Fun with optical spectrum analyzers, spectrometers and monochromators

[_Wim_]

After some small discussion in the buy/sell/wanted forum (https://www.eevblog.com/forum/buysellwanted/wtb-used-old-osa-(optical-spectrum-analyzer)/) I thought this is a better place to continue...

How do you feed the laser into the fiber that goes into the OSA?
I reckon you can't just shoot a couple of wats down the fiber into the sensor as the energy will destroy the sensor.
The source I'm measuring will be around the 40 to 20watts range.. so for sure not something you want to give to a sensor.

Do you place the fibers ending close to the laser and take on the scatter? Or do you attenuate the energy somehow??

In my case the issue is more getting enough power into the MS9710c as the other way around. The MS9710c has a single mode fiber input, and both lasers I shown in the previous post only have a couple of mWatt power. So to get this coupled into a fiber I have to use a fiber coupler (see pic attached). Mine is unfortunately normally for multimode fibers, which makes it hard to get sufficient coupling into a single mode fiber.

With a 20 Watt fiber, you have to be VERY careful! I do not have much experience with such high power, but letting it scatter around is for sure not a good idea. Normally this should only be pointed at a beam dump. To extract a little power from the main beam I am not sure how this is typically done.  Beam splitters are typically max 1/99, so that still leaves way too much power over.  In any case, wear high quality safety glasses at all times!

I bought a fiber (vis-nir) from ebay with sma connector (seeing on the anritsu there is a threaded cap on there made me assume it is sma)
 

SMA will be unlikely, I would guess it will be a FC-input connector (multi-mode). This will be easier to couple into that single mode, I have been thinking if it would be possible to convert the MS9710C to multimode also.

The update rate on the anritsu seemed a bit slow for the measure on a video I recieved (sweep) but perhaps it can be sped up reducing it's resolution ?

As with any spectrum analyzer, speed and resolution are indeed a trade off.

[masterpj55]

yes 20watts to 40watts of invisible 1064~808nm or higher is quite risky!
These are pump diodes who do not come with a fiber couple or anything to correct the beam other then a FAC (fast axis correction) lens placed in front of the semiconductor output.

You really need laser protective glasses with the highest OD rating for the wavelenghts you are working with (not a chinese gamble of glasses).
I bought my glasses from uvex (germany) and one from laserscope (USA). That have the OD rating all listed.

The idea is to scatter it on a safe surface (concrete or a brick beam dump box etc) and then have the fiber take in the scattered light in a room where only I would be present during, pre measure it with a power meter to see how much makes it in the fiber and then put it into the spectrum analyzer. I have heard of beam splitters indeed but getting them and getting them for each application might be a lot more troublesome while potentially moving the mount closer or further from the source could basically act as a means to control how much energy makes it into the fiber this way.

As an alternative way: Perhaps multiple beam splitter stages are used to reduce the beam.. gradually more splitters are added and each stage measured with a handheld power meter till a desired intensity is reached? I have a laser power meter (continuous measuring) and also a handheld power meter (interval measuring) to do such a thing.

Most pumping diode sources have an terrible divergence by the way since they are meant to go through various optical stages still.
I was thinking to 3d print some kind of ring clamp (that doesnt twist the fiber) for holding the fiber in place.

I might have some friends that have fiber couplers for sale so I'll ask around.. but usually i work around the 300 ~400~500mw minimum so i reckon getting enough power isnt a big problem.

Can an OSA handle 5miliwatts safetly? 10miliwatts? I have no manual and I rarely hear mention of what powers an OSA should accept.

[Thomas]

We have a couple Yokogawa AQ6370C OSA's here at work, max power input is +20dBm which is 100mW.

[_Wim_]

We have a couple Yokogawa AQ6370C OSA's here at work, max power input is +20dBm which is 100mW.

Same for my MS9710C, 20dbm with internal attenuator on, 10dbm (10mWatt) without the attenuator

[_Wim_]

I might have some friends that have fiber couplers for sale so I'll ask around.. but usually i work around the 300 ~400~500mw minimum so i reckon getting enough power isnt a big problem.

