How to under-drive a laser diode?

[SL4P]

Seems easy  - but I want to make sure of the right way...  I have no idea!

I'm replacing the IR laser diode in a piece of gear (808nm 40mW spec), but can't find an identical part easily.
If I drive a 100mW diode from the same source - I'm guessing it will push out 100mW ?

What is the recommended method to de-rate a higher power diode in this situation to emit the lower required output consistently and reliably?

As they are current driven - is it as easy as  adding serial resistance? or....

[NiHaoMike]

It can be done and in fact is often done in optical drives to calibrate the (writing) laser power to just the right level needed by the disk.

[edpalmer42]

I'm far from a laser diode expert, but I don't think it works that way.

The driver for the existing laser diode will be designed to provide a certain amount of current.  Regardless of what diode you install, the driver is going to put out that same current.  If you increase the power capability of a diode, the voltage drop doesn't change much - you just pump more current through it.  But the driver won't allow that, so your 100 mw laser is now reduced to a 40 mw laser.

But wait, there's more!

Laser diodes are finicky beasts.  In order to work at all, you have to drive them almost to the point of destruction.  So what's going to happen when you try to drive a 100 mw diode with a driver for a 40 mw diode?  Maybe nothing.  That is, maybe there won't be enough current to make the 100 mw diode wake up and do anything.

Ed

[SL4P]

I'm far from a laser diode expert, but I don't think it works that way.

The driver for the existing laser diode will be designed to provide a certain amount of current.  Regardless of what diode you install, the driver is going to put out that same current.  If you increase the power capability of a diode, the voltage drop doesn't change much - you just pump more current through it.  But the driver won't allow that, so your 100 mw laser is now reduced to a 40 mw laser.

But wait, there's more!

Laser diodes are finicky beasts.  In order to work at all, you have to drive them almost to the point of destruction.  So what's going to happen when you try to drive a 100 mw diode with a driver for a 40 mw diode?  Maybe nothing.  That is, maybe there won't be enough current to make the 100 mw diode wake up and do anything.

Ed

That's kinda what I was expecting -especially with a non-linear curve of output to drive...
Any ideas appreciated.

[ajb]

What's the application, and how close to 100mW do you need it to be?  Laser diodes really need to be driven with a constant current, even more so than LEDs.  If you have datasheets for both diodes, there should be a "slope efficiency" (optical power vs current) figure for each that allow you to at least get into the right ballpark.  For close power matching, you'll need a power meter.  If both diodes are of the same type, it's unlikely that the difference in efficiency will be substantial, so just straight replacing the one with the other (assuming will likely be pretty close.  The good news is, since you're looking at putting in a higher power diode, you should have some margin for error before you overdrive it to the point of damage.

Laser diodes are finicky beasts.  In order to work at all, you have to drive them almost to the point of destruction.

No, you can drive laser diodes well below their rated current and have them work just fine.  The trick is to get above the lasing threshold, which might be 5-30% of rated If.  Below that threshold, a laser diode is just a very expensive LED.  Some of the parameters may shift a bit between Ith and Ifmax, but for a static (non-modulated) application you should be able to deal with that.

I only deal with visible wavelengths, though, so take the above with a grain of salt--I'm not sure to what extent the common 808nm species might behave differently.  For best results, wait for LaserSteve to chime in.

[SL4P]

Thanks - a bit to think about and prompts me to build a simple test jig for after the fact.

[ajb]

Oh, and of course there's the obligatory precaution that even 40mW is more than sufficient to cause minor burns to the skin and permanent retinal damage, and 808nm is invisible, so special care should be taken to avoid injury.  808nm is well within the transmission spectrum of the gooey bits of the human eye, so it has no trouble reaching the retina.

[Kleinstein]

The specs of laser diodes have more parameters than just the maximum power. Other important parameters are the optical interface - the position of the chip inside the case and the degree of divergence. In an optical drive the laser needs to get focused very well, so the optical parameters must match, and very possibly alignment is needed.

When driven with a constant current the 100 mW diode will give about the same power as the 40 mW Laser, but some difference is possible. With a 100 mW diode there is likely some safety margin not to blow the diode due to to much power.

[Pjotr]

It is not that simple. Laser diodes have a threshold current to get the laser action. Below that current it is just a LED. But that threshold is not near destruction, otherwise it would be impossible to modulate a laser diode. On the other hand when you go over the maximum rated current they will be destroyed very quickly. So you need them to use well within their rated working current.

[LaserSteve]

Once your above lasing threshold, the diode output generally follows the current till you get close to the knee of the current curve, then increasing current decreases output power and then comes failure mode where the diode still conducts but you blow the mirror surface off the lasing face. In principal, you would just have to back off on the drive current to the newer, brighter, diode.

Generally if you have a constant current source with a trim pot for setting current, exchanging diodes is fairly easy.
Provided you start at a very low current setting and can monitor the current by some means.

We generally use a "Dummy Diode"  test rig consisting of a two ohm 2 watt precision resistor and three or more 1N4001 in series to simulate the diode drop. The number of 1N4001 used is based on the forward voltage of the new diode from its spec sheet.  This lets us calibrate the constant driver without endangering the laser diode.

What would catch most people is that diode cans are generally pressed into the brass collimator mount with a force fit. So you
have to get the old diode out without destroying the collimator mount. This usually can be done, but you will destroy the old diode in the process. You will also have to refocus the collimator.

Not to mention there are 9 different pinouts in common use, of which I generally see the same three over and over.

Adding a series resistor will not help you. The diodes run off a constant current, not a constant voltage source. They have a negative delta and will thus run away to destruction without current limiting.

All bets are off if the diode driver runs in constant light mode instead of constant current mode. This happens if there is a feedback photodiode mounted inside the diode can or off a beamsplitter down stream. At that point your into some serious reverse engineering.

Roithener Lasers in Germany usually can source the older diodes, inexpensively.  Its worth a Google.



Steve