Extracting Diodes From Nichia diode bank

[Ricker]

Would you mind telling us how you cold extract them ABarnett? Or maybe a PM if you don't want on here?

 

[Ricker]

Yeah I'm about to start chopping. I wanted to double wrap the diodes before I started. I used a plastic type masking tape to cover the window and made a little bubble so the tape won't leave a residue of adhesive. Then made the caps from shrink tube. Now I can go at it without having to worry to much.

Wow great job, that's what I'm working on right now, covering the lenses.

 

[ABarnett]

I'm just going to use a dremel with the thin cutting disk. I get a bag of ice and wrap it in a towel so I can cool it down. Aside from that I just be as careful as possible and chop away.

Also if you think you may have scratched the window or messed it up at all you might want to decan the diode. I've had these blow just because of a spec of dust being on it.
Once it blows that's it. This is a picture of an NUBM44's widow that went.

 

[Ricker]

Yeah that's basically what I'm doing too.

 

[RedCowboy]

I know about popped windows

I think I can use my carbide metal cutting blade on my table saw and rip a 1/8 inch wide grove the depth in my pic above at the bond line between the diode and block right down the side, I will set my fence to cut at the bond line, if I need a 2nd cut I make a 2nd cut, just right down the side and with a sharpened screwdriver pop them out through that cut.


I love the heatshrink caps, I will borrow that thanks. :thanks: +rep owed when it will let me again.

 

[Ricker]

I know about popped windows

I think I can use my metal carbide metal cutting blace on my table saw and rip a 1/8 inch wide grove the depth in my pic above at the bond line between the diode and block, just right down the side and with a sharpened screwdriver pop them out through that cut.

You mean cut right along the line of the bottom of the diode where the pins start? then pry the diode out? that might work! just be careful not to bend the diode frame.

 

[ABarnett]

Yeah try to expose at least 2/3 of the base of the diode. Where the diode is soldered to the block. If you leave anymore than that you may bend the diode. The only diode I've wrecked was one that wasn't exposed quite enough. Once you get enough of the block cleared away then you can use a flathead screwdriver or something flat and tap it lightly until it breaks free.
Diodes bend pretty easily. They are brass or something.

 

[ABarnett]

Off topic but I just shined a laser at the phosphor on the color wheel and it's absolutely crazy. Soooo bright and green. It's nuts. I wasn't expecting that.

 

[Ricker]

Wow that is too cool! I'm going to have to grab mine out of it and try!

 

[Ricker]

When the diode is done, do you suggest putting some canned air to it or anything?

 

[RedCowboy]

My dead nubm44 and nubm07e stick to a magnet, like the backs are steel.

This is a cool pic ABarnett

 

[ABarnett]

I don't know how much work you did extracting the diodes without having the protected but canned air is better than touching them. If I was you I would just closely examine each diodes window and if it's anything except flawless you may want to decan it. I don't really know but I would rather have a decanned diode over a blown one.

 

[Ricker]

They were covered 90% of the time, a couple times the tape came off while there was dust in the air, I ordered a micoscope, not sure how powerful it has to be to see any dust particles.

 

[RedCowboy]

If it's a usb scope they get a good close image, you can hold the diode near the scope lens and really get a good close up.

Also a basic 10x jewelers loupe will work well, I have a 10x 20x and 30x from flebay that work great.

All my dead diodes stick to a magnet, the backing seems to be ferrous.

 

[Ricker]

This isn't tedious at all... :tired::undecided:

 

[Ricker]

I wish First Contact was a bit cheaper. :/ Amazon.com: Photonic Cleaning Technologies RFCAK Red First Contact Amazon Kit: Industrial & Scientific