I thinks it is easier to work with a power level that is just below the max the OSA can handle. Typically adjustment of a fiber coupler can be VERY sensitive, making the reading jump between almost nothing and too high. This is especially the case when coupling into a single mode fiber. As said before, due to the age of your MS96A, I suspect this will be a multimode fiber input, and this makes it normally quite a bit easier (9µm diameter vs 50µm diameter make a huge difference in surface area to couple into).

You really need laser protective glasses with the highest OD rating for the wavelenghts you are working with (not a chinese gamble of glasses).
I bought my glasses from uvex (germany) and one from laserscope (USA). That have the OD rating all listed.

Good to hear you have high quality glasses, but take into account that a 20 watt invisible beam in not only dangerous for your eyes...

As an alternative way: Perhaps multiple beam splitter stages are used to reduce the beam.. gradually more splitters are added and each stage measured with a handheld power meter till a desired intensity is reached? I have a laser power meter (continuous measuring) and also a handheld power meter (interval measuring) to do such a thing.

About using splitters or other optical components, take also into account they have maximum allowed power density (power/surface area), so the wider your beam, the easier they can handle in. A beam expander could be an option to achieve this.

Good idea about using the power meter first. If you put the open end of the fiber you are coupling into just before the power meter, it will give you an accurate reading of the power coupled into the fiber (and due to the fast update, it is also way easier to adjust the coupler). Once adjusted correctly, you should be ok to attach the fiber to the MS96A

[masterpj55]

Absolutely! even skin contact is very dangerous!  I have a surgical laser that I work with thats about 150watts of 1064 which is more dangerous then this pump diode. I cover areas that potentially can scatter a little bit of IR even (like the prism that splits up IR from green at the last stage) with cinefoil to reduce the scatter and therefore reduce the involved risks.. getting your hand into a beam is a nono.. using a special camera to even see the light that you normally cannot see also helps a ton.

In that case the beam is more collimated/focused so its worse then a pump source but still: its a lot of energy!

I have a beam expander somewhere in a drawer bin that will do the job! Beam expanders are usually used to expand a beam to reduce divergence to then steer a beam with larger galvo scanner mirrors.
Not new to high powered lasers but am new with this kind of measuring (I normally keep it to power meters without knowing the exact exact wavelength other then what datasheets specify. (on the 2nd hand market however you sometimes deal with binned diodes that aren't doing their promised wavelength and require more temperature calibration just like a DPSS (Diode pumped solid state) laser to gain the best modulation and efficiency.

I had the idea of buying a large fiber so the surface of the fiber is larger to catch more light without any crucial laser coupling..
I might do some test bench setup without the OSA and just a fiber with the couplings and see what the power meter picks up under various tests benches and see how good my control is over adjusting the power without adjusting the source.

And yes I would first meter what that fiber end is providing and then plug it into the osa only when I'm certain what is going through that fiber wouldn't damage it.
It's good to know it's a 10mw to 100mw tolerance for the osa sensor/monochromator. It's sadly a bit of a gamble because mine doesn't come with a manual and manuals I cannot find online. I've seen them on ebay sellers images but I doubt they would part with just their manual if they have a device for sale... Bit amazed nobody cared to scan the booklet in and upload it online at any point in time.

[_Wim_]

I had the idea of buying a large fiber so the surface of the fiber is larger to catch more light without any crucial laser coupling..
I might do some test bench setup without the OSA and just a fiber with the couplings and see what the power meter picks up under various tests benches and see how good my control is over adjusting the power without adjusting the source.

Going from a smaller fiber to a larger (moving in the direction of the light) is no issue, but the other way around makes the setup very unstable as the amount of light that gets though is then depended on the internal reflections of the fiber.

So if your MS96A is a MM 50µm input, I would not connect a larger fiber. Smaller is ok. For you wavelengths typical communication fibers patch cables are ok, which are dirt cheap, so no reason not to use them.

using a special camera to even see the light that you normally cannot see also helps a ton.

