Create an Account  |   Log In

View All »Matching Part Numbers

Your Shopping Cart is Empty

Tapered Amplifier

  • 850 nm Center Wavelength
  • 1 W Output Power
  • Controlled, Collimated Output Beam
  • 14-Pin Butterfly Package

FC/APC Connector


FC/APC Connector

Related Items

Please Wait
Item # TPA850P10
Center Wavelength 850 nm
Small Signal Gaina 24 dB
Amplification Bandwidth CWL ± 5 nm
Operating Current 2.5 A
Output Power 1 W
Output Polarization State TEb
Case Operating Temperaturec 10 to 40 °C
Fiber PM780-HP
Fiber Length 1 m
Connector FC/APC, 2.0 mm Narrow Key
Internal Package Thermistor 10 kΩ
Steinhart-Hart Coefficients A: 1.1292 x 10-3
B: 2.3411 x 10-4
C: 8.7755 x 10-8
TEC Operating Current (Tcase = 25 °C) 2 A
TEC Operating Voltage (Tcase = 25 °C) 3 V
Internal Package Thermistore 10 kΩ
Laser Class 4
Compatible Controller LDC2500B
  • At 2 mW seed power
  • The polarization is parallel to the package's base.
  • Requires an adequate heat sink and non-condensing atmosphere.
Limited Stock Icon

These items will be retired without replacement when stock is depleted. If you require this part for line production, please contact our OEM Team.


  • Output Centered at 850 nm
  • 1 W Output Power
  • PM Fiber Input, FC/APC Connector
  • 14-Pin Butterfly Package

This Thorlabs' Tapered Amplifier consists of an optical amplifier integrated into an industry-standard, 14-pin butterfly package. Available at a center wavelength of 850 nm, this modular tapered amplifier is easy to integrate into larger systems. The output of the amplifier is free space. Thorlabs recommends using an optical isolator (IO-3-850-HP) to prevent back reflections from damaging the amplifier.

The input to the butterfly package is fiber coupled to simplify alignment procedures that customers typically have to perform when working with a traditional tapered amplifier. In addition, the butterfly package protects the amplifier itself from damage and contamination, thus yielding an extended lifetime. Thorlabs' tapered amplifier design also incorporates collimating and beam-shaping optics to produce a nearly circular, collimated output beam. The FC/APC connectors and PM fiber input enables connection to any type of seed laser, such as home-built or custom External Cavity Lasers, such as Thorlabs' line of ECL kits. For maximal coupling, the polarization axis of the input beam should be aligned to the slow axis of the tapered amplifier's PM fiber. For this amplifier, the slow axis is aligned to the key on the fiber connector. For more information on the needed input laser power and the associated power output from the tapered amplifier, please see the Graphs tab.

This tapered amplifier specifies both a maximum current and maximum power. These are tandem specs and therefore both must be observed simultaneously. In the normal course of operation, the chip will typically reach maximum power before reaching the current threshold. For example, with a seed laser power of 5 mW, the TPA850P10 will typically output about 1 W with a drive current of about 2.5 A (the maximum current output of the LDC2500B). Additionally, it is recommended that the seed power be a few mW in power (5 - 10 mW is a decent range for most tapered amplifiers) in order to reduce the load on the TEC. Due to inefficiencies in the chip, at very low seed powers (<1 mW) there is considerable heat generation and thus considerable effort on part of the TEC to properly maintain chip temperature. As the seed power increases, these inefficiencies are overcome and the heat generation is reduced. Over about 10 mW, there is no significant gain in thermal efficiencies; thus there is no reason, in terms of temperature regulation, to go above 10 mW input power.

