Laser diode reliability and/or failures.

[Marco Polo]

When I think of a "common" single mode laser diode I think of something like the Opnext HL63603TG, the Osram PL450 and PL520, and 405nm blu-ray diodes. I've asked around a bit and from all appearances, it seems like diode failures due solely to being overdriven aren't particularly commonplace. The general consensus seems to be that the leading killer of this type of diode is that something went wrong with the build. The laser diodes themselves seem to be fairly durable and reliable even when run out-of-spec.

I'm curious because I've seen it stated it a couple of threads lately that "driving the diodes this hard shortens their life spans." While I don't doubt that, how much does it shorten their life spans?

I have the previously mentioned red, green, and blue lasers and I haven't had a single problem with them. I fool around with them all the time, so they have definitely accumulated some hours. They are all running at 300mA at the diode. Only the blue gets warm, and even so, it takes a while. It pulls ~800mA at the tailcap.

Has anyone had diode failures that were most likely due to being driven beyond rated specs, rather than due to some sort of misadventure (e.g. heat, ESD, general user idiocy, etc.) ?

The only laser diodes that seem prone to dying are 405nm's. How are those 700mA BDR-209's holding up?

 

[Grix]

Diodes have become much more robust lately. On some 445nm diodes I have accidentally touched both leads with my bare, ungrounded hands at the same time multiple times and they still work fine, that was unheard of a few years ago. Also, diodes that are pretty much impossible to instakill by overcurrent have become pretty popular. On these, instead of having a linear piv graph, the output power growth slows down and actually decreases after a certain point if you keep increasing the current, and works fine if you turn the current back down. Earlier, every diode would just suddenly die if you increased the current too much.

That being said, diodes still do die sometimes when in use, and it's reasonable to attribute these deaths to overdriving. After all, nearly every diode we use are overdriven by like 100% over spec sheet settings, you can't expect them to last forever under those conditions. It's really impossible to say exactly how much life expectancy is decreased with current increases, as it depends on heatsinking, invidual diode variation, driver quality, duty cycle etc.

 

[chipdouglas]

haha I think about this too. Glad you brought it up. All of my deaths have been caused by my poor building. Once my lasers are up and running, I've never had one die.

When I think of a "common" single mode laser diode I think of something like the Opnext HL63603TG, the Osram PL450 and PL520, and 405nm blu-ray diodes. I've asked around a bit and from all appearances, it seems like diode failures due solely to being overdriven aren't particularly commonplace. The general consensus seems to be that the leading killer of this type of diode is that something went wrong with the build. The laser diodes themselves seem to be fairly durable and reliable even when run out-of-spec.

I'm curious because I've seen it stated it a couple of threads lately that "driving the diodes this hard shortens their life spans." While I don't doubt that, how much does it shorten their life spans?

I have the previously mentioned red, green, and blue lasers and I haven't had a single problem with them. I fool around with them all the time, so they have definitely accumulated some hours. They are all running at 300mA at the diode. Only the blue gets warm, and even so, it takes a while. It pulls ~800mA at the tailcap.

Has anyone had diode failures that were most likely due to being driven beyond rated specs, rather than due to some sort of misadventure (e.g. heat, ESD, general user idiocy, etc.) ?

The only laser diodes that seem prone to dying are 405nm's. How are those 700mA BDR-209's holding up?

 

[ultimatekaiser]

the average diode is usually rated to last about 5000-10000 hours if run at spec... each diode would be different however, and their average lifetime is surely a function of current being fed through it, even with proper heat sinking. but by how much would likely vary diode model to model....and even a bit unit to unit....hene's and argons are the same way, even sometimes as much as by a factor of 10.

 

[NKO29]

I've only ever had one diode randomly die... Never broken one myself thankfully I never properly ground myself, work on carpet and ESD has never damaged anything i've worked on (including my $700 (at the time) gtx 780s). User error seems to be the death of diodes, which includes letting them get too hot.

 

[The Lightning Stalker]

I had 2 diodes extracted from the same drive that came out as LEDs the other day. At first I
was mystified, but then I remembered that the ESD strap was accidentally unplugged during
that extraction. The first diode I ever extracted was also the first one that went LED. It was a
SmOC and I didn't use the strap at all. That learned me real quick. :spank:

 

[Pi R Squared]

I agree that its mostly heat, ESD, and general user idiocy that causes most failures. I suppose some testing would be possible but there are 15 to 20 different diodes we commonly use, and I don't think testing one of each would mean anything. There are just too many variables involved.

