Typical

For complete specifications, please view the Specs and Graphs tabs.
This center wavelength allows an operating wavelength of 1528 - 1562 nm; for details please see the Specs tab.

Item #	Symbol	BOA1004S and BOA1004P
Item #	Symbol	Min	Typical	Max
Operating Current	I_OP	-	600 mA	750 mA
Center Wavelength	λ_C	1530 nm	1550 nm	1570 nm
ASE Optical 3 dB Bandwidth	BW	80 nm	85 nm	-
Saturation Output Power^a (@ -3 dB)	P_SAT	13 dBm	15 dBm	-
Small Signal Gain (@ P_IN = -20 dBm, λ = 1550 nm)	G	23 dB	27 dB	-
Gain Ripple (RMS) @ I_OP	δG	-	0.05 dB	0.3 dB
Polarization Extinction Ratio	PER	-	18 dB	-
Noise Figure	NF	-	7.5 dB	9 dB
Forward Voltage	V_F	-	1.3 V	1.6 V
Chip Length	-	-	1.5 mm	-
Waveguide Refractive Index	-	-	3.2	-
TEC Operation (Typical/Max @ T_CASE = 25/70 °C)
TEC Current	I_TEC	-	0.13 A	1.5 A
TEC Voltage	V_TEC	-	0.28 V	4.0 V
Thermistor Resistance	R_TH	-	10 kΩ	-

The maximum amount of CW power that can be extracted is approximately 3 dB higher than the saturation power. Please see the Optical Amplifiers tab for more information.

Item #	Symbol	SOA1013S			SOA1117S and SOA1117P
Item #	Symbol	Min	Typical	Max	Min	Typical	Max
Operating Current	I_OP	-	500 mA	750 mA	-	500 mA	600 mA
Operating Wavelength Range		1528 nm	-	1562 nm	1528 nm	-	1562 nm
Center Wavelength	λ_C	-	1500 nm^a	-	-	1550 nm	-
Saturation Output Power^b (@ -3 dB)	P_SAT	12 dBm	14 dBm	-	6 dBm^c	9 dBm^c	-
Small Signal Gain (Over C-Band @ Pin = -20 dBm)	G	10 dB	13 dB	-	15 dB	20 dB	-
Gain Flatness (Over C-Band @ Pin = -20 dBm)	ΔG	-	5 dB	7 dB	-	-	-
Gain Ripple (p-p) @ I_OP, λ_C	δG	-	0.1 dB	0.5 dB	-	0.2 dB	0.5 dB
Polarization Dependent Gain	PDG	-	1.0 dB	1.8 dB	-	1 dB	2.5 dB
Noise Figure	NF	-	8 dB	9.5 dB	-	10 dB	11.5 dB
Forward Voltage	V_F	-	1.6 V	1.8 V	-	1.4 V	2.0 V
Chip Length	-	-	1.5 mm	-	-	1.0 mm	-
Waveguide Refractive Index	-	-	3.2	-	-	3.2	-
TEC Operation (Typical/Max @ T_CASE = 25/70 °C)
TEC Current	I_TEC	-	0.23 A	1.5 A	-	0.2 A	1.5 A
TEC Voltage	V_TEC	-	0.5 V	4.0 V	-	0.4 V	4.0 V
Thermistor Resistance	R_TH	-	10 kΩ	-	-	10 kΩ	-

The center wavelength of 1500 nm was selected because this center wavelength yields the specified saturation power for the specified operating wavelength range of 1528 - 1562 nm.
The maximum amount of CW power that can be extracted is approximately 3 dB higher than the saturation power. Please see the Optical Amplifiers tab for more information.
Saturation output power specified across operating wavelength range.

Item #	Symbol	BOA1007C and BOA1007H
Item #	Symbol	Min	Typical	Max
Operating Current	I_OP	-	600 mA	750 mA
Central Wavelength	λ_C	1530 nm	1550 nm	1580 nm
Optical 3 dB Bandwidth	BW	80 nm	85 nm	-
Saturation Output Power^a (@ -3 dB)	P_SAT	15 dBm	18 dBm	-
Small Signal Gain (@ P_IN = -20 dBm, λ = 1550 nm)	G	26 dB	30 dB	-
Gain Ripple (RMS) @ I_OP	δG	-	0.05 dB	0.3 dB
Polarization Extinction Ratio	PER	-	18 dB	-
Chip Noise Figure	NF	-	6.0 dB	8.0 dB
Forward Voltage	V_F	-	1.3 V	1.6 V
Chip Length	L	-	1.5 mm	-
Waveguide Refractive Index		-	3.2	-
Lateral Beam Exit Angle	Θ_EXT	-	19.5°	-
Beam Divergence Angle (FWHM)
Transverse	Θ_T	26°	34°	42°
Lateral	Θ_L	10°	14°	30°

The maximum amount of CW power that can be extracted is approximately 3 dB higher than the saturation power. Please see the Optical Amplifiers tab for more information.