These chips have a maximum operating temperature range of 10 to 40 °C. It should be noted that the package temperature must be kept low enough (<35 °C) for the internal TEC to properly pump heat away from the chip. Our LDC2500B dual current and TEC controller (available below) is designed specifically to meet the thermal and power needs of our tapered amplifier. It has two TEC's and specially designed heat sinks to address the problem of overheating. The case temperature of the tapered amplifiers is controlled by the second TEC in order to stabilize their gain and to keep the chip temperature under control (typically 25 - 30 °C). Other integrated driver systems, such as our CLD1015, and mounts, such as our LM14S2, cannot provide sufficient thermal control and could result in damage to the tapered amplifier if employed.

Raman Spectroscopy Applications

Raman Spectroscopy: The Approach

Raman spectroscopy is a spectroscopic technique that detects photons that have undergone Raman scattering. Since Raman scattering is relatively weak compared to Rayleigh scattering, one of the main historical problems with Raman spectroscopy had been separating out the weak Raman signal from the strong Rayleigh signal. Today, this problem is easily remedied with notch or edgepass filters. Similarly, recording the Raman spectrum has been aided greatly by the advent of CCD spectrometers. The image below shows a Raman Spectrometer system constructed using Thorlabs components, designed for 780 nm light using a tunable laser kit fed into a TPA780P20 Tapered Amplifier.

In this side-scattering configuration, the polarization of the laser was set vertically with respect to the table (horizontally polarized light cannot scatter horizontally). Isopropyl alcohol, contained in a cuvette and mounted in a CVH100 Cuvette Holder, was used as a sample. The cuvette holder allowed optical access to all four sides of the cuvette, making it ideal for a Raman spectrometer. The scattered light was collected by a fiber and fed into Thorlabs' CCS200 spectrometer. The Raman spectra for isopropyl alcohol, measured with this 780 nm Raman spectrometer, is presented to the lower right.

Raman Spectrometer
Click to Enlarge

A Raman spectrometer constructed from Thorlabs components.
Raman Spectrum for Isopropyl Alcohol
Click to Enlarge

Raman spectrum for isopropyl alcohol, measured with the 780 nm Raman spectrometer discussed above.

Power in Raman Spectroscopy

Power is important in Raman measurements. Raman scattering is a weak, low-probability event with a 1/λ4 wavelength-dependent efficiency. Sensitivity and integration time for data accumulation improve with increasing power, as long as the notch or edgepass filter can sufficiently attenuate the strong Rayleigh signal and the power is below the damage/saturation thresholds of the devices. To address this need, Thorlabs offers several laser diode packages for 785 nm that have output powers on the order of several hundred milliwatts, such as the LD785-SE400 or LD785-SEV300, capable of delivering 400 mW and 300 mW of power, respectively.


Laser Sources for Raman Spectroscopy

When using a laser to produce Raman scattering in a sample, sufficient power is important to improve the sensitivity and integration time of measurements. Thorlabs has invested in developing our GaAs material design and processing capabilities in order to produce reliable devices while improving the output power and available wavelengths.

A selection of Thorlabs' laser sources useful for Raman spectroscopy is outlined below.

GaAs-Based Fabry-Perot and DFB Semiconductor Laser Diodes
Our line of GaAs-based gain chips and diode lasers offers the possibility to improve the signal quality of an instrument by dropping in a higher power replacement or to design a new instrument around a customized diode. Traditionally, many Raman spectrometers have been built around 785 nm chips. Thorlabs offers two Ø9 mm TO can laser diodes at 785 nm, with one diode providing 400 mW of power (LD785-SE400), and the other, a wavelength-stabilized laser diode, providing 300 mW of power (LD785-SEV300).

For some applications, other wavelengths may be desired to balance the 1/λ4 Raman signal dependence with the background fluorescence of a particular sample. The GaAs material system can be designed to produce wavelengths from 630 nm up to about 1050 nm. In these cases, Thorlabs' OEM manufacturing capabilities may be able to develop a solution to suit a particular application. Please contact for more information or to discuss an application.

Tapered Amplifiers
To achieve powers of 1 W in a free-space beam, it is possible to use a tapered gain medium. Thorlabs offers a tapered amplifier for use with seed sources at 850 nm with maximum output powers of 1 W.