For your 808nm laser any normal CCD camera with the IR filter removed will work perfectly. If you buy a black & white industrial cmos camera these do not have any IR filter by default, and can be a very useful tool when playing around with lasers. Attached an example where I used a CMOS camera to characterize the beam quality of an 850nm 0.5mWatt laser (Anritsu MG0934E)
)

Software used:
https://micro-manager.org/
https://sourceforge.net/projects/beamprofiler/
Camera: uEye Camera UI-5240CP-M-GL REV1

For higher wavelengths I just repaired an Electrophysics Micronviewer 7290A  camera, see:
https://www.eevblog.com/forum/thermal-imaging/swir-electrophysics-micronviewer-7290a-user-manual/
https://www.eevblog.com/forum/thermal-imaging/electrophysics-micronviewer-7290a-heat-imaging-test-and-inherent-display-lag/

If I would come across the user manual of the MS96A, I will for sure post it here. Did you try to ask Anritsu?

[masterpj55]

Have a CCD indeed with the IR filter removed.

Regarding input it says 0.6  ~ 1.6 μm but i reckon that is the range it supports for input. I don't know if there is a way to determine multimode (other then through the manufacturer I reckon).

I wonder if those fiber patch cables really work.. aren't they made of glass.. From memory glass absorbs most of the deep IR wavelengths no?
Do you know a good source for these cables? most of the fiber cables I found with the wavelength range cost 120 USD and up "used" .. however they are also something like 200um to 600μm

Some really valuable information and makes sense on the internal reflection bit.
I think i should be getting a 62.5 μm/125 μm fiber patch cable with FC connector (The analyser hasnt come in yet) I found one on amazon.. it seems their wavelength rating is 850nm to 1300nm .. wonder if it reaches 808nm just fine.

Edited becaused of searching for patch cables got me cheap results)

[_Wim_]

Regarding input it says 0.6  ~ 1.6 μm but i reckon that is the range it supports for input. I don't know if there is a way to determine multimode (other then through the manufacturer I reckon).

Typically the fiber is also connecterized inside the OSA. If you can remove the fiber, it easy to see the dimension of the fiber with a fiber microscope. A fiber microscope is also very useful to check for damage & dirt, and they are very cheap on Ebay: https://www.benl.ebay.be/itm/Westover-Scientific-FM-L200-200X-Fiber-Microscope/174183811639?hash=item288e2a0a37:g:4voAAOSwtp9ePJtH

I think i should be getting a 62.5 μm/125 μm fiber patch cable with FC connector (The analyser hasnt come in yet) I found one on amazon.. it seems their wavelength rating is 850nm to 1300nm .. wonder if it reaches 808nm just fine.

808nm is for sure no issue. Attached the measured spectrum of an halogen bulb with a 2m MM fiber patch cable (=REF in plot) and with two identical fiber patch cables connected in series (so 4m total) (=Live spectrum in plot). A slight additional attenuation can be seen below 650, but it is not spectacular, so these are for sure quite usable even below is absolute accuracy is not a requirement.

[Kleinstein]

The usualy multimode fiber are still similar quarz material. So they should be good up to at least 1600 nm with low absorbtion. There are some very cheap ones for audio signals - these may use a glass  + plastics combination and these may not work well in the deeper IR, though the core is still glass.

For the single mode fibers it is no only the smaller area, but also the smaller NA that makes a difference. AFAIR this give an additional factor of about 10 for the coupling (in)efficiency.

[_Wim_]

The usualy multimode fiber are still similar quarz material. So they should be good up to at least 1600 nm with low absorbtion. There are some very cheap ones for audio signals - these may use a glass  + plastics combination and these may not work well in the deeper IR, though the core is still glass.

I have indeed used these for 1550nm laser sources without any noticeable issue.

For the single mode fibers it is no only the smaller area, but also the smaller NA that makes a difference. AFAIR this give an additional factor of about 10 for the coupling (in)efficiency.

Thanks, did not realize that their was such a big difference in NA also.

