Some questions about fully build laser vs ESD

[BlackOps]

hello, i successfully completed an pocket laser burning pen. It had an SLD3232VF 405nm 50mW 5.6mm laser diode. Case negative.

And here is what happened. (Yea its pretty dumb) so i had an piezo electric device extracted from an lighter. And an fully builded laser burner (aluminum host). Now, i could throw this laser of an airplane and it would still work, thats how i designed and build it. How is it not able to withstand an shock from an piezo electric device?

So i was playing with the piezo until i saw my laser on my desk and i tough why not (diode was $2 anyways). And then it survived the 1st shock at the tail cap, (laser was ON) the second one, on the focusing knob, it survived as expected (ground) now, i tough LOL, so i tried where the electronics are (BMS micro controller, dual gate transistor and the boost driver + LDO, neat stuff) and the diode went crazy stuttering and then it turned into a nice led like 1 month ago. It would have being nice if it at least would have stuttered but no, it went to ****.

But here is the main question, how could this happen? It was against a fully build, watertight, probably military grade laser, and it died? I did this piezo thing on my iPod touch 4th gen (every corner) back in 2013 to test it, and it still works today. Even on my Xbox 360. How could this piece of **** not survive this.

The driver is working perfectly fine, the micro controller used for the safety feature works good and cuts the power off at exactly 2.80v. Driver has been tested after this disaster.
driver has been fully tested, no voltage output shift, no heat, current regulation works. Switching works absolutely fine.

Now i want to know, how an laser diode couldn't survive this, while every other semiconductor used in this laser was undamaged, note, i shocked it at the area where the electronics are, and the electronics are builded into the (entire) aixiz housing, i used every part from the aixiz housing, so the electronics where isolated within 2 tubes. Those survived.

I ordered new diodes (this time a couple spares) so i could rebuild this master. But i want this to be cleared out so i could solve that issue and the diode replacement process is 15 min anyways. Any explanation is highly appreciated.

 

[diachi]

Laser diodes are very sensitive to ESD - more so than a lot of other semiconductors.

 

[BlackOps]

Laser diodes are very sensitive to ESD - more so than a lot of other semiconductors.

I understand, but even when its builded into an aluminum housing? The semiconductor was isolated. The only thing connected to the diode, from the outer case, is common ground. Does it even die from ground?

 

[Cyparagon]

Are you saying the diode died a month after you zapped it?

the diode went crazy stuttering...
It would have being nice if it at least would have stuttered but no...

clear as birefringent calcite.

 

[BlackOps]

Are you saying the diode died a month after you zapped it?

clear as birefringent calcite.

Hi, no lol. A month ago, i destroyed an diode, now, its in the exact same condition again. As i zapped the diode {somehow}. Thats what i mean. But do you have an theory for why it died while isolated in a host? (Laser was case negative by the way).

Or is it just that laser diodes are the most sensitive semiconductors out here? I mean i can't destroy an advanced computer processor in an iPod like this. Still works after 3 years. It would be awesome if this could be cleared out as i might be able to patch this issue. Thanks anyways

 

[diachi]

Are you saying the diode died a month after you zapped it?






clear as birefringent calcite.

Yeah ... not the most intelligible post ever... Put it down to the language barrier!

Guess it could still be ESD damage even if it did die a month later. Just wasn't enough to initially kill the diode, instead causing it to degrade more rapidly over time.

Don't see why you'd zap sensitive electronics for shits and giggles though ... Unless you don't care about potentially destroying them.

Or is it just that laser diodes are the most sensitive semiconductors out here? I mean i can't destroy an advanced computer processor in an iPod like this. Still works after 3 years. It would be awesome if this could be cleared out as i might be able to patch this issue. Thanks anyways

For starters ... the processor in the iPod is most likely isolated from the case - your diode is not. May also be less sensitive to ESD in the first place.

Best way to "patch the issue" is not to zap your laser with a Piezo...

 

[BlackOps]

Yeah ... not the most intelligible post ever... Put it down to the language barrier!

Guess it could still be ESD damage even if it did die a month later. Just wasn't enough to initially kill the diode, instead causing it to degrade more rapidly over time.

Don't see why you'd zap sensitive electronics for shits and giggles though ... Unless you don't care about potentially destroying them.



For starters ... the processor in the iPod is most likely isolated from the case - your diode is not. May also be less sensitive to ESD.

Best way to "patch the issue" is not to zap your laser with a Piezo...

Hi, thanks for your reply. Yea, My primary language is Dutch. I will edit that section.

A processor in an iPod is connected to more then 20 ground planes and those are connecting with screws and therefore touching the case. I like to experiment. I did this piezo electric experiment so many times. This is the first time something died.

Yea i know, its stupid but i at least learned something and i have to say, its worth it. The diode died nothing, else so its 10 min work (cleaning the diode window is gonna take me 45 mins) thats the only thing i'm mad about. Cleaning the optics.