The butterfly package offers two major advantages to working with bare chips. First, the input side is fiber coupled so the seed can be fed into the amplifier via the FC/APC connector. Second, a lens system within the package on the output side helps to collimate and correct the strong astigmatism of the bare chip.

More information about the tapered amplifiers can be found here.


Delivering Light to the Sample

The Raman spectroscopy experiment described above was constructed using components from Thorlabs' extensive line of optomechanical components, optic mounts, and optics. The setup was built on a breadboard using our SM1 lens tube and 30 mm cage systems. Components were mounted using our Ø1/2" and Ø1" post assemblies and bases. Alternatively, our extensive line of optical rails provides another option for building a support structure for your experiment. In addition to our standard line of optic mounts, Polaris® low-drift kinematic mirror mounts are available for mounting optics in the system that require long-term alignment stability, particularly in conditions where the temperature may cycle in between experimental runs.

Thorlabs offers a wide range of mirrors which can be used to guide light through the system. Besides our plano broadband-dielectric- and metallic-coated mirrors, we also offer concave and off-axis parabolic mirrors that can be used to focus or collimate light within the system without introducing chromatic aberration.

Raman scattering produces a relatively weak signal compared to optical signals produced by other mechanisms, such Rayleigh scattering or fluorescence from the sample. Filters that can remove unwanted signals are an important part of a Raman spectroscopy system. Thorlabs' line of premium hard-coated edgepass filters and bandpass filters provide a solution. Our line includes longpass filters with cutoff wavelengths between 400 nm and 1500 nm and bandpass filters with center wavelengths between 400 nm and 1064 nm.

Fiber patch cables can be a convenient way to simplify light introduction or collection in Raman spectroscopy applications. They are available with a variety of fiber core sizes, connector types, and cable lengths. If you cannot find a patch cable suitable for your application, you can design your own custom patch cable using our custom cable configurator.

Protecting the Laser from Back Reflections
Back reflections from the optics in a setup can re-enter the cavity of a laser, potentially causing mode-hopping, amplitude modulation, frequency shifts, or even damage to the laser source. Optical isolators are passive magneto-optic devices that only allow light to travel in one direction, protecting the laser source from back reflections or other signals that may occur in the system after the isolator. Thorlabs' optical isolators are available in fiber coupled and free-space configurations with center wavelengths ranging from the UV to the IR. While our most popular isolators are shipped from stock to provide faster service, custom isolators are also available. Visit the Custom Isolator page for more information or to request a quote.

Sample Holder

Thorlabs' CVH100 cuvette holder was used to integrate sample solutions into the Raman spectrometer described above. The holder can be used with both macro and micro cuvettes and features four light ports. A filter can be added via a filter holder that fits into a slot on the top of the CVH100. The cuvette holder also comes with a fiber adapter that includes a collimating lens, allowing output light to be fiber coupled into a detector.


Thorlabs offers a line of CCD spectrometers that can be used to record the spectrum produced by a Raman spectroscopy setup. These compact spectrometers feature an SMA connectorized input and come with an SMA-to-SMA connectorized fiber patch cable. Each spectrometer is wavelength and amplitude calibrated. The spectrometers are controlled via a software package that includes a graphical user interface and extensive set of drivers for data acquisition and analysis.


Vibrational States of CO2 Molecule

Raman Spectroscopy: The Basics

Discovered by Krishna and Raman in 1928, Raman spectroscopy has given rise to a multitude of specific techniques, from Linear Raman Spectroscopy to Coherent Anti-Stokes Raman Spectroscopy, and proves itself to be a powerful tool for spectroscopic analysis. One of the most common applications of Raman spectroscopy is to measure vibrational, rotational, and other low-frequency modes of a system (e.g., molecules).

