Are diodes killed by mW or mA?

[laser_freak]

We've all killed at least one diode. I know that when you power it with way to much current, the little wires connecting the pins to the actual laser diode chip will act as a fuse. However, when they don't act as a fuse (and it isn't the fault of ESD), is it the optical power or the electrical power that kills the diodes? Maybe a combination of the two?

I've heard that one limitation holding back the power of LDs is self degradation caused by optical power.

It's probably not a straight forward answer, but I want to hear your opinions.

 

[aryntha]

Well, the optical power is a function of the electrical input minus efficiency. In the long run, on either a micro level or macro level -- diodes are killed by heat. That heat can cause degradation of the lasing surfaces (the PN junction) or the die wires.

There may be other "failure modes" that don't involve any heat at all ( maybe someone else can explain exactly what happens internally when a diode 'goes LED', I presume that the laser surfaces get damaged and while the PN junction emits light, no lasing happens) ... but mW is more of a "result" than a factor that will cause direct damage. (Unless the beam is reflected back into the diode... resulting in heat once again.)

 

[qumefox]

Generally they die from excessive optical power unless you burn off the bonding wires. The coatings that make up the facet mirrors can only take so much power before they burn off. When a LD goes 'LED', what has happened is that one or both facets have been destroyed, either by over current, or a nanosecond burst from ESD. the PN junction still emits the same amount of light, however, since there is no longer a working optical cavity, you no longer have a laser.

 

[hakzaw1]

I caused an early death to one 445 by the Law Gravity- only a foot or so --from the couch onto the carpet but sometimes that is all it takes..
also a few years back having no knowledge of duty/rest cycling I killed my first greenie in one night of overuse. Ouch!! 40$ back then too.

 

[qumefox]

Strange. I wouldn't have thought they would be that sensitive to shock as to be hurt by landing on something soft like carpet. Was it a plain diode, or mounted in a host?

 

[aryntha]

With the fall to the carpet, it could have been ESD, rather than the drop.

 

[millirad]

Interesting discussion. I had not realized that the LD had more optics than those provided by the aixis module. So they have a set of thin film cavity mirrors called facets, to provide coherence. Great info.

 

[qumefox]

Well, they aren't really separate optics. The ends of the diode material itself are polished facets with the HR and OC coatings. As stated before. Without a working cavity, they're just LED's.

This is oversimplified though. There are other differences than just having an optical cavity that separates LED's from LD's. If that was the only difference, we'd already have LD's in almost every wavelength.

 

[ossumguywill]

LDs don't really have optics (besides the window), the only optics they have are necessary for operation as a laser (that would be a mirror and a semi-mirrored surface as well as side optics).

 

[pullbangdead]

Yes.

Lasers can be killed by the light, by the current, by heat, any of it.

Excessive current while being held at a constant low temperature still kills. Excessive heat kills. Too much light kills regardless of heat and current, as well. They kill in different ways with some different symptoms, but all very effective.

 

[lasersbee]

We've all killed at least one diode. I know that when you power it with way to much current, the little wires connecting the pins to the actual laser diode chip will act as a fuse. However, when they don't act as a fuse (and it isn't the fault of ESD), is it the optical power or the electrical power that kills the diodes? Maybe a combination of the two?

I've heard that one limitation holding back the power of LDs is self degradation caused by optical power.

It's probably not a straight forward answer, but I want to hear your opinions.

All the above explanations are correct...IMO...

The little wires that come out of a Laser Diode are NOT fuses
and can NOT act as fuses...
To blow one of those pins in an instant would require tens of amps
if not more... and the Laser Diode PN junction would be damaged
much faster than the wires could melt and open...

Fuses in general are very slow safety devices...

Jerry

 

[laser_freak]

All the above explanations are correct...IMO...

The little wires that come out of a Laser Diode are NOT fuses
and can NOT act as fuses...
To blow one of those pins in an instant would require tens of amps
if not more... and the Laser Diode PN junction would be damaged
much faster than the wires could melt and open...

Fuses in general are very slow safety devices...