Click to Enlarge
Figure 1: Diagram of the BOA1007C (top down view). The anode and cathode sections of the chip are highlighted here. Please note that the BOA1007H has the same chip structure, and thus, its anode and cathode sections can be similarly determined.

Note: All plots illustrate typical performance, and individual units may have slightly different performance, within the parameters outlined on the Specs tab.

1050 nm BOA Pin Out

Mechanical Drawing and Pin Assignments for Butterfly Package

Comparison of a SOA to a standard Fabry-Perot Laser Diode

Booster optical amplifiers (BOAs) and semiconductor optical amplifiers (SOAs) are single-pass, traveling-wave amplifiers that perform well with both monochromatic and multi-wavelength signals. Since BOAs only amplify one state of polarization, they are best suited for applications where the input polarization of the light is known. For applications where the input polarization is unknown or fluctuates, a Semiconductor Optical Amplifier (SOA) is required. However, the gain, noise, bandwidth, and saturation power specifications of a BOA are superior to that of a SOA because of the design features that make the SOA polarization insensitive.

BOAs and SOAs are similar in design to Fabry-Perot Laser Diodes, the difference being that Fabry-Perot laser diodes have reflective coatings on both end faces of the semiconductor chip. The optical feedback from the reflective end faces establishes a cavity in which lasing can occur. SOAs and BOAs have an anti-reflection (AR) coating on both end faces of the semiconductor chip. The AR coatings limit the optical feedback into the chip so that lasing does not occur.

As is typical for all amplifiers, BOAs/SOAs operate in two regimes: a linear, flat, constant gain regime and a non-linear, saturated output regime. When used to amplify a modulated signal, the linear regime is typically used to eliminate pattern-dependent distortion, multi-channel cross-talk, and transient response issues common to EDFAs. The non-linear regime is used to take advantage of the highly non-linear attributes of the semiconductor gain medium (cross-gain modulation, cross phase modulation) to perform wavelength conversion, optical 3R regeneration, header recognition, and other high-speed optical signal processing functions.

For a continuous wave input signal, the amount of power that can be produced by the amplifier is determined by the saturation output power (P_sat) parameter. P_sat is defined as the output power at which the small-signal gain has been compressed by 3 dB. The maximum amount of CW power that can be extracted is approximately 3 dB higher than the saturation power.

SOA Linear vs Non-linear Regimes

Posted Comments:

alexey Kokhanovskiy (posted 2020-12-01 00:26:02.43)

Dear Sir/Mme We have purchased BOA1004P. What state of the polarization does BOA amplifies? That goes along fast or slow axis of PM fiber? With respect, Alexey Kokhanovskiy

YLohia (posted 2020-12-01 10:20:21.0)

Hello Alexey, thank you for contacting Thorlabs. The BOA amplifies the polarization state along the fast axis of the fiber.

Vasili Savitski (posted 2020-04-07 05:11:46.033)

Hi. Could you please give some numbers on the rise time of your BOA and SOA working in pulsed regime? Thank you.

YLohia (posted 2020-04-07 09:24:54.0)

Hello, thank you for contacting Thorlabs. We have seen rise times on the order of hundreds of picoseconds for these BOAs in the past. Unfortunately, we do not have a formal rise time spec for these. I have reached out to you directly to discuss this further.

DH Lee (posted 2019-08-08 19:24:22.277)

Can you show me the width and height of the ridge waveguide?

