G-ball removal testing

[deadshadow]

Re: B-ball removal testing

very interesting testings,nubm05 is the cheaper and will do the job right,as for sealant i would suggest to use some thermal silicon,its not vaporizing and also has nice grip+ thermal conductivity.generally you see the degrading results in around a month so different gasses will give us the main idea.

personally i think the factory gball doesn't have any specific gasses,its just air free(vacuum)

 

[Benm]

Re: B-ball removal testing

What makes you think its vacuum?

It's perfectly possible of course, and you may not even notice it when you decan one with such a small volume, but then again what would be reason to have a vacuum in there in the first place? I think it would make manufacture more more difficult.

There could be a partial vacuum there if they are sealed at high temperature and the air in them contracts as they cool down after being sealed - no idea.

I still think that diodes are mostly canned (with a planar window or permanently fixed lens) is to keep contamination by particles and perhaps water out, not to prevent oxygen or CO2 coming in.

Most semiconductors just are not sensitive to oxidation by normal air. Yet virtual all chips are encased one way or another (in a package, under a blob on a pcb etc). Good reasons for this would be: preventing ingress of water, and in the case of powerful optics, dust. Another would be to keep -light- out: you can de-can an older metal transistor just fine and it'll work forever, but shining light on it will turn in on and mess with any circuit it is in.

I'm quite sure of the latter as i've used standard salvaged transistors as light sensors by decanning them long ago - those came free of charge whilst things like photodiodes, transistors and even ldr's came at a price at that point.

Decanning didn't hurt them one bit though, in a closed case they would work just was well as the ones with the can intact (think TO-39 and TO-18 form factor transistors here).

 

[LewDude]

Re: B-ball removal testing

Subscribed! Pretty cool you have access to the tools / enviro needed to do this. As mentioned before, the nubm05 is likely the best bet price-wise.
Good luck!!

 

[RedCowboy]

Re: B-ball removal testing

The glass in the window of the diode can runs all the way across the top of the can, a vacuum could cause that window to implode and cause the seal to suck in air, why have 14 pounds working against the seal when they can have an inert gas at equal pressure and no load on the window, so I doubt it's a vacuum.

 

[paul1598419]

Re: B-ball removal testing

I doubt it too, but it would only be 14 lb/inche^2. Since the window is not an inch^2, it wouldn't have 14 lbs against it. Depending on how the widow is fixed to the can, it could be possible to have a vacuum in it. I doubt it because the expense of building them that way would likely be prohibitive.

 

[deadshadow]

Re: B-ball removal testing

my guess is mostly stands to the fact that Chinese sellers heat the whole bank till the soldering melts(around 200c?).Gballs are too easy to extracted cause they don't have good grip at all,so i could imagine those cans start flying and popping by gas expanding at 200C if those wasn't vacuum(like popcorn:crackup.
also when the 1st batch of gball diodes came out i remember someone on the forum keep asking the manufacturer if they are filled with any gass,they did replied with "no" but they refused to give any further explanation at all.
the answer could be partially true.

2nd scenario is that are filled with pure oxygen without the extra stuff that "air" contains but that doesn't answer why those cans don't pop at high temperatures.

 

[kecked]

Re: B-ball removal testing

I don’t see a vacuum or pressure as these have to survive processing and part involves air delivery. That could change the gball position. So let’s get this designed up and agreed cuz I’m not buying these diodes twice.

Control unchanged diode at recommended drive
Decanned diodes
1 air
2 nitrogen
3 argon
4 decanned in argon and placed in argon filled box.
5 decanned in argon and regular lens in regular housing Ie nothing else special
6 decanned in air but then cleaned of dust with can air before adding lens
Etc....

So I can do some of the special ones and others can do the less equipment intense ones. This divides the cost and labor. I say we publish the results in photonics mag.

Does anyone have dead diodes they can send me. I want to practice decanning as I have not done it. Also want to look at failure of the die.

 

[RedCowboy]

Re: B-ball removal testing

I doubt it too, but it would only be 14 lb/inche^2. Since the window is not an inch^2, it wouldn't have 14 lbs against it. Depending on how the widow is fixed to the can, it could be possible to have a vacuum in it. I doubt it because the expense of building them that way would likely be prohibitive.

Actually it's 14.7 lbs. per sq. inch average at sea level, but I assumed most people knew what I was talking about and I was talking about the atmospheric pressure working against the cans seal as well as the window.