Raman scattering by molecules is a type of inelastic scattering, in which the final and initial energy states are different, resulting in the molecules being in a different quantum state. Raman scattering is in contrast to Rayleigh scattering, an elastic scattering event, in which the final and initial energy states are the same (i.e., energy is conserved), and the molecules remain in the same quantum state. Both Rayleigh and Raman scattering are dependent upon the polarizability of a molecule: the stronger the polarizability of a molecule, the larger the scattering cross section. While both Rayleigh and Raman scattering are second order processes that scale as 1/λ4, the scattering rate for Rayleigh scattering is on the order of 103 times greater than that for Raman scattering [1]. Typically, in Raman spectroscopy, the stronger Rayleigh signal must be extricated since it carries little pertinent information about the vibrational modes.

Stokes and Anti-Stokes Radiation

Since Raman spectroscopy requires the polarizability change to be a function of normal coordinates, one of its limitations is that it cannot measure direct dipole transitions. Because of this, Raman spectroscopy is sometimes utilized with other techniques to fully measure the vibrational and rotational states of a molecule. For example, in the CO2 molecule, of the three vibrational states depicted in the figure to the right, only ν1 (symmetric stretching) is Raman active. The other two vibrational states (bending and anti-symmetric stretching) are infrared active [2]; thus, Raman and infrared spectroscopy comprise complementary measurements.

Raman Spectrum for Acetone
Click to Enlarge

Raman spectrum for Acetone, measured with a 532 nm Raman Spectrometer (bottom), and compared to published results (top).

Raman scattering is a two-photon process, wherein the incident photon (hνi) is absorbed by the molecule, and the molecule is excited to a "virtual" level (not necessarily a stationary Eigenstate). Once promoted to this virtual level, the molecule will decay to an excited state and emit a "scattered" photon (hνs). In general, the molecule begins in the ground state, and thus, the energy of the scattered photon is less than that of the incident photon. The energy difference is related to the vibrational, rotational, or electronic energy of the molecule [2]. The emission of a scattered photon possessing less energy than the incident photon is called Stokes radiation, whereas the emission of a scattered photon possessing more energy than the incident photon is known as anti-Stokes radiation. The figure to the left depicts Stokes and anti-Stokes radiation. Since anti-Stokes radiation requires that the molecule already be in the excited state before scattering, the peak intensity of the anti-Stokes signal is lower than peak intensity of the Stokes signal.

The graphs to the right show the results of a typical Raman spectrum for acetone, taken with Thorlabs' DJ532-40 laser diode, compared to published results. For standard linear Raman spectroscopy, information about the molecule is obtained through several measurements. The linewidth of the scattered radiation can yield a plethora of diverse information about the system. For example, in a gas sample, the linewidth can represent Doppler width, collisional broadening, natural linewidth, etc. Polarization analysis of the Raman spectrum also yields additional information about anisotropy and the polarizability tensor. Additionally, information about molecular orientation or vibrational symmetry can be extracted from polarization analysis. Finally, the intensity of the Raman lines relates to the scattering cross section and population density of molecules in the initial state.

[1] D. W. Ball, Spectroscopy 16(2), 28 - 30 (2001)
[2] W. Demtroder: Laser Spectroscopy Volume 2, 4th Edition (Springer-Verlag, Berlin, Heidelberg, 2008)
[3] G. Dent and E. Smith: Modern Raman Spectroscopy: A Practical Approach, (Wiley, Chichester, United Kingdom, 2005)
[4] I. R. Lewis and H. Edwards: Handbook of Raman Spactroscopy, (CRC Press, 2001)
[5] R. L. McCreery: Raman Spectroscopy for Chemical Analysis, (John Wiley & Sons, Inc., 2000)