YLohia (posted 2019-08-09 09:56:59.0)

Hello, thank you for contacting Thorlabs. Unfortunately, we cannot share this information as it is proprietary. We do, however, post the chip length, refractive index, and transverse/lateral beam divergence angles for the BOA1007C (see Specs tab for more information).

yedukondalu patha (posted 2019-05-16 08:09:06.867)

Could you please send the manual, how to operate SOA1117S. Thanks, Yedukondalu P

YLohia (posted 2019-05-16 09:02:47.0)

Hello Yedukondalu, thank you for contacting Thorlabs. Unfortunately, we do not offer a manual for this as the operation is very similar to that of a regular butterfly pigtail laser diode. I have reached out to you directly to discuss your application further and offer any assistance I can.

kimyeonhwa1006 (posted 2018-10-15 21:07:20.387)

Hi, I have a question for SOA graphs. I'm just wondering that the reason why the SOA 1013S and SOA1117S has different tendency according to Gain vs wavelength. The graph of SOA 1013S in appendix tends to be decreased while that of SOA 1117S tends to be increased. I'm waiting for your reply. Thank you.

YLohia (posted 2018-10-19 12:08:41.0)

Hello, thank you for contacting Thorlabs. The reason is that the peak of the ASE for SOA1013S happens at ~1500-1510 and that of the SOA1117S is at ~1550nm. Thus, based on the 1528-1562nm (C-band) region shown in the plots in question, the plot decreases for SOA1013S and increases for SOA1117S. Both of SOA1013S and SOA1117S are for C-band applications and we only test gain across this window. For the SOA1013S, which is marketed as a linear device (due to its high saturation power), we intentionally set peak gain wavelength at ~1500nm to obtain higher saturation power in the C-band. The saturation power tends to be higher on the longer wavelength side of the peak.

kyeongpyo.lee (posted 2018-09-19 05:10:10.58)

I was testing the small signal gain by using BOA1007H, and I found that whatever the input power is (i.e. -50dBm or -20dBm) its amplified output is fixed at about -13dBm for 200mA current supply. I cannot understand this situation. Is it possible to do so?

YLohia (posted 2018-09-25 10:05:02.0)

Hello, thank you for contacting Thorlabs. Please note that the performance of the BOA is highly dependent on its junction temperature and, thus, the unit should be used under properly temperature controlled environments. The BOA1007H will already start to produce gain (around 10-20dB) even with a 200mA drive current at 1550nm. It is quite likely that your input coupling efficiency to the BOA is not high enough, which would imply that the chip is not receiving enough light to amplify. The measured power then will be the BOA's ASE power itself, which can be around 5mW at 500-600mA of drive current. This is the fixed output you are seeing at these low input power levels. If you find that the power of your signal of interest is not high enough compared to the ASE, you may want to use spectral filters to attenuate the ASE signal. FBH1550-12 or FB1550-12 would be good choices to look into for the 1550nm center wavelength.

mdrolet (posted 2018-09-09 22:39:48.94)

Hi, Do you have any graph showing amplification in pulsed regime? Thanks, Mathieu

YLohia (posted 2018-09-17 02:22:57.0)

Hello Mathieu, thank you for contacting Thorlabs. Based on our discussion, your 150ns pulse width with 40 kHz rep rate supplied to this SOA will be perfectly fine, assuming that you use appropriate drivers/mounts. I have reached out to you directly with an eye diagram for an informal test we performed on one of our SOAs in the past/

jchg2758 (posted 2018-08-27 07:28:15.16)

Can I know the general recovery time of those c-band SOAs? I want to know how fast pulse train could be applied to this amplifier. Thanks.

YLohia (posted 2018-08-27 10:53:12.0)

Thank you for contacting Thorlabs. SOA1117S gain recovery time is not a parameter that we specify or measure, but we expect it to be in the 300-500 ps range (potentially on the faster side).

robert.supe (posted 2017-10-02 10:13:51.683)

in regards to product BOA1007C. We have several units and the height of each chip varies greatly. from 905um to 765um. I am unable to locate a height specification or tolerance on any of the data sheet provided on the website for the product. In order to proceed with our product development we need to know what the tolerances and specifications are for this product. Are there any which can be provided?

tfrisch (posted 2017-10-02 06:22:35.0)

Hello, thank you for contacting Thorlabs. We will reach out to you directly to discuss these tolerances.

hvdbrom (posted 2015-12-15 17:24:51.02)

Is this amplifier meant for CW or for digital signals? We need to amplify digital signals with pulse repetition rate of 10 Gb/s or 20 Gb/s.

besembeson (posted 2015-12-16 09:52:24.0)

Response from Bweh at Thorlabs USA: Yes these can amplify a 10Gb/s or 20Gb/s modulated optical input as well as cw. In fact, the original application of the SOA design was a telecom alternative to EDFAs. However, the peak power levels will be limited to the cw saturation power.

bdada (posted 2011-09-09 11:09:00.0)

