NEW LOC Laser diode review.

[Hemlock_Mike]

Benm -- I've pushed both diodes which includes the one you have momentarely to 1 Amp to try killing it. It didn't act like a fuse but the power dropped to something like 26 mW but it was there for only a second. The power returned as normal.
I'm wondering if there is a strange facit which dumps or cuts back with energy density. It would be fast like a dye cell Q suitch.

HMike

 

[LASER66]

Diode manufacturers have done a lot of work on facet/ cavity absorption. I know some of SANYO's work, when they did red, as well as Opnext/ Hitachi. Indeed, the diodes will "fold over" instead of failing. I have some SANYO data around here somewhere. It was in one of their older catalogs. There may be some slight damage going on, but not COD. That is why they did the work, to protect the diode. Example: I went to a sales call to a large laser manufacturer. Myself, the local SANYO rep, as well as members of the Japan team were there. The customer wanted some higher powered 635nm diodes. This was a few yars ago, so we are talking about 40 to 50mW, let's say. Well the graphs for power extended to maybe 100mW, then "folded."
I can get the names of the technology, and I may even have the confidential documents I think. I looked on-line just now and did not see anything to link to at the moment.
Anyway, this is just an indication to LPF members, that they may have reached the 2X rated CW power point! I say, that if we sell a diode with date rated for XmW CW, then LPF will be running it at 2XmW CW!
LPF runs diodes in CW mode at the pulsed rated number.
Please remember, most customer demand the 10 to 50K hours of litime seen in the data sheets.
I will look for the data.
Will

 

[Benm]

Well... some more data!

I tried equal CW vs average pulsed power testing. Results are interesting in the sense that they further deny anything termal going on here (measured at 1 kHz pwm):

300 CW - 188
400@75% - 169
500@67% - 183
600@50% - 147
900@33% - 90

400 CW - 228
500@80% - 220
600@67% - 185
800@50% - 130
1A @40% - 95

I've operated the diode at several seconds on 900 and 1000 mA do dial in the current, it suffered no ill effect whatsoever. I've been running it at 400 mA for a couple of hours too, no degradation, no nothing.

I did however find some interesting readings when i looked at current and diode voltage on a scope, that might hold a clue to what is going on here. I have to take a closer look and post some scopeshots, but i'm starting to believe there IS something there, although you wouldnt notice it at CW operation.

 

[Benm]

As for the electrical readings.. i have attached some scope shots with this post. Lower trace is current, upper trace is voltage (in arbitrary divisions, just set to fit the scope nicely).

The last 2 screenshots show the current the driver put out at 200 kHz in to 5 ohms resistance, and 5 skottky diodes respectively. Its not picture perfect, but by no means as weird as the 200 kHz shots of the lpc 822 at 200 mA. This driver was designed with ~50 kHz modulation in mind, so the little bumps and dips are to be expected even into a dummy load.

I have some ideas about what is going on here, but i would like to have some of your ideas based on the scope shots before i go into it further - as i don't want to steer anyones direction of thinking.

Meanwhile, if anyone has suggestions for further experiments, post those too, as i'm running out

 

[Hemlock_Mike]

Benm -- As frequency increases, stray Cap and Inductive factors play a strange role. Good luck in the low RF range.
Meanwhile, I have my diode 2 running at a MOST stable output of 235 mW for 7 straight hours @ 400 mA. Never a flicker noted in 7 Hours.

I just light a match with it for those of you who care!!!

HMike

 

[Benm]

At least they are durable then, also in terms of runtime.

At 200 kHz i'm seeing some reactive loading indeed, could be the length of the wires and/or the capacity of the laser diode to blame... but the driver should be able to compensate for that if it was fast enough.

Maybe i'll try with a two-transistor fixed current source and modulate that, it should be a least an order of magnitude faster than the current opamp driver. Gain/bandwidth product for the opamp used is 4.5 MHz, so i'm not surprised it has problems with odd loads at a few 100 kHz.

 

[flogged]

Mike -

Thanks for posting these numbers. Looks like a decent diode, especially considering the price (!)

It appears 400-450mA will reliably produce 1/4 watt of power.

If I might suggest another metric for evaluating the diode? It's simply the most efficient point in the operation of the diode (mW per mA).

So:

Current: mA/mW (mW per mA)
100: 0.5
150: 0.61
200: 0.67
250: 0.72
300: 0.7
350: 0.64
400: 0.6
450: 0.53
500: 0.47
550: 0.39

This reveals that 250-300 mA of current is the most efficient point of operation, where the curve reaches max acceleration. This is likely where you'll have a happy diode with a long life. Beyond this efficiency drops and more power is wasted heating the die.

Another way to evaluate this is by looking at the quality of the light. At optimal (spec'd) power levels the light output will be very clean, saturated and 'speckly'.. beyond optimal power levels the linewidth tends to broaden, color becomes more 'milky' and not as speckly.