[cdev]

Only semi-related but still thought I would ask for an artist friend who is very into color.

At the consumer level, not for science, what are the cheap options for measuring and matching colors, that are compatible with a bunch of different computing platforms, and affordable?

A color measuring tool, for the visual arts, things like matching paint, monitor and sensor profiling, etc.

[_Wim_]

Only semi-related but still thought I would ask for an artist friend who is very into color.

At the consumer level, not for science, what are the cheap options for measuring and matching colors, that are compatible with a bunch of different computing platforms, and affordable?

A color measuring tool, for the visual arts, things like matching paint, monitor and sensor profiling, etc.

I would say typically something that works together with a smartphone like this: https://www.sherwin-williams.com/painting-contractors/color/color-tools/colorsnap-match-details

How accurate this is, I cannot say, I have absolutely no experience in this.

EDIT: should have watched the video first, this device can only match colors to the brand paint, so probably of no use for an artist. But the typical smartphone consumer apps is still probably the way to go.

[masterpj55]

Regarding input it says 0.6  ~ 1.6 μm but i reckon that is the range it supports for input. I don't know if there is a way to determine multimode (other then through the manufacturer I reckon).

Typically the fiber is also connecterized inside the OSA. If you can remove the fiber, it easy to see the dimension of the fiber with a fiber microscope. A fiber microscope is also very useful to check for damage & dirt, and they are very cheap on Ebay: https://www.benl.ebay.be/itm/Westover-Scientific-FM-L200-200X-Fiber-Microscope/174183811639?hash=item288e2a0a37:g:4voAAOSwtp9ePJtH

I think i should be getting a 62.5 μm/125 μm fiber patch cable with FC connector (The analyser hasnt come in yet) I found one on amazon.. it seems their wavelength rating is 850nm to 1300nm .. wonder if it reaches 808nm just fine.

808nm is for sure no issue. Attached the measured spectrum of an halogen bulb with a 2m MM fiber patch cable (=REF in plot) and with two identical fiber patch cables connected in series (so 4m total) (=Live spectrum in plot). A slight additional attenuation can be seen below 650, but it is not spectacular, so these are for sure quite usable even below is absolute accuracy is not a requirement.

Very very usefull! thank you! i bought the fiber microscope.. a tool at that cost for the future also to inspect laser fibers is very useful.

[_Wim_]

Very very usefull! thank you! i bought the fiber microscope.. a tool at that cost for the future also to inspect laser fibers is very useful.

Don't forget to turn off the laser first! 

[_Wim_]

Don't forget that to see the dimensions of the actual core you have to illuminate the fiber from the other end also.

Nice instruction how to inspect fibers:
https://www.thefoa.org/tech/ref/testing/test/scope.html

[_Wim_]

Another thing to watch out for with fiber patch cables, is the difference between non-angled (flat/PC/UPC) and angled (APC) connecters. For example you can have FC/PC connectors and FC/APC connectors. They will mechanically fit, but damage can occur when mixed up. APC connectors are typically green colored, but that is not always the case, so always check the tip of your fiber patch cable (and know what type you should plug in to your device). The different type of non-angled connectors can normally be mixed without any risk of damage.

More detailed description of the differences can be found here:
http://www.fiber-blog.de/en/270-what-kind-of-fiber-polish-be-used-pc-upc-or-apc.html

[masterpj55]

I assume i need the flat one .
Thanks will have a look!

[_Wim_]

I assume i need the flat one .
Thanks will have a look!

For multimode, I do not think APC exists, this is only for single mode fibers. And as the MS96A is most probable a multimode input, you will indeed need a none angled one. But it is good to be aware of the many different types.

[masterpj55]

Seem to struggle to find the flat one.. Most manufacturers don't specify it or have the PC one rather then the full flat one.

[_Wim_]

PC is ok. UPC also. As long as it is no APC you should be fine. These can be mixed. Core fiber is flat part of all of these, so it make good connection even if types are not identical

[masterpj55]

reviving the threada a bit. got a fiber.. it reads with LEDs fine.
Im not daring to put a laser through it yet until i have means to get the power down.