Response from Buki at Thorlabs: Thank you for using our Feedback tool. We apologize for not providing the drawing for the BOA1007H on our website. We will add it shortly. In the meantime, please refer to SAF1093H, which has the same mechanical drawing as the BOA1007H. You can access the drawing on the right side of the SAF1093H product page linked below. http://www.thorlabs.com/thorProduct.cfm?partNumber=SAF1093H

jikim (posted 2011-09-09 11:34:06.0)

I would like to ask the mechanical dimensions of a BOA1007H. Before ordering such parameters are important for the design of our experiment. Thanks in advance.

jjurado (posted 2011-05-18 13:08:00.0)

Response from Javier at Thorlabs to jikim: Thank you very much for contacting us. You are correct, since SOA is polarization dependent, the gain at fix current and wavelength will vary if the polarization state of the input changes. So, the maximum and minimum gain curves just describe the window of gain changes vs. polarization state. The difference between the max. and min, gain values correspond, then, to the polarization dependent gain (PDG) spec in the Specs tab.

jikim (posted 2011-05-18 15:00:18.0)

In the performance plots of a SOA1013S, the maximum and minimum gain are shown. What does it mean? Are they the gains for different input polarizations, i.e. for TM and TE modes?

Thorlabs (posted 2010-08-26 14:54:46.0)

Response from Javier at Thorlabs to jikim: The chip lengths for all our SOA/BOAs is 1.5 mm, except for the SOA1117, which has a length = 1.0 mm. I will contact you directly regarding the refractive index, since this information is not currently readily available.

jikim (posted 2010-08-26 10:58:25.0)

Could you specify the length of the chip and its refractive index at 1550 nm?

Adam (posted 2010-04-23 12:15:21.0)

A response from Adam at Thorlabs to jikim: I apologize for the confusion. The fiber to chip loss is 2.5dB and the chip to fiber loss is also 2.5dB. The total dB loss from fiber to fiber is 5dB.

jikim (posted 2010-04-23 09:00:15.0)

Do you mean that fiber-to-chip loss is 2.5 dB and chip-to-fiber loss is also 2.5 dB? Or the loss of each side consisting of a surface of the chip and a fiber?

Adam (posted 2010-04-20 16:44:21.0)

A response from Adam at Thorlabs to Jikim. The coupling losses at each side is 2.5dB.

jikim (posted 2010-04-20 10:38:13.0)

Dear Thorlabs, I have a simple question on a fibre-coupled BOA (BOA1004S). In this device, a pair of fibers is connected. Thus, there should be a certain amount of the coupling losses, e.g. fiber to the chip and chip to the fiber. Could you let me know such coupling losses? They are very important parameter in my laser system. And Im looking forward to hearing from you soon. Best Regards, Jae-Ihn Kim =============================================== Dr. Jae-Ihn Kim - Scientific Assistant - Technische Universitaet Kaiserslautern Fachbereich Physik Erwin-Schroedinger-Str. (Postfach 3049) 67653 Kaiserslautern - Germany Tel: (49) - 631 - 205 - 2319 Fax: (49) - 631 - 205 - 3903 http://www.physik.uni-kl.de/bergmann/ ===============================================

Adam (posted 2010-03-18 16:56:29.0)

A response from Adam at Thorlabs to Sema: Previously, Covega has produced BOAs at this range, 1060nm. Thorlabs Quantum Electronics is planning on releasing a similar BOA in the next couple months. Will this work? At this time, there are no plans for the release of a SOA at this wavelength range, but if we can get more information about your application, I can suggest this as a new product idea. That being said, do you need a BOA or a SOA? Previously, most of our customer at this wavelength were either (1) making a laser for OCT or (2) amplifying 1064nm seed laser-- both of which a BOA is better suited than an SOA. I will email you to get more information for your application so we can suggest the most suitable product.

sema (posted 2010-03-18 10:20:02.0)

Covega also used to produce SOAs for 1060 nm. Are these devices still available?

View All »Matching Part Numbers

C-Band Optical Amplifiers (BOAs and SOAs), 1550 nm

Features

Booster Optical Amplifiers

Semiconductor Optical Amplifiers

BOA1004S and BOA1004P Graphs

SOA1013S Graphs

SOA1117S and SOA1117P Graphs

BOA1007C and BOA1007H Graphs

Booster Optical Amplifiers, Butterfly Package

Semiconductor Optical Amplifiers, Butterfly Package

Optical Amplifier Chips