Also I have seen some of the GBall diodes GBall cans pretty much fall off to the touch, seems the Chinese ( cook em out ) method is melting the can loose in some cases.

 

[paul1598419]

Re: B-ball removal testing

@kecked.

You can edit the title of this thread to change the "B" into a "G". If it is going to continue, you might as well do that as it looks strange.

Edit: Yes, I knew it was 14.7 lbs/inch^2, but left it the way you cut off the last significant figure because for my purposes it didn't matter.

 

[Benm]

Re: B-ball removal testing

I doubt that pressure or vacuum would play any role here, these volumes are tiny.

Pressure problems increase with the cube of something. You can fairly easily demonstrate that with a syringe: If you get a 10 ml one and leave 1 ml of air in, and then close it off and pull the plunger out to 10 ml, the pressure inside would be 0.1 atm. Despite being made out of fairly thin plastic the thing hardly even deforms under that.

Seal off a tanker truck and lower the pressure in that to 0.1 atm and it will implode with a pretty impressive bang. Something like an oil barrel will also crumple up when doing that.


When testing decanned diodes i don't think this has to be overly complex, as long as you have a cleanroom/chamber. Basically you'd have to test only 4 conditions:

- unmodified [control]
- decanned in argon and ran in argon [oxidation damage test]
- decanned, and ran in clean air (i.e. particle free as in a semiconductor fab) [control for dust/soot/etc]
- decanned in clean air and running in "normal" air with the usual pollution that can cake on [confirming it's the pollution that does the damage, if any].

But before even doing that it may be interesting just to see if any gas is in these diodes, and if it sealed properly. You could just submerge a diode in decently deep water (or in shallow water under a partial vacuum) and see if any gas leaks out.. if it does it means it's not even sealed to begin with and further testing may be pointless.

If no bubbles escape from the diode, clip the case open under water, and look if you see any bubbles: if you see none it confirms a vacuum, if you do see any it confirms there is -some- gas in there, and would make the testing interesting.

My bet it's just air in there really - some LG drives also feature can-less red laser diodes that seem to work just fine, as long as you don't get any crap on them.

 

[kecked]

Ok I decanned a num07e in a nitrogen glovebox and let it run for almost 2 months straight at 750ma. So not hard at all. Came to look today and it has died. I sealed the whole thing in a glass jar so not even dirt could get in.

I decanned a second one but that one I drew a vacuum on. Same diode same conditions. It's still running. I do remove the vacuum to test power and reseal. The diode dropped about 10% in power over the time but so did the one that died. Both were on the same battery pack using 317 as drivers. figured no modulation no problem.

I'm out of diodes to repeat the nitrogen test. power hung around 700-800mw at 750ma. The jars never got warm. were in basic 12mm housings with glass g2 lens. Sorry no pictures. can't bring camera in space where glove box is located. They would require me to autoclave the phone.

Conclusion...non at all. Not enough data.

 

[paul1598419]

You need more heat sinking than a 12 mm module. That is not nearly enough for .75 amps. If you place the module inside a bench heat sink it should be able to run uninterrupted or in a lab host with a fan for active cooling.

 

[kecked]

Yea I need to do a better job planning. Just wanted to try anything. Interestingly it isn’t hot to the touch. Very warm but not burning hot. All I had on hand was 14 gauge wire so that might have helped as I wound some around the housing to stabilize the connections. Nothing was soldered in place. I should have done that before bringing it in. Well they are booting me out of the box soon so it won’t matter.

 

[paul1598419]

At 3/4 of an amp I would expect the module to get quite hot as it isn't much to sink and remove heat from the diode. Maybe you weren't actually getting that much current. You said one of them died, so it could have gotten hot. It is nice of you to run tests as there has been a lot of speculation about what would work best in decanned diodes.

 

[RedCowboy]

Slightly off topic, the NUBM42 and NUBM49 have awful divergence, about twice the bar width as the NUBM44, so at 15 feet the NUBM44 with G2 makes a 1.75 - 2.0 inch long bar, and the NUBM49 makes a 3.5 inch long bar, the 49 is doing 5.9w @ 4.5a and the 44 is doing 6.75w+, the 49 is more like 450-455nm but the divergence is awful, this is with the GBall can/lens removed.

 

[kecked]

I’ll try to save up some pennies and get more diodes but I have to wait till December to get back in the glove box. It’s dust and static I am concerned with or I’d just use a plastic bag. I can plan this out better too.