Tapered Amplifier Diagram

Posted Comments:
jsw419  (posted 2018-04-06 11:02:53.937)
Please could you inform me of the beam divergence, regarding your tapered amplifiers (TPA780P20)
YLohia  (posted 2018-04-10 09:25:21.0)
Hello, thank you for contacting Thorlabs. The typical FWHM divergences for the TPA780P20 are: Parallel: 10 degrees Perpendicular: 33 degrees
chenav  (posted 2018-02-13 21:29:39.333)
Can the TPA780P20 amplifier be used at 795 nm? Do you have any graphs showing gain vs wavelength? Unfortunately the web page and data sheet give conflicting information, citing 20 nm and 10 nm amplification bandwidth respectively.
YLohia  (posted 2018-02-22 05:18:40.0)
Response from Yashasvi at Thorlabs USA: Hello, thank you for bringing this inconsistency in specs to our attention. The +/-5nm bandwidth on given on the spec sheet is correct. Our Tech Marketing team is working on fixing the misprint. Unfortunately, we do not currently have the Gain vs. Wavelength data.
chenav  (posted 2017-08-26 19:30:17.24)
I read comments about "thermally activated aging" of the TA chip when operated without a seed laser. I wanted to verify if this is a problem also when the seed laser is shuttered with very high duty cycle, i.e. ~95% ON / 5% OFF with a period of 1 second. Would it be safe to assume, to the best of your knowledge, that detrimental thermal effects on the chip would be minimal under these circumstances? Thank you.
tfrisch  (posted 2017-09-20 01:15:54.0)
Hello, thank you for contacting Thorlabs. In addition to thermal damage, running the tapered amplifier without a seed will cause stored up energy and back reflections can cause immediate damage to the chip. This would be an instant failure rather than reduced life. I'll reach out to you about your application to discuss whether your duty cycle and unseeded time would be cause for concern.
user  (posted 2016-07-11 14:02:21.453)
Hi, I just have a short question regarding the TA chip. I understand the lifetime is supposed to be 10 years, but in the event of something catastrophic occuring such as improper seeding or overheating (and burnout) would a purchaser be able to access the chip to replace it? Or would a whole new package have to be purchased?
bethany  (posted 2016-05-18 14:34:50.307)
I have a related question to alexeyzaytsev's question below. Do you have any more information about the behavior and lifetime effects of using a 1ns pulsed input to seed the TA? (We can generate such a pulse using an EOM on a CW laser, but would like more power than the EOM allows.) Any information about peak power, rep rate, temperature handling, extinction/background and effects on the diode lifetime would be appreciated. Thanks!
besembeson  (posted 2016-05-19 02:57:44.0)
Response from Bweh at Thorlabs USA: The gain chips should be okay for the 1ns pulsed seed laser. I would recommend however keeping the average input power to under 10mW. This could be higher depending on the amplifier drive current but the average output power should always be under 1W or 2W, model dependent. Running the tapered amplifier without a seed laser could cause damage due to thermally activated aging so the lifetime of the diode going through input power cycles could be affected but we can't determine by how much as it seems the time constant for the thermal effects will be much longer than the repetition rate. Also, should the repetition rate be too low, instances of self-lasing could occur that has the potential of damaging the chip. We have not done studies with such ns seed lasers to comment on the other performance characteristics requested.
gkatsop  (posted 2015-09-22 22:22:33.817)
Hello. Wavelength question: We would be very interested in an amplifier for 1315nm. Is it possible/plausible to make one? Thank you, George K.
besembeson  (posted 2015-10-07 10:23:07.0)
Response from Bweh at Thorlabs USA: Thanks for your interest. I will follow-up with you regarding this.
hammae  (posted 2015-06-19 08:04:05.02)
What are the electrical properties of the TPA780P20 Tapered Amplifier gain chip? IV curve? capacitance? electrical optical modulation BW ? I would like to design a bias-tee or a modulator for this chip and I need to evaluate these parameters.
besembeson  (posted 2015-09-21 01:36:42.0)
Response from Bweh at Thorlabs USA: I will followup with you regarding these specifications and further discuss your application.
cqtch  (posted 2015-04-22 12:40:22.643)
Dear Thorlabs Team, I have a question concerning the beam profile of the TPA780P20. In our case the power output seems fine, but there is a considerable structure on the beam profile. Also, close to the TA (order of some cm) the beam is first focussing and then expanding. Is there a way to tune the beam profile by for instance the temperature or is there a way to change the collimation ? We would like to couple the TA into a fiber and my guess was that we should be able to couple a considerable amount nevertheless with the current profile we couple about 40%. Thank you very much. regards, Christoph
jlow  (posted 2015-04-29 03:48:51.0)
