Getting Started with Optics?

[CaptnYellowShirt]

That's what I'm doing. Specifically using large q-switched lasers for high speed illumination.

Here's the issue though: I don't know anything about optics. I forget if we even covered basic optics in Physics I/II

So I'm looking for a primer or book on the subject. Practical issues in:  setting up fiber optic launches, collimation of components, selecting the right components when considering the high power nature of the experiment, etc etc.

Any suggestions?

[retrolefty]

I don't know anything about optics.

Easy, rule #1, don't scratch them. 

[CaptnYellowShirt]

I got my methanol and kimwipes.

[TVman]

Don't bend them to much in one angle.

[SeanB]

I got my methanol and kimwipes.

Careful of coatings that dissolve in alcohols. Many antireflective coatings do so. 

Get some protective glasses that absorb the wavelength in use. Eyes come in pairs, and getting new ones is very expensive, and typically takes 30 years or so, and a good amount of teaching of the replacement.

[jeremy]

Some of my research is in an optics lab.

First of all: optics stuff is horrendously expensive. Having a single lens cost $500 is not too unusual. Lasers costing many thousands or tens of thousands doesn't really cause much pause either. The major trouble you get yourself into going cheap is tolerances. You're not paying $5000 for a red solid state laser, you're paying a few hundred in parts for the laser and the rest is on a guarantee that it works exactly as specified in the datasheet. If you treat a cheap red laser you got from ebay as perfectly 650nm, you're gonna be in a world of hurt, particularly with diode lasers. In general, they will be poorly aligned in their cavity, have multimodal operation (they emit light of more than one "colour"), any lenses will be mounted loosely and polished badly and the laser will have a wide tolerance on its fundamental mode (the colour that you see) which will depend strongly on things like drive current, temperature, etc. You also need a vibration-dampening table, but you could probably get away with just doing it on a big slab of rock.

Honestly, I think your best bet is to trawl ebay for any optics labs that are getting rid of old gear. Except no-one ever throws out lenses, or anything that is still useful for that matter; light hasn't changed much over the years.  sorry

Also, for the record we use IPA and Acetone for 99% of solvent related jobs.

[CaptnYellowShirt]

Any suggestions on a book?

[corrado33]

One of my colleagues likes "Optics" by Hecht. And it was used by our most recent Optics course.  (If you're crafty, you can find it online. I didn't of course, because that'd be illegal.)

http://www.amazon.com/Optics-4th-Edition-Eugene-Hecht/dp/0805385665

Oh and I agree, people NEVER throw away optics. We have optics in our lab that are burnt that my lab-mates won't let me throw away. Stupid if you ask me.

And when people DO throw away optics, it's generally because they have no idea what they are, and you have no way of finding out.

"Oh what's this random 1/2 in. mirror with no markings on it?" Me: "Trash." Without exhaustive testing (lots of time) you'll never determine what a random optic is.

If you REALLY want to buy cheap new optics, look at Thorlabs.  They'll have the cheapest stuff you can find. (And they often send you thor snacks in their packages!) (This is a big deal for a graduate student.)

EDIT: Yesterday when the Leica (microscope manufacturer) representative was in our lab, he told us NOT to use Kim Wipes on optics. Said they're too abrasive. Do yourself a favor and buy yourself some lens paper if/when you buy optics. The canned air stuff is good for getting dust off of optics (gently.)

[jeremy]

Hrm, it's really difficult to get a book on the practical aspects. I learned it all (and I don't know that much) by watching people who did and asking lots of questions. Probably a good place to look would be in google patents as a start. If nothing else, they often have references to books in them. If you'd like to read about how diode lasers work, I that this book is a relatively easy read. In terms of just optics, I'm fairly sure that Optics by Hecht is the standard prescription.

I hope your EM is up to scratch

Do you have a university library account? An alumni one or similar? If so, you definitely will want to dig into some review articles on high speed pulsed lasers. You may also get lucky and find some articles online at arxiv.

FYI, the latest tech in fast laser pulses is Kerr lens modelocking. You can get pulses on the order of femtoseconds. Also search for titanium-sapphire lasers, these are the ones I have seen most commonly.

