Driving Lasers

I doubt that $2 laser pointers have an optical feedback loop. Constant-current should be fine.

Do you need a super stable light level?

I suspect that constant electrical power does not result in constant optical output over temperature.

Try blasting a laser pointer (or an LED) with a heat gun and freeze spray. They are dramatically more efficient when cold.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Serious lasers tend of come with a built-in photo-diode. Lasers are double ended devices, and while the mirror at non-output end of the cavity is even more reflective than the one that lets a bit of light out to form your beam, there can still be enough leakage through it provide a feedback signal.

If the part itself is unhelpful, putting the beam through a Brewster angled surface - typically a micro-scope slide - can still give you enough reflected light for amplitude control without wrecking the beam.

One thing to keep in mid with your constant current source is that the lasing region is very compact, and even a very brief over-current can burn it out.

Am 13.12.25 um 05:47 schrieb Bill Sloman:

Yes, it is usually the frequency-selective optical reflector grid that dies in ms. Faster than we are used from semiconductors.

Gerhard

A nice juicy carpet shock can blow the front facet off much faster than that (although the fragments will still be moving for awhile afterwards.

Cheers

Phil Hobbs

No one seems to have come up with a viable solution, so I've been giving it some extra thought. I considered the use of a semi-silvered mirror at 45 degrees to the plane of the beam and sampling the light level off that reflection, but then realized the loss through the mirror would be unacceptable. The only other idea I can think of is to use a mirror just off the plane of the beam which can swivel around for a split second every few seconds to deflect the beam into the path of the diode. That seems cumbersome and clunky but it's the only thing I can come up with, not being a designer of any description.

You’re very likely to blow up more lasers trying to do constant power. Start with constant current, then wrap a constant power loop around it.

Cheers

Phil Hobbs

Fair enough, but it still doesn't get around the issue of interrupting the beam to sample its output strength. Optical power and electrical power consumed are not reliably correlated due to temperature changes in the laser diode.

I'm thinking that one would get a bettter optical output tempco by driving a laser diode from a negative resistance source. Or, at least, some proper impedance. It wouldn't take long to measure things.

(Measuring IS a bit more difficult than talking)

Or a current source twiddeled by a temp sensor.

But why does the OP need extremely stable optical output?

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

The microscope slide at the Brewster angle strikes me as likely to offer an acceptable solution.

Getting the angle of incidence to precisely the Brewster angle for the wavelength of your laser and the particular glass in you microscope slide may take a bit of fiddling. The laser beam will presumably be polarised anyway, and you will have rotate the microscope slide to get angle of incidences to line up with the plane of polarisation of the light coming out of the laser. I don't recall that any of the packages we played with had a clearly marked plane of polarisation, but it's while ago, and I didn't have any responsibility for the optics.

Bounce it off a bit of window glass like the rest of us. ;)

Five percent of the light is better than good enough for the job.

Cheers

Phil Hobbs

I could probably live without it, to be honest. I tend to over-think things massively. I could probably run those diodes fine at 10mA below their maximum constant current provided the current was reasonably stable. Hell, I've got dozens of the things; I'll blow a few up and find out empirically like everyone else does if they're honest. :)

The OP is Cursitor Doom. He doesn't seem to want to keep the output power particularly stable - it's more just a matter of monitoring it to keep in a more or less sensible range.

It would be interesting to take a few measurements, like required current to get constant optical output at a few temperatures, noting laser voltage.

You'd need a photodiode or something.

Spritz some freeze spray on a laser diode or an LED. It's dramatic.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

I think that was how it was done in

There also was a small optical attenuator towards the fiber output to isolate the lasing process from unwanted feedback.

The LD had to swim in the fiber connector. It took a lot of 3D EM modeling of the flexible Kapton(R) tape to keep nice eyes @ 10 GB/s.

Gerhard

Yes, I'm sure that will suffice in this instance.

Thanks to a previous thread and Phil's input there, I have several of those too.

Is it cool? ;-)

I've tested a bunch and some of the $15 ones are very good.

They have single-fiber WDMs under $15 too.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Am 13.12.25 um 22:44 schrieb john larkin:

Ours were XFP and > 20 years ago. There were only DFB lasers for us. Price goal was ~ $300.

The SFP group next door could use VCSELs for their wavelengths. That's a completely different game with the low currents. Resistors instead of bias tees, no bias regulation in software...

Cheers, Gerhard