Response from Jeremy at Thorlabs: The structure in the beam you observe is normal as can be seen in the Beam Profiles tab on the webpage. Also, the beam profile does evolve at different distances as you have seen. Unfortunately there's no way to change the beam collimation since the collimation lens is inside the butterfly package. The output beam profile can be changed slightly with different drive current but that can possibly change the beam pointing slightly as well. To fiber couple this, the best approach is to use fibers that have specially fused glass sections at the input (so-called ‘end capped’ fibers) so that the power density is reduced at the actual air to glass interface. If you attempt to use standard single mode fibers where you are trying to focus light down into the small core of the fiber then it’s advisable to start off at low powers and gradually increase the current as the alignment is continually optimized.
sonnenfroh  (posted 2015-04-03 12:32:57.067)
Are other wavelengths possible? I would be very interested in a device at 830 nm.
jlow  (posted 2015-04-14 08:40:22.0)
Response from Jeremy at Thorlabs: We do not currently have a 830nm TPA but we have recently started the development for this.
gnishi  (posted 2014-07-09 10:38:38.1)
Dear Sir: I have a question for pulse driving of TPA780P20 for pulse picking, modulation or shaping of a seeding laser. Our seeding laser will be a pulsed laser diode (FWHM~500ps, rep.1-10MHz, power about 10-100uW) and I would like to ampilify the pulse energy as well as to reduce the pulse width of seeding laser. I consider that this can be done if the TPA780P20 driving current can be drive by very narrow pulse current, such as the avalanche pulse generator. Do you have any information or suggestion on the pulse driving of this product? Regards, Goro
jlow  (posted 2014-08-07 03:51:41.0)
Response from Jeremy at Thorlabs: The butterfly package itself is limited in bandwidth to around 1GHz so it will probably not work to further shorten your short pulse length. There's also some concern as to the peak power of your pulse since it's not clear how you define your power in the post. We will contact you to discuss about this further.
bruno.viaris  (posted 2013-11-14 12:50:22.49)
Dear Thorlabs team, it seems that the LDC2500B is no longer available, so I suppose that a standard butterfly mount like LM14S2 can be used, but what kind of temperature controller should I use? Thanks.
jlow  (posted 2013-11-14 10:54:45.0)
Response from Jeremy at Thorlabs: The LM14S2 should not be used with the tapered amplifier because it cannot handle the heat load given out by the tapered amplifier. We are going to have the LDC2500B available again within a few weeks.
laugh007smile  (posted 2013-05-09 09:38:51.22)
Dear Thorlabs team, i notice that TPA780P20 is coupled with fiber at input and free space beam at output.So my question is can this TA couple with fiber at output derectily like input? Another question is I would like to know the beam size at the collimator and beam divergence.Can you send me more data about this TA. Regards,Smile.
tcohen  (posted 2013-08-05 11:04:00.0)
Response from Tim at Thorlabs: In general the beam profile emitted from the tapered output side of the chip has quite a lot of structure and within the same lot of amplifier chips there are differences in the quality of beam that they produce - particularly with respect to the distribution of low intensity side lobes that appear in the lateral beam profile. Coupling these side modes into fiber will therefore be difficult but the free space quickly becomes a well-defined single Gaussian like central lobe – as the side lobes disperse. We have sent you information on the beam evolution and will continue to add more information to the website.
b.biedermann  (posted 2013-05-03 09:55:00.37)
Dear Thorlabs team, how much output power can I expect when seeding the TPA850 with 10mW at 871nm? Thank you, Benjamin
cdaly  (posted 2013-05-21 09:59:00.0)
Thank you for using our feedback tool. Unfortunately, we cannot specify a power level at this time. From our experience, we believe 871nm might be too far away from the gain peak to give good amplification and mode purity, against competition from the tendency to self-lase.
alexeyzaytsev  (posted 2013-05-01 23:54:09.597)
Dear Thorlabs team, Can I use a pulse laser of 2-3 ns duration and 1-10 kHz repetition rate as a seed? If yes, how much peak power I should expect on the output? Thank you and best regards, Alexey
tcohen  (posted 2013-08-05 11:07:00.0)
Response from Tim at Thorlabs: The chip dynamics should be easily fast enough to deal with a pulsed input. The peak power should not exceed a level where the time averaged output power is greater than 1W. However, TPA units can be damaged when run without a seed source due to thermally activated aging and so the duty cycle will likely play a factor. Unseeded 2W chips will generally give under 1W of optical output and the rest has to be dissipated as heat (you can see this in the TEC current increase that is seen when the seed is switched off). Although the temperature at the chip on submount thermistor is maintained by the increased TEC current, the ridge temperature itself may run hotter. For this pulsed input, however, it is probably ok since the time constants for thermal effects will be very long compared to the on/off times. More testing needs to be done to confirm this theory.
user  (posted 2013-03-25 11:22:18.65)
You quote "extended lifetime" in your literature. Do you have any data to back that up? What is the expected lifetime?
jlow  (posted 2013-04-15 08:50:00.0)
