1. Home
  2. Electronics Design
  3. Strong HF-Modulation of a Laser Diode to prevent Mod e hopping in a Mach–Zehnder interferometer (path length di fference 25cm)

Strong HF-Modulation of a Laser Diode to prevent Mod e hopping in a Mach–Zehnder interferometer (path length di fference 25cm)

Hello!

I want to use a Laser Diode in a Mach=96Zehnder interferometer where one path length is modulated by a moving Mirror. The resulting interference signal is sinus shape like, narrowband and has a frequency of around 1kHz.

When I exchange the HeNe-Laser and use a green DPSS-Laser then I get a good sinus shape like, narrowband signal. But the DPSS-Laser needs around 10 minutes warm-up time for a stable Laser-Beam-Direction. During the warm-up time the beam moves a little bit.

I use a collimator connected to a single mode fibre as detector and so a directional movement of the Laser-beam results in an extreme change in fibre coupling efficiency. So I want to change to a Laser Diode because it=92s smaller then a NeNe- Laser and produces less heat and Diode should have no directional movement of the Laser-beam during warm-up.

I have tested some Laser-Modules:

My 850nm Laser-Module (single mode diode 5mW) produce a very stable sinus shape like, narrowband Signal with no viewable spikes or noise by changing the supply current from threshold to maximum current but is invisible an difficult to adjust.

My 660nm Laser-Module (single mode diode 75mW) produces a very unstable non sinus Signal that becomes better at some very small current ranges. (I expect it comes from Mode-hopping.)

In =93Building Electro-Optical Systems:Making It All Work, Second Edition=94 by Phil Hobbs he writes in Chapter 2.14.3 that a strong modulation of the Laser-Diode-Current in the Frequency range of

300-500MHz suppresses mode-hopping but decreases the coherence length.

The Mach=96Zehnder interferometer that I use has optical path length difference of around 25cm.

If I take a single mode Laser diode I can expect a spectral line width of around 100MHz. This should result in a coherence length of around

1.5meter. This should be enough for my Interferometer.

If I modulate the Laser diode with 300MHz I should get a coherence length of around 18cm. Is this correct or does I make a calculation error? I ask because Phil Hobbs=92s write that the coherence length decreases extreme.

For the modulation I wanted to use EL6203 or EL6204 from Intersil but it is impossible to get it in Europe.

I need a Laser that has no beam-movement during warm up and has little intensity-noise and almost no mode hopping. The wavelength is roughly stabilised by stabilising the case temperature. It must have a coherence length of more than 25cm.

Thanks for your help and thanks to Phil Hobbs for his great book. Harald Noack

Reply to
Harald Noack
Loading thread data ...

This is a bit off-topic for sci.electronics.design, so I'm cross-posting to sci.optics and setting followups there.

The effects of UHF modulation on lasers depends on how strong it is. What I was talking about in the book is very strong modulation, essentially turning the laser on and off at 300 MHz. That forces the laser to start up from noise on every cycle, so that there's no phase coherence between adjacent pulses. That's what gives the mode-hopping immunity. However, when you turn it on, you'll excite longitudinal side modes of the laser, as well as putting a gigantic chirp on the main mode.

Weaker modulation can be pretty useful too, but requires a bit of tuning. The idea is to put the first Bessel null of the modulation at the distance of the strongest back-reflector. (I did that in the ISICL sensor, and it worked fine but needed computer-controlled tweaking. I have a note-to-self to write up that trick for the third edition, but haven't done it yet.)

The sweet spot for diode laser performance is 750-850 nm. Maybe you could use a webcam to do your beam aiming? That would be my first choice.

Second choice would be to use the visible diode plus a polarizing beam splitter and quarter-wave plate. That'll get you about 40 dB of isolation, which may be enough, assuming it's a long-path reflection that's responsible. VLDs aren't very stable beasts to begin with, and can easily mode hop due to back-reflection from the collimating lens.

Third would be to write a bit of code to explore the (temperature, bias current) space to find islands of stability, and work there. That may be your only option with that VLD if it's limited by feedback internal to the collimator package.

The other thing is that an interferometer with a 25 cm path difference is a pretty decent thermometer and barometer, and large Mach-Zehnders are also vulnerable to air currents and microphonics, but you know that already since you've built it.

Usually 750-850 nm plus a webcam is by far the best-performing approach.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510
845-480-2058

hobbs at electrooptical dot net
http://electrooptical.net
Reply to
Phil Hobbs

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.