Breaking diodes? How easy is it?

[RedlumX]

.
The capacitor on the board will blow the arse out of the diode if the driver has been previously powered with an open circuit diode.

I guess this is one of the more frequent causes - you fiddle with a connector, there is a brief interruption of the connection (a few ms suffice), the driver feels it must supply more current and drives the output up to almost power supply level, then the connection comes back - and poof. An output capacitor makes this even worse.

Professional drivers often have an open-circuit detection and switch off immediately in this situation.

One simple remedy is to have a Zener directly at the driver board, before any connector, with a voltage just a tad above the intended forward voltage of the laser diode (one needs to cut and try). Then no matter what happens, the driver never ramps the voltage fully up. Of course the Zener must have a proper power rating.

Otherwise a large cap, like 10uF ceramic, directly across the diode helps to absorb the current inrush (it should be considerably larger than the output cap of the driver). Of course that may not be a good solution if one wants to modulate the driver. Then a Lasorb is certainly the best choice.

[heruursciences]

I have killed many diodes this way, nice working laser, pulled the main plug, put the laser back on power and death diode
Since I switch off the 5V with the interlock instead of the mains I never had a problem with killed diodes.

Preventing this is easy, have a loading resistor in the output of you PSU that draws about 5 to 10% load on your power 5V supply, It is good protocall to switch off the 5V first, but doing this will prevent such things as tripping over the power cord, or somebody accidently unplugging your rig from cooking your poor diode.

[SideFX]

Ok I've really done it this time..

yesterday evening i killed 2 445's after eachother. It was built using aixiz in a z-bolt mount, driven by a flexmod N2.
I was mounting the whole assembly in my projector, moving it around to get the beam roughly lined and focused up, driving it with around 300mA for a small beam.

All of the sudden it went black! There is only a tiny blue dot visible if you look into the diode, and it draws only around 1mA current. No matter at what modulation.
The driver is fine, when i replace the diode the current goes back to the 300mA I set before..

I thought I might have shorted something while moving everything around.. causing spikes or anything.. So i got my spare and started all over. Made sure all leads were properly isolated.. At first everything is running fine.. even running it at 1200mA for a few minutes.. All the tinkering and moving around was done with the diode powered off.. but I need to turn it on when doing alignment.
And again after a short while it went black..

Any thoughts on this?
* I run the flexmod from a 12V power supply, instead of 5V i've done earlier. Could this be a problem?
* And the diode casing isn't connected to a pin right? So I couldn't shorted something out through the heatsink?
* Is it possible that when the flexmod transistor accidentally touches the casing (which is earthed) the diode blows? Of course I've taken precuations so this doesn't happen.. but I'm trying to rule things out..

..and is there someone who can supply me with a couple of new diodes..?

oh and PS sorry for hijacking your thread norty.. don't know it was worthy of a new thread :s

[Morgan]

As a by the by, I made the school boy error of soldering in reverse bias to a boost driver outputting 900mA and powered up. I thought this would've been completely fatal but once corrected the A130 diode was happy to give around 900mW optical output. I certainly wouldn't recommend this and won't be making the same mistake twice but note worthy. A very hardy diode!M

[norty303]

@SideFX - Are you using a LasOrb?

[Xytrell]

And does the flexmod have sufficient heat sinking? Maybe the thermal protection kicked in when it had to drop 8V instead of 1V, and switching the diode was long enough for it to reset itself?

That's my wild guess, anyway.

[mikkojay]

The flexmod regulator needs to be elecrically isolated from everything. If it touches pretty much anything, I believe bad things will happen. I have a tendency to lose those little isolator washers and sticker-pads that come with them, but they are an absolute must.
-Mike

[SideFX]

Hi Guys, thanks for giving it a thought

@SideFX - Are you using a LasOrb?

no haven't until now. When I get new diodes I will certainly do so. Don't know if it would've helped in this situation.. but I already spent to much on diodes..

And does the flexmod have sufficient heat sinking? Maybe the thermal protection kicked in when it had to drop 8V instead of 1V, and switching the diode was long enough for it to reset itself?

That's my wild guess, anyway.

Yes the flexmod is heatsinked. I don't think its thermal protection kicking in.. when I short the defective LD with a normal diode, the current goes back to what I expect, so the driver is delivering

The flexmod regulator needs to be elecrically isolated from everything. If it touches pretty much anything, I believe bad things will happen. I have a tendency to lose those little isolator washers and sticker-pads that come with them, but they are an absolute must.
-Mike

Maybe this is the problem.. The heatsink is directly mounted on the casing. there is a sticker-pad between the transistor and the heatsink.. and I'm using plastic bolts.. but the pad is damaged quite a bit.. I can image some electrons are passing through here

Ok TODO list when new diodes arrive (when I find someone who will supply them hint hint ) - proper isolation between transistor and casing, - LASORB, - back to 5V.. have good experience with that.

[mccarrot]

It could also be you power supply, some Switched PSU's gives a small voltage spike when mains is cut and kills the diode as the driver has no protection. (have killed many diodes this way!)

better drivers have there own regulated PSU on board.


Make sure you cut the secondary side of the PSU first before cutting the primary mains.

[shrad]

I had bad experiences with a bad PSU, killing opnext diodes, so I can only agree

a good PSU has an isolated output and there's no voltage between V+ and earth expect the regulated DC output voltage if you ground both V- and earthing