Response from Jeremy at Thorlabs: Thorlabs buys the Tapered Amplifier Chips from a third party. The basic semiconductor technology has been investigated extensively, verifying in certain cases/designs/specifications calculated lifetimes of >10yrs. For reliable operation of Tapered Amplifier Structures, it is important that a seed laser is always present, as operation without an amplified output of a seed laser will shorten device lifetime. The packaging of the devices is carried out by Thorlabs at the Thorlabs Quantum Electronics facility, and lifetime tests have not been performed. However, the packaging process employs similar techniques that are currently used on Thorlabs’ SOA and BOA devices that we currently manufacture, which are in turn qualified to Telcordia 468 standard for use in telecoms systems’
alexfestbaum  (posted 2013-02-24 19:13:08.903)
Hey Thorlabs team, I would also be interested in the evolution data of the beam profile from the output facet as you mentioned in the last comment. Regards, Alex
tcohen  (posted 2013-03-07 15:55:00.0)
Response from Tim at Thorlabs to Alex: Thank you for contacting us! You can find a small summary of the profile behavior earlier in the feedback. We have some data compiled and I’ll send you that directly right now. We are also looking to build up the presentation on the web, so thank you for sharing your very relevant question with us. In the meantime, we can be contacted and can share this data immediately by emailing us at
tcohen  (posted 2013-01-14 13:46:00.0)
Response from Tim at Thorlabs: We may be able to supply a device without the lid and without any output optics if it is of interest. We will contact you to discuss the details. As for the beam profile, we have data showing the evolution of the profile from 200mm to ~1m. Very close to the package (a few mm) there is some noticeable structure in the beam, with a central lobe but also prominent side lobes. From 200mm or so the side lobes are not so prominent and you are mainly looking at a relatively well behaved central lobe. The vertical divergence will vary in assembly. However, the horizontal beam waist is always wider and as the horizontal divergence is smaller, the larger divergence of the vertical axis will give us a circular beam profile at ~300mm. We will contact you to discuss the custom and to provide these beam profiles.
jikim  (posted 2013-01-08 18:23:24.897)
I would like to know the beam size at the collimator and beam divergence. And is it possible to customize the collimator? Or could you manufacture an amplifier at 780 nm without collimator so that I could implement appropriate lens for our application? If it is, please quote a module.
tcohen  (posted 2012-10-18 12:29:00.0)
Response from Tim at Thorlabs: Thank you for contacting us. I have sent you some data and we will look into providing more information on our website.
mcox  (posted 2012-10-12 06:50:21.673)
Please could you inform me of the beam divergence, regarding your tapered amplifiers (TPA780P20)
sharrell  (posted 2012-03-22 14:26:00.0)
Response from Sean at Thorlabs to parkse: Thank you for your feedback. Our design team informed us that the amount of cooling power from the fan is necessary to dissipate the heat and keep the driver running properly without thermal runaway. If the fan is too noisy, there are two possible solutions. The first is to mount the driver onto a water-cooled breadboard ( A second solution would be to mount the driver onto a large solid aluminum breadboard which would act as a heatsink. In both cases, it would be possible to then disable the fan. We have contacted you directly to provide further support.
bdada  (posted 2012-03-21 17:52:00.0)
Response from Buki at Thorlabs to parkse: Thank you for your feedback. I have passed your comments on to our production team and we will contact you shortly to discuss this further.
parkse  (posted 2012-03-21 02:49:05.0)
I'm using TPA780P20 with LDC2500B. The cooling fan in driver(LDC2500B) is too noisy for some application such as presision laser frequemcy stabilization.
sharrell  (posted 2012-02-27 15:39:00.0)
A Response from Sean at Thorlabs: Thank you from your feedback. Regarding your second question, we added additional information to our webpage shortly after your comment stating that The beam profile data were measured 300 mm from the facet. Regarding your first question: Firstly an isolator (IO-3-780-HP or IO-3-850-HP, depending on wavelength) should always be present between the amplifier and the fiber. From experiments that we have done, we recommend coupling light into a fiber by placing the coupling optic (a Thorlabs 352230-B lens) approximately 300mm from the amplifier package and using an XYZ positioner to fine position a fiber (780HP single mode fiber) to collect the focused light from the lens. Our recommendation is that the collecting fiber is uncoated, but is angle polished (APC 8 degree polish) to prevent reflection both back into the source and also internal to the fiber; the fiber axis then has to be at an angle of 3.64 degrees to the incoming focused beam axis to handle refraction at the air fiber interface. I apologize for the delay in posting some of this information. We do not have your email, but if you have any other questions, please contact us at
user  (posted 2012-01-30 08:20:24.0)
Which fiber collimator in your cataloge can be used with the LDs in terms of damage threashold? How is the beam profile measured (with your collimate lens)?

