Exactly how small is your spot size?  Are you sure it is smaller than the fibre core? What is the divergence of the beam?  Is the laser single transverse mode?

Have you checked the fibre ends for damage or dirt with a hand magnifier or low power microscope (with the laser turned off).

Is the spot focused onto the end of the fibre?  How does the coupled power vary as you adjust the translation stage?  Depending on the fibre length, you might be coupling into the cladding, which could give a false maximum as you adjust the transverse fibre position.  Does the coupled power change as you adjust the axial position?  Try Increasing the axial separation, search for a new optimum transverse position, then re-adjust axial focus and repeat.

What are the laser wavelength, fibre type and fibre length?  It sounds like you are using polymer clad silica core fibre.  Low-OH silica is preferred for longer wavelengths, but shows increased attenuation at shorter visible and UV wavelengths.  High-OH (flame hydrolysed) silica typically has lower attenuation for visible and near UV wavelengths, but has strong absorption around 950 nm, 1250 nm and longer near infra-red wavelengths.  https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=6845

Can you confirm that when the laser is  "set to ~60 mW" you measure 60 mW directly from the laser with your detector?

Hi Alan thank you so much for your response, addressing your comments:

- The spot size is approx. 2-3mm in diameter, I don't have a lens with a short enough focal length yet to make the spot smaller at the moment...  It's beam profile is Gaussian but with the addition of a straight line that goes from just outside the spot on top to just outside on the bottom.  I was just reading that a small cylindrical lens could help remove the vertical line from the beam profile (?)

- I'm not sure what the divergence of the beam is, it seems very collimated as the same small spot can be seen if iIshine the laser onto the wall at a distance

- Yes I believe the laser is single transverse mode

- I will check the ends for damage!

- Yes the spot is focused on the end of the fibre which has an SMA connector at its end

(this is the fibre: https://www.thorlabs.com/thorproduct.cfm?partnumber=M20L02&pn=M20L02#1510 -> 2 metres long, 0.39 NA, 400um core, low OH)

- The coupled power reaches a maximum when I move the stage holding the fibre to where it looks like the spot is centred well on the fibre.  I can't distinguish whether it's aimed at the cladding at that point (too small).  When I change the axial position I get a maximum coupled power when it appears that the fibre direction is in line with the laser beam (0 degrees from beam) 

- The laser's wavelength is 641nm and yes the power coming from the laser in free space was measured with a power meter and read 60mW 

- I am going to try what you said about adjusting the axial position, then the translation stage, axial & repeat tomorrow!

Thanks again!

What kind of laser are you using? Solid-state, semiconductor? Semiconductor lasers diverge very easily. 

Most likely, you need a microscope objective of 5-10X magnification to downsize your 2-3mm spot size. 

Sarah,

With a 2-3 mm diameter 60 mW collimated Gaussian beam, you should have no trouble coupling around 1 mW into a 400 micron fibre core - even without a lens.  https://en.wikipedia.org/wiki/Gaussian_beam#Power_through_an_aperture

When you said "very very small", I assumed a focused spot with diameter less than 100 microns.  If it is less than 5 mm diameter when projected onto a wall more than 1 metre distant, then adjusting the axial displacement of the translation stage by a few mm will not have much effect.  2 mm beam diameter is too large for selective coupling into the fibre cladding to explain your lack of light.

Does the fibre transmit light?  If you aim one end of the patch cord towards a window, or hold it near an incandescent lamp or an LED torch, you should see a bright spot on the other end, corresponding to the 400 micron core.  I have a 200 micron core cable here, and the core clearly transitions from dark to light when I hold the other end a few mm from a computer screen, and move it from a dark to a white area of the screen.

Is there another patch cord you could try? 2 m of fibre should not introduce significant attenuation at 641 nm.

I would have expected at least 0.1 mW even with coarse visual alignment.  At 641 nm, that should be enough to project a bright red spot onto a white card held a few mm from the fibre end.  Don't try this if there is any risk of specular reflection into your eye unless you can turn down the laser or attenuate the launch beam to a safe level.

Did you use the same power meter to measure the free space beam as you used for the transmitted power?  One concern was that the transmitted power might overload the detector.  The fact that you see variation as you centre the fibre or rotate the SMA mount makes this unlikely, but still worth checking.

How much power do you need from the fibre?  Once you have a working patch cord, you do not need a particularly short focal length to improve your coupling efficiency.  Anything between 4 mm and 500 mm focal length will help - assuming you have enough bench space to locate the fibre at the focus of the longer focal lengths.  A short focal length will increase the divergence of the beam at the fibre output, if this is significant in your application.