[jeremy]

I should note: if you happen to be a machinist/you know a machinist, then you might be in luck. Basically everything in an optics lab: lasers, lenses, oscilloscope/multimeter/measurey things and a bunch of steel/anodized aluminium structures with 1/4-20 socket head screws. If you have the machines and the knowhow to make things with tight tolerances, you could probably get a fairly decent setup going (alas, I wish I had a metal lathe...). A good lot of stuff that people purchase from places like thorlabs is aluminium clamps, rods, stages, etc to keep everything aligned.

You may be able to get started buying a "breadboard" (yep, that's what it's called!) from here: http://www.thorlabs.com/navigation.cfm?guide_id=40

[corrado33]

I should note: if you happen to be a machinist/you know a machinist, then you might be in luck. Basically everything in an optics lab: lasers, lenses, oscilloscope/multimeter/measurey things and a bunch of steel/anodized aluminium structures with 1/4-20 socket head screws. If you have the machines and the knowhow to make things with tight tolerances, you could probably get a fairly decent setup going (alas, I wish I had a metal lathe...). A good lot of stuff that people purchase from places like thorlabs is aluminium clamps, rods, stages, etc to keep everything aligned.

You may be able to get started buying a "breadboard" (yep, that's what it's called!) from here: http://www.thorlabs.com/navigation.cfm?guide_id=40

I always marveled at how much people spent on optics holders. They're simply aluminum blocks with very fine threaded screws that adjust the angle in different axis. (And a spring thrown in for good measure.) As for the "breadboard". That's extremely easy to make, assuming you have some time, you're good at threading holes, and have a drill press (to ensure perpendicular holes).

[jeremy]

I should note: if you happen to be a machinist/you know a machinist, then you might be in luck. Basically everything in an optics lab: lasers, lenses, oscilloscope/multimeter/measurey things and a bunch of steel/anodized aluminium structures with 1/4-20 socket head screws. If you have the machines and the knowhow to make things with tight tolerances, you could probably get a fairly decent setup going (alas, I wish I had a metal lathe...). A good lot of stuff that people purchase from places like thorlabs is aluminium clamps, rods, stages, etc to keep everything aligned.

You may be able to get started buying a "breadboard" (yep, that's what it's called!) from here: http://www.thorlabs.com/navigation.cfm?guide_id=40

I always marveled at how much people spent on optics holders. They're simply aluminum blocks with very fine threaded screws that adjust the angle in different axis. (And a spring thrown in for good measure.) As for the "breadboard". That's extremely easy to make, assuming you have some time, you're good at threading holes, and have a drill press (to ensure perpendicular holes).

Ah but you see, that is not true. When I screw things onto the optical table, I rely on the fact that all of the holes are exactly in line (to tens of micron or better) so I can line up a laser with the centre of a lens. The top surface needs to be ground flat, otherwise the tilt will mess with the holes anyway. I don't really consider a drill press to drill perpendicular holes well enough. We've always used a big german CNC mill when we need custom stuff done.

I have a colleague who just threw away a few days work because one of the holes on the optical table was slightly off. He was pretty annoyed 

[corrado33]

Ah but you see, that is not true. When I screw things onto the optical table, I rely on the fact that all of the holes are exactly in line (to tens of micron or better) so I can line up a laser with the centre of a lens. The top surface needs to be ground flat, otherwise the tilt will mess with the holes anyway. I don't really consider a drill press to drill perpendicular holes well enough. We've always used a big german CNC mill when we need custom stuff done.

I have a colleague who just threw away a few days work because one of the holes on the optical table was slightly off. He was pretty annoyed 

Very true, but for a hobbyist who's just getting into optics, I don't think hole placement to the micron will be needed. Nice? Yes. Necessary? No. My laser setup is much more forgiving than yours (thankfully). That much precision may be required for some optics setups, but definitely not for all. Then again, if your drill press doesn't drill holes perpendicular, then that's a problem. 

We've had quite a few small "breadboards" built for a few different applications. Sure it was done by a machine shop, but they work just fine.

[CaptnYellowShirt]

Thanks for the suggestions! I'll check Hecht out.

