Speaking* of green laser diodes.

the DVD trick is neat,

It's a little tougher with laser modes, but perfectly doable. It's very instructive to put a photodiode where it measures just one mode and see just how noisy it is. Typically the fluctuations are of order unity.

I've also used it to measure the tuning range of single-frequency diodes between mode jumps. You just tune till the spot suddenly changes position, and voila.

Cheers

Phil Hobbs

Right, I've done that with a spectrometer and single mode diodes.

Can I ask a question about these multiple longitudinal modes? How the Heck does That happen!? It implies multiple optical path lengths. I know the carrier density (how much current you are pushing through the diode) can change the optical length*... but still at one time it seems like there should be one optical length.

Is it lasing at many modes at once.. or jumping around between modes all the time? I think if it's the later, there's some chance you could do some external feedback and make it single mode at low currents, (above threshold)

George H.

That question seems to make sense only if you think of lasers along the lines of stimulated emission of identical photons.

I tend to think of it as oscillations in a medium that has gain over a certain bandwidth that covers many modes at once. Each mode has a slightly different *frequency*. The path lengths are the same, but the wavelengths are not.

Jeroen Belleman

DK link provided.

There's a deeplink at the end so be careful.

No. As you noted upthread, you can have 2000, 2001, 2002, etc cycles per round trip. OPL is very nearly the same for all, but they're at different frequencies.

OPL is measured in metres. The eigenmodes of a given cavity form a comb in 1/lambda. In the absence of dispersion, that's an equally spaced frequency comb.

Generally it's lasing in many modes at once, though I suppose it's possible to make a system that has one mode at a time. A nominally single-frequency laser can hop between modes, as we both know very well. ;)

A lot of lasers are single-mode near threshold. At any given temperature, there will be one mode that crosses threshold first, and if the gain delta between modes is big enough, it'll maintain itself stably.

I recommend Wolfgang's pages, and , for a lot more about diode laser behaviour.

It's somewhat rambling and poorly organized, but his plots of laser noise vs. temperature and frequency are super illuminating. Specifically, he evaluates a lot of diodes for ECDL operation at

Cheers

Phil Hobbs

Well, that is how I think about it. There's nothing 'wrong' with the photon picture of nature. And sometimes it's a much easier picture.

Hmm, there's some Fabry-Perot (FP) cavity that defines the modes. Feedback is needed, it's not a single pass gain medium. (I think... It was years ago, I tried to measure the output facet reflection coef. of a diode laser.. 90% ?)

Page 6 does the finesse, and I guesstimated the FSR above.. (Well I didn't put in the index of refraction (or the factor of two))

So once one of these modes gets above threshold.. (which one that is may be a little random.) It starts to lase... and stimulated emission as you say, takes over. Maybe it's a failure of my model (which is fine by me) but I can't see how a laser does multiple longitudinal modes at once. Some oscillation, progression between modes.. that I could understand. (It could even be chaotic mode hopping.)

George H.

OK, I think it's a mistake in my model. Should I think of it as more gain than one mode can handle? On a microscopic level, (as you said above) there are nodes in any single mode. An electron hanging around at a node, looking to make a transition to the valence band, may find a different mode to dance with.

Thanks for the links. I'll check 'em out.

George H.

Huh.. I'll have to read a bunch of that. My experience with one diode laser is that the position of the coupling lens is the most tweaky adjustment. And putting a layer of teflon tape on the threads of the Thor labs lens helps a bunch in removing hysteresis in the threads. (and other wobbles)

George H.

Yup, backreflections from the lens (and even the window in the package) are a problem. Stick-slip in the threads is a nuisance, though I haven't tried your teflon tape trick.

Cheers

Phil Hobbs

To make the high power blues and greens usable we often need cylindrical le ns pairs or anamorphic prisms. The diode die is quite wide in the big ones, and a fast axis divergence of 35 degrees or more from the uncollimated dio de is not unusable. To get the traditional "tight" beam from these we use a 3-4 mm focal length lens.

Steve

lens pairs or anamorphic prisms. The diode die is quite wide in the big one s, and a fast axis divergence of 35 degrees or more from the uncollimated d iode is not unusable. To get the traditional "tight" beam from these we use a 3-4 mm focal length lens.

Hi Steve, if that was in response to my lens comment, I should clarify and say I was talking about the lens in a ECDL. It controls the coupling between the diode and the grating.

From Phil's link. The osram PL520 works in a ECDL.

Fun stuff.

George h.

Teflon (PTFE) tape is available from plumbing suppliers.

It lubricates the thread, so you can do the joints up more tightly - to the point of bursting the joint if you overdo it (or so I was warned - I've never done it or seen it done).

My experience was that it took six layers of tape to make a water-tight seal (which might have been cheap screw threads that had big gaps to fill - optical instruments may get better quality taps and dies than plumbing supplies).

In this case a single layer of teflon tape is used on the threads of a Thor labs lens tube. Setting the lens position 'just right' is critical. The tape mitigated the backlash. Which makes finding and setting the correct position much easier.

George H.

Presumably the backlash represents situations where the metal-to-metal contact had welded or dug in, and you've had to apply a lot of torque to break the weld, which has then moved the two threads along for an appreciable distance before they dig in again.

Putting in a lubricating layer - so you don't get metal-to-metal contact or a risk of local welding - could be expected to minimise this.

A couple of layers of Teflon tape might work better than just one - obviously, if you try too put in too much you won't be able to get the threads to engage, but anything short of that should serve the purpose.

On Apr 12, 2019, Phil Hobbs wrote (in article):

As for back reflections from windows (and maybe lenses), there is a trick from photography to cause reflections to fall in harmless places. Basically, if one displaces object to one side the image will migrate the other way, and back reflections will no longer re-enter the lens. This is an application of the Scheimpflug principle:

Joe Gwinn

I often just aperture them to get rid of the astigmatism. In the lab I usually use a Mitutoyo microscope lens as a DL collimator. You can get fast-axis collimating lenses, but they aren't easy to use.

Cheers

Phil Hobbs

Sort of. In a laser collimator, light reflected back down the beam axis hits the laser exactly, regardless of minor lens misalignment.

It can help a bit with the first-surface reflection from the collimator, especially if the first surface is concave.

Problem is, in a LD collimator the coma builds up pretty fast as you go off axis.

Cheers

Phil Hobbs

snipped-for-privacy@ieee.org wrote in news: snipped-for-privacy@googlegroups.com:

query=

(5

The pipe ends teflon tape is typically used on have a thread placed on them which is slightly tapered on the first few threads.

Also with internal threaded pipes in the family. So the match up starts out very open and loose and works toward a full on interference fit.

Six layers is fine, because teflon has a 'cold flow' attribute, so the seal eventually would compress out literally hydraulically because of the closure of the tapers.

George Herold wrote in news:ecb652d8-2646-4c1a- snipped-for-privacy@googlegroups.com:

Much better term to use in this instance than was "hysteresis".