Tapered Amplifier

Based on your currency / country selection, your order will ship from Newton, New Jersey  
+1 Qty Docs Part Number - Universal Price Available / Ships
TPA850P10 Support Documentation
TPA850P10850 nm Tapered Amplifier, 1 W, 10 nm BW, Butterfly Pkg, PM Fiber, FC/APC

Tapered Amplifier Controller

  • Designed for Thorlabs' Tapered Amplifiers
  • Integrated Current and TEC Controllers
  • Compatible with 14-Pin Butterfly Packages
  • Drive Current up to 2.5 A
  • Dual TEC for Chip and Package Temperature Control
  • USB Connectivity or Standalone Operation

Thorlabs offers the LDC2500B Tapered Amplifier Controller that is ideal for use with the Tapered Amplifiers, as well as other standard 14-pin butterfly packages where case temperature control is required. For more information on this product, please see the complete presentation here. Please contact Tech Support if you have any questions.

Based on your currency / country selection, your order will ship from Newton, New Jersey  
+1 Qty Docs Part Number - Universal Price Available / Ships
LDC2500B Support Documentation
LDC2500BTapered Amplifier Current and TEC Controller, 2.5 A, 14-Pin Butterfly Package
Log In  |   My Account  |   Contact Us  |   Careers  |   Privacy Policy  |   Home  |   FAQ  |   Site Index
Regional Websites:East Coast US | West Coast US | Asia | China | Japan
Copyright 1999-2018 Thorlabs, Inc.
Sales: 1-973-300-3000
Technical Support: 1-973-300-3000

High Quality Thorlabs Logo 1000px:Save this Image