Speaking* of green laser diodes.

Most short-wavelength diode lasers have multiple longitudinal modes. That's not necessarily a bad thing--it makes them much less sensitive to mode hopping due to backreflections. You sure don't want to use them for tunable-diode spectroscopy, though.

Cheers

Phil Hobbs

If you do want a single mode output you do have to stabilise the diode temp erature - the temperature inside the diode where the gain medium is, thoug h since you can't get at it, it in practice you do tend to stabilise the ex ternal temperature around the diode as well as the power dissipation in the diode junction.

Hmm OK, I don't have any experience with multi mode lasers. (well some HeNe's) So the coherence length is going to stink. A tunable diode laser thing won't work because there are too many nearby modes? I need a good diode laser book, or review article.

George H.

mperature - the temperature inside the diode where the gain medium is, tho ugh since you can't get at it, it in practice you do tend to stabilise the external temperature around the diode as well as the power dissipation in t he diode junction.

The forward voltage of the diode is 6-7 volts! At currents of ~>100 mA... there's going to be a bit of heat to get rid of. I'm thinking a TEC will be required.

George H.

to

temperature - the temperature inside the diode where the gain medium is, t hough since you can't get at it, it in practice you do tend to stabilise th e external temperature around the diode as well as the power dissipation in the diode junction.

My 1996 millidegree controller - with TEC - got adapted to keep a laser dio de at constant temperature to allow it to stay in a single mode, pointing i n the same direction.

The application didn't need a stable wavelength, but it did need the laser beam to point in exactly the same direction, and mode-hopping shifted the b eam direction around. Not all that much, but enough to matter.

Even back then, the literature was full of TEC-based laser-diode temperatur e stabilising circuits. Since then there have been a couple of integrated c ircuits that handle most of the job, not all that well, but well enough for a lot of applications.

Yup, there are probably a lot of modes. What's more, there will be a lot of mode-partition noise--generally much more than the total intensity noise of the laser. The reason is interesting: the pumping rate constrains the total output power, but not the way it's partitioned between the modes (hence the name).

The modes are strongly coupled, and power moves back and forth between them. Thus etalon fringes and that sort of thing will demodulate the mode partition noise into huge amounts of RIN, even though the undisturbed beam is actually pretty quiet-looking.

Cheers

Phil Hobbs

Are there simple optical accessories that can quiet the situation?

Say Phil, did you mean multiple transverse modes in the above? It's just because I'm having a hard time understanding how one can get lasing in different longitudinal modes. (My simple model is that the longitudinal mode with the highest gain 'wins' and sucks all the gain out of the other modes... The single mode DL's I use have many longitudinal modes.)

Whereas if there were different transverse modes, then those would occupy different areas in the gain medium, and I can then picture the thing lasing in different modes, (using different portions of the gain medium, and not 'sucking' the gain away from each other... Though it does get a little complicated looking... and I can imagine that the modes interact with each other some, as you said in your later post.

George H.

An attenuator will give less noise... and less signal :^) Otherwise I think you are stuck. I think they put apertures inside the laser cavity to suppress some of the transverse modes. From here,

Great! You have to buy 200 pieces, minimum!

It is a crystal doubler (or maybe tripler) on an IR diode.

Jon

Avnet Europe has 70 of the 30mW ones in stock for 24 EUR each, but they want 60 EUR for shipping.

--Spehro Pefhany

First of all, if you want some to play with, Google "DTR's Laser Shop" and don't tell me how your stunned to see 1 Watt Green diodes..

Next up, Littman Metcalf cavities using a grazing incidence grating configuration for feedback works fine up to modest currents.

In the newer multimode, high power blues and greens, it is not unusual to see 3.5 nm FHWM spectrums unless there is feedback.

Even the little Osram single modes tune a few nanometers. One of my friends likes dunking the modern single mode diodes in LN2 to get drastic wavelength shifts while running at low currents.

Steve

Well they don't have stock. You can probably buy direct from Osram.. (200 min. :^)

No, I think this is a 'direct' 'bandgap', maybe a riff on blue laser diodes? (With 7 V forward there is more going on than just a band gap.)

George H.

Not really. You could select one mode with an etalon and apply amplitude feedback, but the fading can be pretty deep so I wouldn't expect that to work well. The best method to get a quiet green laser is to use a doubled diode-pumped YAG.

Cheers

Phil Hobbs

It's multiple longitudinal modes, so apertures don't help. Modern diodes are all index-guided, so transverse modes are pretty well controlled. You do get spontaneous emission in higher transverse modes, which interfere with the main beam and causes "wiggle noise".

Cheers

Phil Hobbs

Nope, longitudinal. In a Fabry-Perot laser, there are nodes and antinodes of the standing wave, leading to spatial hole burning, where the carriers are locally depleted near the antinodes. Different modes have different antinode positions, but they do overlap quite a lot, which makes the longitudinal modes strongly coupled.

The rate equations are second order--the mode amplitude goes like the time integral of the population inversion, which goes like the time integral of the pumping rate minus the total amplitude. So depending on the coefficients, it can oscillate. Some lasers, such as N2, have no stable operating points--they always produce nanosecond pulses.

Getting rid of this problem is one reason for liking VCSELs. (Fast modulation is the other reason.)

Cheers

Phil Hobbs

Nope. I'm using their 488 nm version, and it's just a wide bandgap diode. Infrared diodes don't drop 6-9V.

Cheers

Phil Hobbs

OK, well, this is new(er) technology. Always something new being created in opto-electronics!

Jon

OK, Thanks. Lets see if the cavity was ~1mm (1000um) I'd get ~2000 wavelengths in the thing and the longitudinal modes are separated by 500nm/2000 ~0.25 nm (plus or minus a factor of 2) Can I see the modes with a spectrometer? That would be fun.

George H.