The diode i had a month ago died as i cleaned the window while it was on and therefore destroying the facet. I left it on so i could see whether there was something on the window or not. I replaced the diode. Now, this one died from a stupid test i did.

One more question, what do you mean with shielded? If this spike could kill an diode, it would definitely kill an ARM processor. An processor (even in a desktop) is connected to case.

Could it be that it sparked over somehow? The diode is case negative, and it had an internal photodiode and (maybe) it was an floating semiconductor or something and welded it self to other layers. Still odd that this doesn't happen on an Xbox 360 diode is connected to case (the same way this is). However thanks for your response.

 

[diachi]

H
One more question, what do you mean with shielded? If this spike could kill an diode, it would definitely kill an ARM processor. An processor (even in a desktop) is connected to case.

Could it be that it sparked over somehow? The diode is case negative, and it had an internal photodiode and (maybe) it was an floating semiconductor or something and welded it self to other layers. Still odd that this doesn't happen on an Xbox 360 diode is connected to case (the same way this is). However thanks for your response.

Perhaps the iPod/Processor has built in ESD protection - the diode does not, usually an additional component although IIRC some diodes do have it built in - most don't.

You said you were still getting some light out? Sounds like catastrophic optical damage (COD) - not damage to the actual junction per se - could just be that your other diode was A) Running closer to it's limits than the 360 diode and whatever ESD got through pushed it past the edge or B) The 360 is just less sensitive to ESD than your other diode.


Hard to say for sure.

 

[BlackOps]

Perhaps the iPod/Processor has built in ESD protection - the diode does not, usually an additional component although IIRC some diodes do have it built in - most don't.

You said you were still getting some light out? Sounds like catastrophic optical damage (COD) - not damage to the actual junction per se - could just be that your other diode was A) Running closer to it's limits than the 360 diode and whatever ESD got through pushed it past the edge or B) The 360 is just less sensitive to ESD than your other diode.


Hard to say for sure.

Hi, thanks for your reply. yea, there are sooooooo many unidentified semiconductors on an iPod circuit board, they are almost all the same size (micro) there is probably 1 or 2 clock signal generators in there, and the rest are TVS's (transient voltage suppresors). I tought they where only used as input protection trough the dock connector. But apparently, it isn't. I wish i could solder that small so i could put such chip in my build, as i'm kinda afraid now, that i will kill my new diode with normal ESD from myself (not piezo lol) and waste another 45mins on cleaning the diode windows.

The diode power was a little lower than its sweat spot, an xbox 360 diode is always underpowered. So this might explain it. It still does emit as much light as my 5mW red does strange enough, so it still kinda lases, but very minimal.

One more question, can i use this dead diode as an test load to set current for my incomming diodes? As the junction probably isn't dead right? That would be awesome. If so, could i test the diode in some way? I would like to put this diode in some use

However thanks!

 

[Benm]

One difference between a laser diode and most processors and other digital chips is that those actually have clamping diodes on any input. If you were to put a voltage over 5 volts on any input pin of a microcontroller, there is an internal diode that actually powers the chip from that input voltage. Somethimes this can actually be done quite reliably, but it's not recommended.

Laser diodes have none of this. If you introduce a voltage spike (for example through inductive coupling between your piezo toy and the laser diode) that can power the diode to optical destruction in forward mode, or destroy it in reverse voltage mode.

As for LED'ed diodes as dummy loads: if the damage is purely optical you can use them. There is a -slight- difference in electrical characteristics due to the diode no longer lasing, but for any rough test that will not mattter much.

This is only true if the damage is optical only (broken facette or something).

 

[BlackOps]

One difference between a laser diode and most processors and other digital chips is that those actually have clamping diodes on any input. If you were to put a voltage over 5 volts on any input pin of a microcontroller, there is an internal diode that actually powers the chip from that input voltage. Somethimes this can actually be done quite reliably, but it's not recommended.

Laser diodes have none of this. If you introduce a voltage spike (for example through inductive coupling between your piezo toy and the laser diode) that can power the diode to optical destruction in forward mode, or destroy it in reverse voltage mode.

As for LED'ed diodes as dummy loads: if the damage is purely optical you can use them. There is a -slight- difference in electrical characteristics due to the diode no longer lasing, but for any rough test that will not mattter much.

This is only true if the damage is optical only (broken facette or something).

Hi, thanks for your reply, that makes sense. I think i'm gonna make an ESD protector for my laser. Its going to be a pain as it has to be micro, hopefully i can make the protection measure as small as the space between the diode pins.

I think its only optically damaged, but i'm not sure yet.

One more question tough, do i have to worry about ESD when my laser is done again? (without the protection measure)

Does the piezo generate more voltage then we can with ESD? If this kills the diode, why don't we use wrist straps or something similar when operating the laser? Kinda disappointed in how sensitive these type of semiconductors are compared with microprocessors which are already sensitive.

However thanks for your reply