We already have the breadboard(s), laser*, chiller, and a basic assortment of lenses and mirrors. The specific problem here is getting that beam inside of a water flume to perform some fluid mechanics research. The lab is dirty, the humidity is high, and the breadboard table seems to be a magnet for opps!-impact.

I'd really like to try a fiber approach, but as several of you have pointed out: this stuff isn't cheap.

You want to learn about electronics? No problem! There's a near limitless supply of components that can be had at prices that are less than the shipping cost. Get an idea and try it.

Want to learn about optics? Heh... A trial-and-error approach may not be warranted here.

So I've pulled the papers on thermal damage to fibers, run the numbers, etc etc. But the bottom line here is, if the setup is misaligned... its not a $0.30 diode on the line... its a $300 fiber. So I have a lot of questions about how people build, test, and troubleshoot these types of rigs. And that's the kind of resource I'm looking for.



*http://photonix.com/series-groups/dm-series/

[CaptnYellowShirt]

So maybe something more like:

Optical System Design, Second Edition by Robert Fischer ?

http://www.amazon.com/Optical-System-Design-Second-Edition/dp/0071472487

[jeremy]

I'm not sure about the book sorry, never seen it (but most of the books I have seen are basically just maths, so to each their own).

Assuming you have the equipment, I think most of optics is just about being super careful about everything. Know your tolerances for everything beforehand and get instruments that can measure them (ie vernier calipers instead of a ruler, a many digit bench dmm instead of a handheld). Then, you just go one step at a time and check all of your previous measurements as you go. If something is out of tolerance, stop and think about why. Then either go back a step, or start again. Don't force any screws and don't bash things in; think about why things are happening. I hope you don't find this belittling or insulting, I'm just trying to stress that this sort of stuff brings precision and carefulness to a whole new level.

If you're doing fluid mechanics, you probably have some background in CAD design. I've found it useful in the past to draw the entire thing beforehand and assemble it in inventor/solidworks/whatever.

And don't forget the golden rule: if something isn't working, it's because your assumptions are wrong. Maybe you assumed that the setup wasn't very sensitive to temperature. Maybe you assumed that a 15mA drive current was "about right" for your purposes. Maybe you assumed the beam was 650nm when it is actually 655nm. You get the idea. Apparently a guy nearby to our lab was trying to measure nanometre displacement without a floating optical table and was getting weird results; do you really think that you stamping on the floor has no effect on a nanometre scale?

The normal measuring and marking rules apply; always measure from a single reference face, beware of parallax error, etc. If you need to brush up on these, just go ask your friendly local machinist/fitter&turner.

I would honestly just give it a go, but replace your DPSS laser with a small diode laser and practice doing the alignment. Make sure the colour is similar if possible. If you have an invisible beam (which Nd:YAG is iirc, but I can't work out if your setup has been frequency doubled), get some detector cards: http://www.thorlabs.com/newgrouppage9.cfm?objectgroup_id=296

Although, if you actually have a 300W pulsed Nd:YAG, you could probably just see if the beam is there by seeing what stuff is melting . The lasers I play with are measured in mW and I'm plenty afraid of those

[CaptnYellowShirt]

The alignment laser is a good idea. I bought a small diode that's only a few nm off in wavelength. But the real problem is the beam out of the DPSS is (of my opinion) really poor quality:

https://drive.google.com/folderview?id=0B_c9c-jzSK16N3R0OGNSMUVBd2c&usp=sharing

Those pictures are taken 10-15cm off the end of the DPSS's head. By eye, at 0.75m the beam has expanded to something around 20mm.

So there's really no way to switch in the small diode.

But I am (just right now) thinking I could defocus the diode to mimic the DPSS. Which I guess would work for everything except the diameter-divergence product?

[miguelvp]

I was looking for a course on edx.org but couldn't find one for optics, but you can take this one, it has all the materials etc, just no one to grade you

http://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-637-optical-signals-devices-and-systems-spring-2003/index.htm

Here is the page on how the OCW program works:
http://ocw.mit.edu/help/get-started-with-ocw/

Also for your optical designs this is like the free Eagle for optics:
The free version is limited to 10 surfaces including the object and the image surfaces.

http://www.lambdares.com/buy/educators-and-students

Not cheap to buy even the light edition $1,250, to get upgrades it's $500 annually.