I bought an optical attenuator through ebay for a great price that has the supported wavelength range to push the power down.. so that's one strike to get power down a bit (my anritsu doesnt have a build in attenuator like newer units).

Going to experiment a bit more with ways to get power  down but still get readings (ive was a bit busy with other things hence my lack of progress)

[_Wim_]

Do you know the specs of your attenuator? The two I have (HP 8156A & Tektronix AO5012) cannot take more power than the OSA itself. If it is a free space optic based attenuator like this setup that is off course something else: https://www.newport.com/f/manual-variable-laser-attenuator-kit

 

[Tomorokoshi]

Only semi-related but still thought I would ask for an artist friend who is very into color.

At the consumer level, not for science, what are the cheap options for measuring and matching colors, that are compatible with a bunch of different computing platforms, and affordable?

A color measuring tool, for the visual arts, things like matching paint, monitor and sensor profiling, etc.

Start with X-Rite:
https://www.xrite.com/

It is probably difficult to find a good color measurement tool that can do both monitors and print well and still be affordable, although I'm not up to date with those technologies anymore.

[Tomorokoshi]

20 watts!

I worked on a project that used a 500 mW diode laser. When focused properly (improperly?) it could easily burn craters into the surface of the lens.

[masterpj55]

Do you know the specs of your attenuator? The two I have (HP 8156A & Tektronix AO5012) cannot take more power than the OSA itself. If it is a free space optic based attenuator like this setup that is off course something else: https://www.newport.com/f/manual-variable-laser-attenuator-kit

Yes I do!
It takes a 50um single mode fiber.
It's the Anritsu MN924A Optical Attenuator which is an optic based attenuator.. it's the only affordable one I could find.
I also had a friend who had a fiber coupler also for 50um single mode.. again not ideal but it's the only affordable fiber coupler I could find with the adjustements.

I did some more studying and some low wattage tests (not hooking it up to the OSA) and am aware coupling into single mode is hard.. I found videos that guided through the process very well.
The losses by trying to couple the beam into the fiber would actually be desirable in this case, along with a beam sampler from Thorlabs are kind of what I want to reduce the power to more acceptable levels. Then let the attenuator do the last hurdles to bring the power down in a more controlled manner.

I'd measure the output of the fiber at the end by using my handheld laser power meter before connecting it up to the OSA.

Thorlabs beam sampler can be found here: https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=913&fbclid=IwAR0cwAQSyAm__ppzoYDHB-BdkK-xIJ8zThJk8qeVR4en4-PNeRumn5DWD10

[_Wim_]

It's the Anritsu MN924A Optical Attenuator which is an optic based attenuator.. it's the only affordable one I could find.
 

That is indeed similar to the attenuators I use. But these can typically only handle something around 20-25 dBm.

I also had a friend who had a fiber coupler also for 50um single mode.. again not ideal but it's the only affordable fiber coupler I could find with the adjustements.
 

50µm is multi-mode, 9µm is single mode.  Your attenuator appears to be single mode,

The losses by trying to couple the beam into the fiber would actually be desirable in this case, along with a beam sampler from Thorlabs are kind of what I want to reduce the power to more acceptable levels. Then let the attenuator do the last hurdles to bring the power down in a more controlled manner.
 

Be careful counting on the losses due to the inefficient coupling. Typical problem is that at first to coupling is so bad you do not even see a signal on the OSA, and while adjusting the coupler (or by some vibrations of the table) you can suddenly hit the sweet spot having way too much power going to the OSA. Fiber couplers are VERY sensitive. If this would happen, I have no idea how your attenuator will then act, but I can imagine that if you severely overload it, it will start passing more power to your OSA, especially with the high laser powers you are talking about!

Don’t forget to calculate your power density. With a 40 Watt laser, it easy to destroy the beam sampler if the beam is too narrow.

It could be an option to place 2 of those beam samples in series so an attenuation of at least 100x is guaranteed independent of the polarization.