Jerry

I think it was either daguin or Hemlock_Mike who had said that in PHRs the little wires connecting the laser die to the pins would act as a fuse when driven at too high of a current. I can't find the original post, but I'm sure it was a reputable member who said so. Maybe I'm remembering the post out of context and the diodes were being driven at much higher currents.

 

[lasersbee]

My mistake... I understood the pins external to the Laser Diode...

Yes the thin wires connecting the Laser Die on the inside of the
Laser Diode can blow by putting too much current through them...
but that is just another way of causing a LD defect.
By the time the internal wires do melt... the Die will have had time
to deteriorate or damage..

If the current to a Laser Diode is kept to spec.. the Laser Die
connecting wire should not blow...


Jerry

 

[qumefox]

Yes.

Lasers can be killed by the light, by the current, by heat, any of it.

Excessive current while being held at a constant low temperature still kills. Excessive heat kills. Too much light kills regardless of heat and current, as well. They kill in different ways with some different symptoms, but all very effective.

Excessive current in itself doesn't kill LD's unless you burn out the bonding wires. When run at excessive current, you have to remember your putting out more light than the coatings on the facets can handle at the same time. This light output is what ultimately kills the diode when it's cooled adequately.

Like I said above, there are only two true killers. Too much heat (PN junction and possibly facet damage) and too much light (Leads to COD on the facets). Running at excessive current just happens to be a trigger of both, though in most applications they're cooled well enough that the latter happens first.

But i'll quit nitpicking now. :na:

 

[pullbangdead]

Excessive current in itself doesn't kill LD's unless you burn out the bonding wires. When run at excessive current, you have to remember your putting out more light than the coatings on the facets can handle at the same time. This light output is what ultimately kills the diode when it's cooled adequately.

Like I said above, there are only two true killers. Too much heat (PN junction and possibly facet damage) and too much light (Leads to COD on the facets). Running at excessive current just happens to be a trigger of both, though in most applications they're cooled well enough that the latter happens first.

But i'll quit nitpicking now. :na:

False.

Current can degrade diodes independently of light output and temperature. It may not be as commonly seen around here, but it definitely happens. Experiments that control for light and heat have definitely been done, showing that current itself can degrade devices. This is true not just in lasers, but in all electronic devices and ICs.

It's especially evident at high current densities, like those that exist in laser diodes.

But I suppose it could be argued that the difference is immaterial here: the source if the heat is the current, whether it's the heat or the current itself that is actually doing the degrading.

 

[seoguy]

I know that when you power it with way to much current, the little wires connecting the pins to the actual laser diode chip will act as a fuse. However, when they don't act as a fuse (and it isn't the fault of ESD), is it the optical power or the electrical power that kills the diodes?

In the case you describe, I think it's more typically excessive optical power, resulting in COD to the output facet.

The speed of light inside the LD is slower than in the atmosphere, and the output facet thus also has to act as the optical equivalent of an RF impedance-matching device. But because this match is far from perfect, a portion of the optical power attempting to exit the LD is instead transformed into waste heat, which is deposited into the output facet.

LD manufacturers use special coatings and other "tricks" to try to make the change in refractive index less abrupt, and to improve the efficiency of this interface. But the match is still imperfect, resulting in part of your optical power being stopped at the output facet, potentially causing damage.

As a result, while excess electrical power can also damage the device, the output facet is often the "Achilles heel" of the LD, in terms of max output power it can produce without significant damage.

A great example of this is the new red LD's that some members have been experimenting with. The are designed with a built-in protection against excessive optical power - above a certain point, the optical power actually goes DOWN with increased current - with all of that excess energy being absorbed electrically by the LD!

To clarify, above a danger threshold, the LD is actively trading increased electrical stress for a reduction in optical output power, presumably to protect against COD.

Indeed, even though optically these are less powerful than an LPC-815, they have been amazingly pushed to as high as 1.1 AMPS of electrical current without destruction!

Clearly (in this case at least), optical damage rather than electrical is the "weak link" in the chain, otherwise this strategy would make no sense!

Unfortunately, this fail-safe design approach also puts the equivalent of a "speed limiter" on the diode, making such LD's far less useful to a hobby that relishes pushing LDs to WELL past their manufacturer rated power levels! :cryyy: