Extracting Diodes From Nichia diode bank

Well, I am now taking apart my second block, and I just noticed that one diode came off the piece that it holds to easy with a simple pry, i noticed that when I was cutting next to it that a little bit of "liquid" bubbled out from next to the diode, then I pried and it came off super easy. Now maybe this was because of A) the vibrations of the dremel, B) the heat from cutting so close to the diode, or C) when installing the diodes they didn't cook it or what-may-have-you long enough, leaving some of this residue, here are some pictures of this goo under a microscope along with where I found it.
















Let me know what you think!

Looks like the can on that one fell victim to your extraction process

VirtueViolater said:

Looks like the can on that one fell victim to your extraction process

Click to expand...

How so? The can is fine and it works great. Are you talking about the tape and plastic on the top of the can to protect the window? A couple of pages ago I posted pictures of this process to keep debris from getting into the diode window.

Ok, must be what im seeing, looks like the can has a depression in it

Interesting, it could be a foreign contaminant from the factory, time will tell if it was a one time deal or something else.

So when are you going to build that multi diode monster?

VirtueViolater said:

Ok, must be what im seeing, looks like the can has a depression in it

Click to expand...

Oooh I see what you're talking about, I checked it a second ago and there's nothing on it, I think it's a reflection from the side of the diode, or my phone perhaps.

RedCowboy said:

Interesting, it could be a foreign contaminant from the factory, time will tell if it was a one time deal or something else.

So when are you going to build that multi diode monster?

Click to expand...

Right, I'm not too sure but any and all information in this thread is a plus!! I froze a diode in the freezer, seemed to have no effect, so I set my soldering iron below the block and diode kind of like a hotplate, using my thermal imager and thermometer gun to get it around 240c, and that didn't seem to do anything either, probably due to not being evenly heated with just the soldering iron under it.

Aaaand hopefully soon!!! I didn't forget you either, I just bought a new car and seeing a new girl, life gets in the way of hobbies sometimes! But I will get what I said I would send you out to you RC! I appreciate all your help along the way.

Just thought Id show two diodes I cut off block. I'm interested to see how they perform. Probably good. They make really good contact because they are soldered and the block material must have pretty high performance heatsinking properties. There is only 2mm between each diode so it's tough to get it separated fully intact.

This may be a noob question but...
Has anyone tried pressing them out? Just use Flaminpyros 9mm out press, get some pressure on it, then heat them individually with a solder iron while under pressure.

I tried to punch one out before but it bent the diode into a dome shape. I think it could be done like that for sure but you would need a punch that fit the hole behind the diode and had holes for the pins. It's soldered fully on the back and around the edges. The metal conducts heat really fast as it should so attempting to heat just one would require you to heat it up super fast. I imagine when they solder them in its done really quick using some type of induction or high frequency method.

This is nice, make a couple thread inserts so you can use off the shelf lens, even do G2's and cyl pairs, this is a neat idea, I wonder how well the block material transfers heat and if the coating has anything to do with it.

I don't know if the coating does anymore than prevent oxidation, it may help transfer heat. It has a grey themail paste where it contacts the heatsink. I wonder how it compares to copper. It does seem to have a lot of bubbles in it.

i have a bad feeling that nichia has been keeping tab on us......

They can only flex so much muscle when it it comes to product control. I'm sure if they get complaints from the anti laser people they would look into it. Also whatever division that deals with product safety and design my take a peak at this just to see how how it is people are defeating their tamper resistant design. But as long as they keep using TO-9 package diodes it doesn't matter what they do to slow the extraction. Also I'm sure it's a very small percent of people that would use visible diodes with malicious intent. There's plenty of companies out there that would be pleased to sell a 1000+ watt laser to anyone.
Nichia and the companies they sell to all have met or exceeded product safety guidelines. After that if a consumer buys said product and decides to use it in a way it wasn't design for they have limited liability. However if they know it's being used in a way that might be potentially hazardous they may be forced to address that if it became a serious problem.
But it's not a problem at all. It's just a fun hobby that requires basic safety.
There's absolutely no reason they shouldn't sell directly to us. It would be much safer for everyone in the long run .

So, I read this thread and nobody mentioned using an alloy to drop the melting point of the solder. Am I missing a reason this won't work?

You can set the block on a hot plate at just 100C apply a little BGA flux and melt an alloy of of bismuth, tin and indium (like ChipQuik) at the back of the diode and just wait for it to amalgamate all the solder. When's it's all molten the diode should press out and you can then soak up all the solder/alloy off the diode with copper braid. All without taking it over 100 degrees.

:beer:

I was thinking to make my socket to fit the notches in the edges of the diodes, it would have to be hardened steel, but a few degrees of rotation should crack it loose, but your idea could work from the pin side on the back.

Here's a block that's been extracted from, the top left never had a diode as this unit is a 7 diode unit, and the 2 inserts are a work in progress, well a test of concept, they are sitting there loose.

So looking at the walls your idea seems very plausible, what do you think?
I think maybe my socket idea was used from walls residue looking un melted, or do you think it was?

I would think a oval punch the size of the rear opening with pin holes drilled in it would work, just insert it over the pins into the oval hole and press it out from the back.

Hey guys!! Any new developments?! Buy looking at RC's picture above, my way of cutting them out with a dremel seems quite primitive. There has got to be an easy heat/freeze/chemical reaction way to pop these out easily that someone knows, who isn't under a Non-Disclosure Contract! Time to get the thinking caps back on! I'll have my pj Friday..ish, and will let you know of any and all progress! Thanks, and sorry for the 5 month bump, but I feel as though this subject has far more potential for us whom are at the stage of extracting our own diodes! Thanks!

Ricker