 

[Benm]

Looking at efficiency can be good, you end up with an optimum somewhat below the maximum possbile level, but its fair enough.

I have not noticed any difference in beam quality when ramping up and down the current though. When you defocus the beam to a large spot on the wall you can see some dark/light boundaries in the edges change a bit, but that is the case for all red diodes i've used so far.

Even at ridiculously high currents, where power is reducing, like 600 or 700 mA, there is nothing noticeable about the light produced, besides the odd thing that you get less light for more current there At that point the effect is so dramatic that you have to check the current meter to see if you're adjusting upwards or downwards - totally weird!

 

[flogged]

Well, I used to be able to evaluate old, < 25mW, red diodes visually.

However at the power levels we're talking about here it's likely not possible. Too bright to tell the difference.

Same things with the blue diodes-- too bright and the eye is not sensitive enough to those freqs..

Anyway, it's best to rely on 'hard' numbers from calibrated instruments.

{EDIT} Forgot to add - the point of peak efficiency is one empirical way to determine the max power where you'll still have a long lived diode. For those of us the want longevity.

 

[Benm]

The peak efficiency idea might prove empircally correct often enough, but there is no real basis for it. The drop off in efficiency is caused by factors unrelated to those that kill laser diodes.

For example, you see a very similar efficiency curve with power LED's, but the efficiency starts dropping long before the manufacturer-recommended normal or maximum current is reached.

In case of a laser diode there are the mirror faces that get destroyed from excess optical power, not the actual light emitting semiconductor itself. These LPC822 have some mechanism built in that prevents that entirely though - any ordinary LOC or LCC would be very dead if i tortured it like this diode. I suppose i could ramp up current as far as i like, until a bond wire goes, which would be a worthless experiment in this case since power is already on its way down before that happens.

 

[flogged]

I disagree. For a given species of diode the lower powered ones will live longer. On average of course. More power == more heat, and this alone will lower the life expectancy of the diode. It's not a linear relationship, sometimes a little more heat will cause the LD to decay exponentially more quickly.

Of course, there will always be freak diodes that may live a long life at high powers. Also, if you actively cool the LD to it's optimal drive temp (usually around 25-30C) this will generally give it a long life.

Most people here are not actively cooling their diodes, so more power almost always means a shorter lifespan.. whether or not this relevant for a pointer is another question.

Don't confuse a laser diode with a LED, different beasts.

 

[Benm]

I doubt these need active cooling - something like a maxsink woule be just fine.

These 822 diodes seem fine for fix-and-forget types of builds... they are not incredibly powerful, but i reckon many people would benefit from such sturdy duides hitting the market. Just feed them ~400 mA and you can forget about anything else.To me these apear to be laser diodes that are as easy to drive as leds, a very good thing for novice builders!

I'm sure you can get better effiecieny and output power from other diodes, if you are willing to take risks. But if these are no more expensive than comparable diodes, i see a very good deal here. Not so much as for uber-powerful reds, but more as abuse proof devices for those building their own drivers.

These prove to be no-brainers in that regard - just feed them 400 mA and forget about them

 

[flogged]

In the case of this particular diode I agree. Simply because the diode is so inexpensive I don't care if it doesn't last for more than a few hours. Even with a lifespan of just a few hours that's a long time if you only use it in short bursts.

I'm gonna order some of these, and I'll be putting at least 400mA through them. I want a quarter watt of red!

Has anyone tested these to see how long they last when driven at 400mA??
Lifetime test or whatnot.

{EDIT} Duh.. I see Hemlock_Mike test to 7 hours. Sold.

 

[Hemlock_Mike]

After 3 more hours today at 400 mA, LED :-( That's about 10 hours total to COD.
I can't believe that this diode went this way -- It was barely warm.

I'll load up the last diode and try it.

RIP --- HMike

 

[flogged]

Sigh. In that case I'll probably drive these diodes (when I get them) around 250-300mA.

Appreciate you testing and sacrificing!

 

[Benm]

Too bad to hear it went led on you.. i ran it for another 2 hours tonight at 400 mA, still going strong so far. In the heatsink i'm using it runs up to about 5 degrees c above ambient, and given the heat in europe these days that'd be 30-40 c all the time.

In any case, the benefits of going beyond 300 mA with these diodes seem very limited... perhaps thats a good point to aim for. As itis, these things deliver well over 200 mW easily, and can take a beating/mistake too. For that combination i'm positive in recommending these diodes for a wide user group. The dwindling power curve can be disappointing to users who want to push things to the max, but these diodes are excellent for first builds where you might make major mistakes in setting current and such.

I'm sorry to hear mike's went led eventually, as i thought these might prove to be reliable reds for projectors and such.