Dissecting an MC-E

[saabluster]

I don't have a lot of time right now to go into a lot of detail but I wanted to share some pictures I have taken of an MC-E that was blown because it was run without having a heatsink. This is the LED given to me by Aircraft800 to do some "research" with. So thank you very much Aircraft800!

I have much more work to do and actually bought a digital microscope just for this project. When I can find more time I will continue this thread with those pictures but for now I will post the ones I took with my camera.

Nothing seems too amiss looking at this picture but....

The LED got so hot that the dies unsoldered themselves. This is a view of the glass dome and the two dies that stuck to the gel. You are looking at the underside of the dies.

This really was very interesting as there was three different failures all in the same package. Two dies unsoldered and stuck to the gel. One more unsoldered itself but managed to keep its bond wires intact. Its the one "standing" up. The bond wires are holding it there. The other one stayed soldered but had the phosphor stick to the gel.

Notice the die that still has the bond wires attached? I wonder what would happen if you powered it up?

Very interesting! I plan to try and take the good die out of this package and mount it directly to a piece of solid silver with my special epoxy. More will come when I can find the time.

 

[LED_Thrift]

Cool and interesting pictures. Thanks for posting this.

 

[cerberuss]

Really interesting post, I was wondering... if one had to take the dome off a good mc-e... ...would that then change the optical characteristics and Possibly make it more of a point source source of light... and behave better in reflectors.. allowing for better throw?

Just wondering... :thinking:

nevermind... found the answer in another thread... throw increases but overall brightness reduces

 

[spencer]

Looks really cool. I will be looking forward to updates.

 

[Aircraft800]

Unbelievable Pictures Michael! I like the new Microscope. How thick is that dome? I think that a flat top would greatly improve the optical characteristics of the MC-E for our purposes. Is there enough material to polish it down flat on top? The dome seems to exagerate the space between the dies, expecially considering it's appears smaller that the P7 which doesn't have as much of a problem controling.

 

[saabluster]

Unbelievable Pictures Michael! I like the new Microscope. How thick is that dome? I think that a flat top would greatly improve the optical characteristics of the MC-E for our purposes. Is there enough material to polish it down flat on top? The dome seems to exagerate the space between the dies, expecially considering it's appears smaller that the P7 which doesn't have as much of a problem controling.

Thanks. These pictures were actually done with just my camera and an aspheric lens I had laying around. There will be more to come. The dome is exactly like the one on the XR-E in that the base of the dome where it meets the package is where it stops. We have been discussing the optical issues in this thread. Polishing or sanding are not really an option. Not saying it couldn't be done there are just easier ways to accomplish this.

 

[Ryanrpm]

Thanks. These pictures were actually done with just my camera and an aspheric lens I had laying around.

I bet you didn't have to search hard to find it! I'm sure you have more than one laying around also.

 

[saabluster]

I bet you didn't have to search hard to find it! I'm sure you have more than one laying around also.

Yeah I have a "few".

 

[TexLite]

Great job on the dissection and the pics!

Is the silicone interface viscous or is it "set"?

Thanks,
Michael

 

[saabluster]

Great job on the dissection and the pics!

Is the silicone interface viscous or is it "set"?

Thanks,
Michael

It is exactly like the XR-E. A gel.

 

[65535]

Does that mean with medium physical force it would pop off? Say grinding the glass dome flat?

 

[saabluster]

Does that mean with medium physical force it would pop off? Say grinding the glass dome flat?

Just taking it to a piece of sandpaper would surely pop off the dome as it is the stickiness/suction created by the gel that holds it in place.

 

[djipe]

Columbo ! Where are you ? ? ? :candle:

Here we have a serial cree kiler


Nices picts, Good work guy.

 

[csshih]

That's not dissection.... it's vivisection!! that cree was still alive. murderer!!

 

[saabluster]

That's not dissection.... it's vivisection!! that cree was still alive. murderer!!

HE WAS DEAD WHEN I FOUND HIM! I SWEAR!








....Must destroy evidence

 

[Gryloc]

I plan to try and take the good die out of this package and mount it directly to a piece of solid silver with my special epoxy. More will come when I can find the time.

Have you thought of attaching the single die to a heatpipe? Then you can solder the heatpipe (or epoxy it) into a standard finned heatsink laying around that has the proper hole size drilled into it. Soldering it could be tricky, but this "special epoxy" sounds like it could do the trick. Actually, could you tell us anything about this epoxy? I am guessing it is electrically conductive and could be applied thinly.

