Laser with low low-frequency noise?

[Peter]

Hi all,

I'm trying to find a laser source with very low close-to-DC noise
(<100 nW rms in a bandwidth between 0.1 - 100Hz for a 1 mW output).
Wavelength should be visible or NIR, and power ideally around 1 mW.

Noise cancellation in the detector is not an option since the devices
we're characterizing with the laser are sensitive to power
fluctuations on the scale of 100 nW.

We've looked into diode lasers and drivers with feedback from a
photodiode (integrated in LD package or external). However, it seems
almost impossible to find anything better than 0.1% RIN (and that
often neglects noise below 10 Hz).

If anyone knows of a source that satisfies these specs, or of a
feedback system that can stabilize power to <0.01% rms in the
0.1-100Hz band, please let me know. Any advice would be greatly
appreciated.

Peter

[Lostgallifreyan]

Post by Peter
Hi all,
I'm trying to find a laser source with very low close-to-DC noise
(<100 nW rms in a bandwidth between 0.1 - 100Hz for a 1 mW output).
Wavelength should be visible or NIR, and power ideally around 1 mW.
Noise cancellation in the detector is not an option since the devices
we're characterizing with the laser are sensitive to power
fluctuations on the scale of 100 nW.
We've looked into diode lasers and drivers with feedback from a
photodiode (integrated in LD package or external). However, it seems
almost impossible to find anything better than 0.1% RIN (and that
often neglects noise below 10 Hz).
If anyone knows of a source that satisfies these specs, or of a
feedback system that can stabilize power to <0.01% rms in the
0.1-100Hz band, please let me know. Any advice would be greatly
appreciated.
Peter

I might be looking at this the wrong way, so if this post wastes time and
effort, feel free to ignore it.

It seems that if you try to stabilise power, your control needs to be
responsive, so that itself might be a source of noise. So instead, if you can
stabilise current, and find a diode that has a nice linear region, or at
least one with no mode disturbance over the range of power you're after, and
little noise of its own, that might work. A simple driver based on an LM317
regulator, with the extra capacitors where possible (data sheet explains
those for voltage regulation, not all will apply in current reguklating
config so far as I know), you might get noise in the current down to
extremely low levels. Which leaves only the diode to worry about.

If you need to modulate the diode power, I don't know enough to advise. While
I built a fast analog proprtional modulator based on the original LM317
current scheme, I never looked at the idea from a perspective of noise
reduction. It might be very low at slow speeds if modified, but I have no
certainty about that. Sam's LaserFAQ includes a few variations on the LM317
current regulator to look at. Line regulation for the IC is very good. Load
regulation less so, but as a lased diode is a very steady load anyway that's
not likely to be a problem.

Thermal noise may be a problem, but a TEC with some stabilising control might
help.

[Lostgallifreyan]

Post by Peter
Hi all,
I'm trying to find a laser source with very low close-to-DC noise
(<100 nW rms in a bandwidth between 0.1 - 100Hz for a 1 mW output).
Wavelength should be visible or NIR, and power ideally around 1 mW.
Noise cancellation in the detector is not an option since the devices
we're characterizing with the laser are sensitive to power
fluctuations on the scale of 100 nW.
We've looked into diode lasers and drivers with feedback from a
photodiode (integrated in LD package or external). However, it seems
almost impossible to find anything better than 0.1% RIN (and that
often neglects noise below 10 Hz).
If anyone knows of a source that satisfies these specs, or of a
feedback system that can stabilize power to <0.01% rms in the
0.1-100Hz band, please let me know. Any advice would be greatly
appreciated.
Peter

Another though, not mentioned in my earlier post:
Diode laser cavities can be upset by retroreflection (even destroyed, if
enough light gets back). Even though you need only a milliwatt, it is worth
using a stronger diode with a large quiet range of output power, then
attenuating it. This can reduce the chance of disturbing the cavity by any
light sent back into it. I'm sure that attenators, like resistors, make noise
all their own, but likely the least of the problem.

This web page might be useful:
http://redlum.xohp.pagesperso-orange.fr/laser/modeanalysis.html
It is based on a search for diodes for use in holography, and stable modes
with low noise are a main focus of the effort. The diodes are often easily
found, and strong enough to leave 1 mW after a fairly strong attenuator.

Again, all my thoughts on this are based on the assumption that trying to
actively stabilise power might be sensing things that just cause noise as the
system tries to respond. If your first need is to make the power level
steady, silent, then you can set it to anything you like so long as it's
within a range that IS sufficiently quiet. And once you set it, it will stay
there. In short, maybe you don't need feedback at all, except to control
current, which is a problem far more widely solved....

[news.eclipse.co.uk]

Post by Peter
Hi all,
I'm trying to find a laser source with very low close-to-DC noise
(<100 nW rms in a bandwidth between 0.1 - 100Hz for a 1 mW output).
Wavelength should be visible or NIR, and power ideally around 1 mW.
Noise cancellation in the detector is not an option since the devices
we're characterizing with the laser are sensitive to power
fluctuations on the scale of 100 nW.
We've looked into diode lasers and drivers with feedback from a
photodiode (integrated in LD package or external). However, it seems
almost impossible to find anything better than 0.1% RIN (and that
often neglects noise below 10 Hz).
If anyone knows of a source that satisfies these specs, or of a
feedback system that can stabilize power to<0.01% rms in the
0.1-100Hz band, please let me know. Any advice would be greatly
appreciated.
Peter

hi Peter

i take it you've by now got something sorted for this.

Your sepcs are not all that hard to achive - 40dB or 1 part in 10^4 is a
pretty standard measurement, getting to 1 part in 10^5 takes care and
getting better requires some effort.

Additional points are that it if you want really low noise then it is
better to work away from DC - laser diode noise is much lower at 1MHz
than at DC - a decent single mode DFB can be better than -120dB per root
Hz. Single mode fibre laser significantly better still.

There are lots of low noise laser driver design which work well.

Some information on measurement you are making woudl be useful to
provide more detailed info.

cheers

David

[Julia Feng]

在 2012年1月6日星期五 UTC+8下午9:04:03，Peter写道：

Post by Peter
Hi all,
I'm trying to find a laser source with very low close-to-DC noise
(<100 nW rms in a bandwidth between 0.1 - 100Hz for a 1 mW output).
Wavelength should be visible or NIR, and power ideally around 1 mW.
Noise cancellation in the detector is not an option since the devices
we're characterizing with the laser are sensitive to power
fluctuations on the scale of 100 nW.
We've looked into diode lasers and drivers with feedback from a
photodiode (integrated in LD package or external). However, it seems
almost impossible to find anything better than 0.1% RIN (and that
often neglects noise below 10 Hz).
If anyone knows of a source that satisfies these specs, or of a
feedback system that can stabilize power to <0.01% rms in the
0.1-100Hz band, please let me know. Any advice would be greatly
appreciated.
Peter

Hello Peter,

All you need to do is just puchase a low noise laser, right? With reference our product, 532nm might be a ideal choice. Now the noise of amplitude is <0.5%, and your needed power 1mW, also low frequency modulation of 0.1-100Hz, I can ask our engineer if you still need to know now.

Hope this help for you.
Julia