Alan

 

[Marco Polo]

Well, my PL450 in the C6 host went LED tonight. The diode current was 300mA. It's the only one of my C6 lasers that put off any noticeable amount of heat, but it never went beyond getting lukewarm. Even had thermal compound between module and heatsink.

I'm not sure if I should try another PL450 at 300mA, or drop the current to 250mA. Could it have been a weak diode?

I have the PL520 running at 300mA as well. It doesn't seem to get warm so hopefully it lasts longer.

If you don't mind, consider posting in this thread whenever you have a diode fail during use. Mention what diode was involved, the drive current, maybe what you were doing with it at the time. With enough failure reports we can get a better idea of how long a given diode lasts at various current levels, which would be good information to have when planning future builds.

 

[The Lightning Stalker]

Reading some older threads, I have found numerous posts about seriously overdriven diodes dying after
20min-1h. The next time I run across one, I'll link to it here.

 

[Pi R Squared]

I have yet to have one fail except a cheap 532nm pen I bought on Amazon. My builds are still working. 300ma is high for a PL450 isn't it? I would try it at 250ma. My PL520 is running at 350ma and has a considerable amount of time on it already, however it has such a large Mrcrouse finned heat sink that I can leave it on 5 minutes and there is barely a trace of heat so that would make a difference. I am now more in favor of using large heat sinks. Two of my lasers are too powerful for the host/heat sinks they are in (links in my sig). If I have a failure I'll try to remember to post it here.

Alan

 

[Atomicrox]

I've owned lasers for around 4 years now and I have had:
-Several mechanical failures (mostly on pen hosts)
-532nm degradations/mode hoppings/instability (due to DPSS)
-Power losses and artifacts due to dirty lenses
-One laser which lost around half the power but still works fine (possibly due to a battery reversal that damaged the driver)
-One 405nm that came LEDed and was replaced by the builder

As far as I remember the only one that could be attributable to spontaneous diode failure is the 405nm.

I'd also like to have more people post their failures here.

 

[Sigurthr]

While those in the beginner DIY or cheap-Chinese-mass-produced sections of our hobby mostly experience thermal, mechanical shock, ESD, or reverse bias failures, you seem to be unaware of a very pertinent failure mode for diode lasers; Catastrophic Optical Damage.

Catastrophic optical damage - Wikipedia, the free encyclopedia

C.O.D. occurs when the mirror coatings or lasing medium absorbs too much light from exceeding the intended power density. Simply put you could chill your diode with LN2 and still kill it by pushing it too hard simply from optical damage.

 

[bostjan]

I've had one 532 nm 5 mW pen fail on me, after a few minutes of use, but I believed it to be a driver issue, since the lasing action still occurs for a second or two before it dims and eventually fades completely out.

I've had a 405 nm pen mechanically fall completely apart, but the module still works just fine, after quite a bit of use and abuse.

I've never had a diode fail, but I'm still quite new to the world of high power laser diodes.

 

[IsaacT]

I have never had a diode die from being overdriven.

My reasons they die:

• When I first started the hobby I let the diodes get too hot because I had a really big soldering iron that was too hot.
• Then there were a few that died of Blunt Force Trauma. One I dropped from too high up. One got crushed by a bad diode press when removing it from the AixiZ housing.
• MAYBE one died from ESD.
• I have had one die from Reverse Polarity.
• One died from a momentary disconnect and reconnect where there was a jump in current as it reconnected.
• And one died from overheating.
• Oh and one died when I accidentally sheared off the diode pin as my hand slipped.

 

[Cyparagon]

Even if you decrease the life by 99.5%, it isn't noticeable when you use it as a pointer. 25-50 hours is a lot for that application. By the time the average user is up to 5 hours of total use, a new diode becomes available.

 

[The Lightning Stalker]

I was testing another 784nm IR diode again and disconnected and reconnected it while the power was still
connected (shorted the leads though, lol) and it LIVED! Guess I was just lucky I guess. There was no
decrease in power output or anything. I was very surprised. It was a crappy little ROHM too.