If the main spot is more or less circular, the vertical line could be due to diffraction from an edge or rectangular aperture.  Does it carry much power?  A cylindrical lens may help, but will also change the divergence of the main Gaussian beam.  If your principal concern is coupling efficiency, I would try a simple spherical converging lens first (or an achromat or low power microscope objective if that is more convenient).

Hope this helps.

Alan.

Thanks Mehmet I may try using an objective lens down the road, and I am using a diode laser currently

Alan - the fibre does transmit light, I see very faint red light being emitted at the distal end when the proximal end is facing the laser

Yes I was measuring the power from the laser and from the fibre using the same power meter - I will check again today to verify if I was overloading it

I need at least 8mW coming out of the fibre for my experiment

Yes space is a bit tight so I am looking to find a short focal length lens

I'm not sure whether the vertical line carries much power - I think I'll go with what you said and stick to spherical converging lenses/microscope objectives!

Someone in the lab may have a spare fibre-coupled laser for me so my troubles may all be over shortly - thanks for your contributions Alan and Mehmet!

Sarah

Sarah,

My guess is a faulty patch cord.  I would expect something like a thousand times more light (even without lens coupling).  Rather than a barely visible faint red light, you should be contemplating the efficacy of your blink reflex and hoping for swift recovery from strong retinal after-images.

If you don't have access to suitable test facilities, consider disposal of the patch cord, or at least label it "suspect until proven otherwise".  Minimum short term bend radius is 20 mm for this type of fibre, 40 mm radius for longer periods.  Could the cord have suffered an internal fracture?

Good luck.

Alan.

Dear Sara,

If still, you having the same problem, your fiber may be defective. Either its two  ends or the cable itself is half-broken.

If you really have no access to a proper focusing lens, look around maybe somebody is wearing strong reading glasses.

Take a fiber out of your experiment and bring to a room with windows. Point one end outside (do it during the daytime), and look at the other end. Do you see a bright end?

greetings every one!

We have to choose the material from which le laser is made; It must emit in the range where the fiber works with min of dispersion and dissipation.

Thank you very much for your responses! I have determined the patch cable to be faulty.  After trading in my patch cable, and better aligning my laser into my fibre with a collimating lens and using 2 silver front-coated mirrors, I have a much better coupling efficiency :-)

One tip if anyone else has this coupling problem is to take another laser and shine it through where the fibre is positioned - if its beam spot does not line up to the centre of the other laser's cavity, then the coupling efficiency will be reduced.

My last issue is that I have backscattering into my laser cavity, creating power fluctuations in my data, and so aiming my backscattered light (w/ random polarization) away from my laser cavity is a trade-off with the amount of light I am able to couple into the fibre itself.  I do not have an optical isolator unfortunately and since the backscattered light is randomly polarized in my setup I can't come up with any clever use of a combination of polarizers, QWP's and beam splitters to stop it from back-reflecting.

Thanks again!

If you have too much light out of the fiber, add a neutral density filter between the laser and the fiber. A scattered light will be absorbed in the filter.

Though the other power fluctuation can be a fiber vibration. How stable is the beam profile after the fiber?

Hi Egidijus, I cannot add a absorption ND filter because I already have such low signal.  Everything is very stable, clamped, and tightly screwed.  The fluctuations look more like power drift rather than noise,

Sarah

You may check if you can use a diverging lens in front of the laser output. That is a lens with convex-concave surfaces. The concave side should face the laser output. 

I will try that! Thanks!

Sarah, what type of fiber are you using (silica, POF?) and what is the wavelength of the laser?  Attenuation of POF at wavelengths higher than ~640nm, for example, is very high.

Hi Malcolm I am using these fibres:

https://www.thorlabs.com/thorproduct.cfm?partnumber=M20L02&pn=M20L02#1510

(0.39 NA, 400um core, low OH)

The laser is 641nm, and I think most of my losses are due to the fact that we do not yet have a perfect collimating lens for the fibre I'm using 

I used to couple the laser light into fiber. Put the fiber just around the laser output port with a adjusting rack. Adjust slightly and carefully. The loss can be minimize to 7dB or less.

Reducing the coupling loss between the laser and the fiber by using a spot-size converter (SSC) or using a thermally expanded core (TEC) fiber improves the L-I characteristics of the laser.

If still having the same coupling problem, check the SMA adapter in your setup.

Good luck! 

The optimum coupling is achieved by adapted the laser source with the acceptance angle of the optical fiber (taking in considerations the laser spot size), and of course the coupling element will increase the power coupling

The numerical aperture (NA) of the coupling lens and that of the fiber should be close for high coupling efficiency.

Conference Paper Underwater wireless-to-plastic optical fiber communication s...