1mW or less IR (980nm or so) laser diode source?

[Pippy]

Darned if I can find a source for low power near-IR laser diodes.

Their are cheap 30mW and above available but that is way way over the power output I require, and as far as I know lasers have quite a narrow range of output power at which they work properly (produce laser light) so it's not possible to just turn down the output power.

Anyone know a source of cheap near-IR 5.6mm laser diodes at all ?

Anywhere between 850nm and 1120nm will do. Anything shorter than about 850nm starts to become visible, which is not suitable for the project.

Collimating IR led's is not that easy, led's are not point sources, and I require a beam of around 20mm diameter over a range of 1 meter to over 20 meter. Beam expanding a laser diode to around 20mm is pretty easy.

[ElectroIrradiator]

Is pulsing a 30 mW diode with a PWM signal an option, or does it have to be a continuous beam? Could give you any required output power below maximum.

[NiHaoMike]

What about fiber optic lasers? You might be able to find some obsolete ones for cheap.

[ejeffrey]

Thorlabs sells a 2 mW VCSEL for $30 https://www.thorlabs.de/thorproduct.cfm?partnumber=VCSEL-850 and a 10 mw 904 laser diode for $25: https://www.thorlabs.de/thorproduct.cfm?partnumber=L904P010.

The VCSEL is not single spatial mode, while the 904 laser is, and they have a few other options -- those are just the first two I found.

If they are available at that price from thorlabs new, you can probably get similar things considerably cheaper elsewhere.

Don't underestimate how much power you need.  If you use a cheap collimation lens and an aperture to circularize the beam you can easily lose 75% of your light.  You get a lot more if you use AR coated lenses and anamorphic prism pairs to circularize the beam, but that is more expensive.  In any case, it is always easy to attenuate the light with a filter, it is harder if you don't have quite as much as you want.

Laser diodes can run over a fairly wide range of output power, although the noise and spectral performance will be worse at the bottom range of power.  I don't know how much you care about spectral purity or noise.  Higher power diodes will also have a larger threshold current, so a 30 mW diode operated at 3 mW will be less efficient than a 3 mW diode.  I don't know if you care about this.

So I think a 30 mW laser is probably overkill, but a 5-10 mW one should be perfectly reasonable for generating a 1 mW beam.  You can operate it at 25-50% of its rated power and have plenty of room to spare for collimation losses.  If you end up with too much power, you can still just put a filter in front.  Some polaroid material makes a simple and cheap attenuator.

[Psi]

You should be able to turn down the power to get 1mW without issue.
A 10mW diode should start to lase well before 1mW.
You will of course need a laser power meter to confirm output power.

Most 1mW laser pointers sold in USA are actually 5-15mW diodes with their power turned down to meet the local laws.

You'll probably need a driver with feedback though. Small diodes are much more sensitive to current/voltage fluctuations and die much easier. You can't just drive them with a constant current the way you can with 200mW diodes. Although if your running them at lower power than their rating it might be ok but the output power will likely wander a little.

I'm pretty sure that, even if you found a source of 1mW diodes, they would actually be 10mW-30mW diodes with just lower drive specs in the datasheet.

[Pippy]

Thanks guys for your replies.

The main reason for wanting a limited output power is for safety reasons, and particularly with IR (the light being invisible) it needs to be perfectly eye safe.

I will experiment with reducing the output power of a higher power device then, only I'll start with red laser diodes to make things safer when experimenting, even though I can see the IR using the IR sensitive camera I have, I like to be able to see what's going on with my own eyes when designing/developing.

Pulsing it not really a problem as it's being pulsed anyway due to superimposing a carrier onto it, a modulated carrier at that. So I could reduce the carrier duty cycle (it's a square wave at present) but if things did fail it could end up with the full 30mW/50mW being emitted, which is properly dangerous when it comes to lasers and/or collimated light.

Hence the reason for wanting a diode that was physically incapable of producing anything much more than a couple of mW's at best under fault conditions.

I've bought one of these 5mW to do some testing with ..

http://www.aixiz.com/store/product_info.php/cPath/67/products_id/365

Only to discover the diode is actually a 50mW one

[ptricks]

Some of the diodes will work when power is lowered, you need to look on the datasheet for the laser threshold current and that will tell you the lowest possible current you have to supply to keep the laser working.   A lot of those 5mW ones will stop working at under 40ma so it doesn't scale in a linear fashion. The diode I have would only be 1mW at 15ma and that is below the threshold so it wouldn't work.
Have you contacted Axiz directly ? They should be able to tell you what will work and were very helpful when I wanted a laser that produce a sharper line than the ones they sold on ebay and places.

One other option might be to use a filter on the output to lower the output.

If you haven't yet take a look at the sams laser faq, it is old but still has lots of good information.
http://www.repairfaq.org/sam/laserfaq.htm

[ptricks]

Have you contacted Axiz directly ? They should be able to tell you what will work and were very helpful when I wanted a laser that produce a sharper line than the ones they sold on ebay and places.

I did email them and ask if they have the 5mW diodes on their own but they just pointed me towards the 50mW diode they use. oh well.

From the sams faq I found information on using a Neutral Density Filter , seems that is what you need to reduce the power on a 5mW down to what you want.

[ElectroIrradiator]

It ought to be possible to make a simple, series current/voltage regulator, just for the laser diode. Put a reservoir cap on the regulator output to even out the current spikes, just large enough to not cause noticeable sag while the diode is 'on' during a brief part of the PWM signal.
The regulator gives you a point to do a safe shutdown if needed. For instance a simple diode/r/c watchdog may fail to detect the PWM signal across the diode for a certain brief period of time, causing automated shutdown.
Additionally, put a sufficiently large power resistor in the unregulated power rail to the voltage regulator. In normal operation, given the maximum 1/30th of max. laser current, the resistor should leave just enough voltage across the regulator to fulfill its dropout requirement plus a bit. 30 times more current compared to normal operation is then simply not going to happen, even should the watchdog fail.

For bonus points make the PWM signal in an external hardware oscillator, and mix it in an gate with the intended serial data signal from the MCU. This makes the PWM 'safety' signal independent of any software errors and MCU crashes/resets.

[ElectroIrradiator]

OK, no worries. Just thought I'd mention it, hope you didn't mind.

ND filter...? Hmm, that might break.

Curiously, when I search a bit, it seems that many of the pre-assembled IR laser modules can adjust the power level down to the 1 mW range...

[ejeffrey]

It wouldn't be too hard to make an 'intrinsically safe' driver that will never exceed a given threshold.  For instance, a voltage regulator with a zener diode and a current limiting resistor.  That should safely limit the current to a maximum value.  You can still do feedback from the power monitor for regulation, but even in a fault condition it will go to a safe maximum.  I wouldn't trust a diode rating for safety purposes anyway -- the power rating is just a level determined by the manufacturer to not damage the output facet.  There is no guarantee that it can't produce more power in the right/wrong situation.

[NiHaoMike]

You can supply it through a RC network sized to only supply the rated current for a low duty cycle. If you're using it for communication, a high peak power is of great advantage since it boosts your SNR.