Congratulations on being able to get a single die with bond wires still attached to it. Its too bad that it meant losing an MC-E for it. I have a dome-less neutral white MC-E that is almost useless to me (not as bright), so it would be neat to see if I can harvest separate dies like this. Hmmm...

Just be careful when eclectically connecting those free bond wires. I learned from Newbie that if you would try to solder the fine gold wires, you will find that they will dissolve into the solder. It will take some sort of low-temperature bonding technique to eclectically connect them. How about low melting point solder (silver based?), electrically conductive epoxy, or mechanical clamps (like crimping a fixed piece of copper or aluminum around the bond wire). Please keep us informed on this. I am extremely interested in this. Back in the day, I created a thread asking if it is possible to de-solder XR-E dies (back in the P2/P4 days) to make a quad die emitter (ceramic substrate with 4 dies like the current MC-E). I had 3d drawings made up and had a plan, but lacked time, money, and skills to attach dies with bond wires and phosphor coating. Its in my thread history, but I do not have a link on hand now. Good luck!

-Tony

 

[saabluster]

Have you thought of attaching the single die to a heatpipe? Then you can solder the heatpipe (or epoxy it) into a standard finned heatsink laying around that has the proper hole size drilled into it. Soldering it could be tricky, but this "special epoxy" sounds like it could do the trick. Actually, could you tell us anything about this epoxy? I am guessing it is electrically conductive and could be applied thinly.

Congratulations on being able to get a single die with bond wires still attached to it. Its too bad that it meant losing an MC-E for it. I have a dome-less neutral white MC-E that is almost useless to me (not as bright), so it would be neat to see if I can harvest separate dies like this. Hmmm...

Just be careful when eclectically connecting those free bond wires. I learned from Newbie that if you would try to solder the fine gold wires, you will find that they will dissolve into the solder. It will take some sort of low-temperature bonding technique to eclectically connect them. How about low melting point solder (silver based?), electrically conductive epoxy, or mechanical clamps (like crimping a fixed piece of copper or aluminum around the bond wire). Please keep us informed on this. I am extremely interested in this. Back in the day, I created a thread asking if it is possible to de-solder XR-E dies (back in the P2/P4 days) to make a quad die emitter (ceramic substrate with 4 dies like the current MC-E). I had 3d drawings made up and had a plan, but lacked time, money, and skills to attach dies with bond wires and phosphor coating. Its in my thread history, but I do not have a link on hand now. Good luck!

-Tony

Funny you mention the heatpipe idea. Last night I bought several heatpipes to do this very thing. I got to thinking about it and decided if I was going to do this I might as well go all out. Yes the epoxy is electrically conductive. Here is a PDF on it. I am using the 6030 hk with a thermal conductivity of 60 W/m°K. I remembered what newbie had said about soldering to the bond wires so I figured I could use my epoxy to also connect to the bond wires. I am trying to decided whether or not to encapsulate it in clear silicone(with no dome so the die size is small) or try and cool the top of the die with a liquid coolant. What do you think? I am going to try and put this in a DEFT so that may rule out the coolant idea but it sure is tempting.

 

[saabluster]

One other idea I have had is to scrape off the phosphor Cree uses and coat the die in quantum dots. I have found several places where you can buy them but they are fairly expensive. If people wanted to go in on it together I might be able to do it but as it stands I cannot afford it at the moment. Sure would be cool wouldn't it?

 

[R33E8]

One other idea I have had is to scrape off the phosphor Cree uses and coat the die in quantum dots. I have found several places where you can buy them but they are fairly expensive. If people wanted to go in on it together I might be able to do it but as it stands I cannot afford it at the moment. Sure would be cool wouldn't it?

Another problem with quantum dots would be finding an appropriate epoxy that doesn't render them useless.. I think they have a problem with clumping together when mixed improperly or without the right substance..

 

[saabluster]

Another problem with quantum dots would be finding an appropriate epoxy that doesn't render them useless.. I think they have a problem with clumping together when mixed improperly or without the right substance..

I was not going to use an epoxy. I think it would be hard to find a clear epoxy that could handle the extremely high heat of the overdriving I plan on doing. If you know of one let me know. My thought was to use a 2 part clear silicone with a thinner added so that I could apply a very thin layer.