XHP50 De-dome = Yes?

My comment referred to your question about efficiency. You will still lose efficiency without the dome.

The cross isn't going to disappear; you've still got four individual dies with space between them. The apparent distance will decrease dedomed, but it will still be there. This is why you can't compare the XPH series to the MT-G2.

the gas de dome failed because the phospher came off, the dies still lit up (2 of them) all 3 worked before de dome, mounting can’t have failed because there’s only 3 pads (thermal,positive,negative) it’s not like the sinkpad XHPs where the pos and neg are split in half
I don’t know, I’m stumped.

I checked all 3 before de dome but only the 1 after with welding goggles, but they all seemed of the same brightness to eye, which would mean they all had 2 dies only lighting up

I’m working on the third one now to see if I can do it!

Anyone any ideas on why only 2 dies lighting up at 6V?!

You might need to resort to the slice method. Best if you can find an appropriate shim to use as a guide for the blade to get an even smooth cut, lubricate the razor, and make the cut in one sliding motion rather than back and forth saw cuts.

The reason I compared the XPH dedome to the MTG2 was when people started to dedome those some found out that the quality of the light got even better after dedome that was unknown before that. Most of the other cree emitters gets worse with an apparent less CRI & often time less nice tint, i want to know how the XPH are going to appear.

And i understand that the 4 dies characteristic aren’t going to disappear, but what we don’t know is the practical implications of a smaller cross artifact, it could be that more reflectors & host will look nice enough with a dedome for example, if a simple dedome can be developed that is :slight_smile:

ok, I’ve done it!
a little messy yes… but thats because i cant remove that white layer, the bond wires are underneath that, ill work on it and see if its possible to clean up and make it look tidier

still have the 2 die problem though!

!!

Are you talking about the "2 die problem" where 2 dies don't work? If so, I'd say that's a pretty big problem.

Yeah I’d say so as well, can’t figure out why though?!

Did you slice this one?

No, I sliced the dome off the top so it was flat, then used gas to get rid of the rest the phospher didn’t come off this time, must be all about timing in the gas with this one

I’m guessing a bond wire broke when slicing(partly why I suggest lubrication, other part is cleaner cut). It’s also possible a pad on one side of the substrate didn’t wet to the Noctogon and you’re only getting current through half the die. Did you test it before dedoming?

At any rate, thank you for sacrificing your time and money for what I like to call "BLF Science".

Like they always say, "someone has to take a hit for the team."

Success!
I switched onto a different MCPCB and we have power to all 4 dies!
It is possible to de dome an XHP!

Congratulations!

Now I'm going to recheck a XHP50 on which I lost two dies as well, see if a fresh reflow will revive iit

Nice, that is some very interesting news :slight_smile:

Maybe your other 2 XPH50’s can be revived also?

Nope the other 2 can’t be revived, one has missing phosphor, (in gas to long?) and the other looks like broken bond wires (hot de dome)

If the silicon floats in gas then maybe put the mcpcb in face down so the phosphor won’t lift off with dome bits. Just a though, maybe irrelevant.

Have this way ever been tested on the MTG2? first slice then a quick gas dedome for the rest…….if it works as good as on the XHP50’s maybe thats an safer way for them also.

Maybe Cereal Killers MT-G2 dedome method will work best for this?

MT-G2 dedome & polish method

I don’t think you’ll be able to use the gas method on MTG2 I tested a scrap MTG2 to see if it would work, it doesn’t, you’ll just end up with a lot of phosphor if not all of it floating off
I sliced the third one before de domed it, but onviouslyndisngbpolish it, however it did look ok just how it was, that is with a flat ‘cloudy’ top (considering the dome is already ‘cloudy’ compared to XML etc)

Ok thanks, good to know.