I just played around with it a long time ago, so I forgotten how it even works

IDL and MaTLab can be used probably, or khoros (later visiquest) but out of those I think only MaTlab survived, also simulink but that's probably expensive and you'll have to have a good understanding of optical matrices.

If you can get your hand on matlab then maybe this?
http://www.mathworks.com/support/books/book61010.html?category=14&language=1

You could enroll on an online edx.org course (free) like this one:
https://www.edx.org/course/ricex/ricex-elec301x-discrete-time-signals-1032

It has a matlab online app embedded on the website course, that might work. Or you can after you register purchase matlab educational edition. I haven't done it but other's have.
And tell them you are registered on an online Rice Universtity course.

Or even that you are enrolled on the MIT online course.

$50 for Matlab
$99 for Matlab and simulink.

Still OSLO seems more visual for optics

Edit: Optical benches are really expensive, maybe you can find a cheap one on ebay.

Edmunds optics have nice ones but I recall everything in there was really pricey. Also very careful with first surface mirrors, they scratch just by looking at them.

[Alexei.Polkhanov]

You may be able to get started buying a "breadboard" (yep, that's what it's called!) from here: http://www.thorlabs.com/navigation.cfm?guide_id=40

Also there are "optical rails"  - cheaper than table, but may work as well for many designs in terms of stability.

[jeremy]

I asked my colleagues who have dealt with high power lasers about this. They recommended two things: firstly, do alignment with a different laser (as I mentioned before), and make sure that you have plenty of beam traps to catch stray beams: http://www.thorlabs.hk/newgrouppage9.cfm?objectgroup_id=1449

As for your laser being bad, I disagree. Every laser I have ever worked with had a fairly significantly diverging beam with no additional optics. This is why the most important lens is the very first one, the collimating lens. You put the laser at the focal point of the lens and then you get a collimated beam out. If you align everything with respect to the collimating lens using a collimated beam, then it doesn't really matter what laser you use (within reason!), you can fairly easily swap between them. It also goes without saying that you should test your laser at the lowest possible power before turning it up.

You mentioned fibers earlier, you will probably need a fiber collimation kit if you are going to go down that road: http://www.thorlabs.com/navigation.cfm?guide_id=27

[johngom12]

I was looking for a course on edx.org but couldn't find one for optics, but you can take this one, it has all the materials etc, just no one to grade you

http://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-637-optical-signals-devices-and-systems-spring-2003/index.htm

Here is the page on how the OCW program works:
http://ocw.mit.edu/help/get-started-with-ocw/

Also for your optical designs this is like the free Eagle for optics:
The free version is limited to 10 surfaces including the object and the image surfaces.

http://www.lambdares.com/buy/educators-and-students

Not cheap to buy even the light edition $1,250, to get upgrades it's $500 annually.

I just played around with it a long time ago, so I forgotten how it even works

IDL and MaTLab can be used probably, or khoros (later visiquest) but out of those I think only MaTlab survived, also simulink but that's probably expensive and you'll have to have a good understanding of optical matrices.

If you can get your hand on matlab then maybe this?
http://www.mathworks.com/support/books/book61010.html?category=14&language=1

You could enroll on an online edx.org course (free) like this one:
https://www.edx.org/course/ricex/ricex-elec301x-discrete-time-signals-1032

It has a matlab online app embedded on the website course, that might work. Or you can after you register purchase matlab educational edition. I haven't done it but other's have.
And tell them you are registered on an online Rice Universtity course.

Or even that you are enrolled on the MIT online course.

$50 for Matlab
$99 for Matlab and simulink.

Still OSLO seems more visual for optics

Edit: Optical benches are really expensive, maybe you can find a cheap one on ebay.

Edit: Beyondtech Resources website has a lot information for free regarding fiber optics information, checkout the ebooks.
https://beyondtech.us/collections/cable-wire-resources

Edmunds optics have nice ones but I recall everything in there was really pricey. Also very careful with first surface mirrors, they scratch just by looking at them.