Saabluster's R&D

[ma tumba]

Grains are inherent to laser beam and there is quite a few means of dealing with it. Generally speaking they introduce some sort of decoherence to smear the grains out.

but in our case when we think of extreme throw, i think that this diffusion is provided by the atmosphere, same effect that causes stars to blink. Plus, the ability of human's eye to see the grains degrades with brightness

 

[Enderman]

You're right.
One of the things I find appealing about long range lights is the beam coming from the device.
With lasers the beam is extremely thin for the first few hundred metres so it's not really visible.

I would only use a laser if it had a large beam expander so that the beam is at least 4" diameter, making it very visible (cause it looks cool )
The beam expander also reduces divergence to give more throw.

WickedLasers has a small beam expander for their lasers which is about 1" diameter, maybe you can try making a larger one with all the lenses you've got.

Color combiners (to join the red, green, and blue lasers) are very common and commercially available so that's not really an issue.
Pretty much every projector has one. Although you might need a special one for lasers.

 

[saabluster]

Took off the whole last week to work on construction and painting our house with my wife. Also celebrated our 18th year of marriage.

Despite pulling 12-13hrs a day working on the house I managed to squeeze in 2-3 hours at night for research. Ordered more equipment. $1400 worth this time. I'll keep the specifics close to the vest for now but this will add greatly to my capabilities. You will be seeing the benefits of this within this very thread soon. For today I have one of my projects into the thermal side. An XPG2 with the substrate thinned to the thickness of a couple pieces of paper. Then to top it off it is mounted to pure diamond which is mounted to solid copper. This is approaching the limits of material science for something that can then be used in an end product.

 

[Devildude]

How close are you to the bond wires being the limiting factor for how many amps can be applied to the die?

 

[Str8stroke]

Very cool stuff. Can't wait to see more.

I always wondered how the bond wires, being so fragile can handle what they do, even on some stock rigs.

 

[wimmer21]

Took off the whole last week to work on construction and painting our house with my wife. Also celebrated our 18th year of marriage.

Despite pulling 12-13hrs a day working on the house I managed to squeeze in 2-3 hours at night for research. Ordered more equipment. $1400 worth this time. I'll keep the specifics close to the vest for now but this will add greatly to my capabilities. You will be seeing the benefits of this within this very thread soon. For today I have one of my projects into the thermal side. An XPG2 with the substrate thinned to the thickness of a couple pieces of paper. Then to top it off it is mounted to pure diamond which is mounted to solid copper. This is approaching the limits of material science for something that can then be used in an end product.

You've got me excited, Michael. Pure diamond on solid copper!?!? And Wow... that substrate... so thin. Reminds me of the ham sandwiches mom used to feed us kids.

Congratulations to you both!

 

[Enderman]

With a copper block that large, why not just divide it into 3 electrically isolated pieces and remove the substrate altogether?
Then you can use 3 separate aluminum heatsinks, or a single anodized one since anodization will prevent the aluminum from shorting the 3 segments.

 

[saabluster]

You're right.
One of the things I find appealing about long range lights is the beam coming from the device.
With lasers the beam is extremely thin for the first few hundred metres so it's not really visible.

I agree wholeheartedly. I love the feeling of wielding a device that can project energy in a viscerally visual way. This is relatively easy to do with lasers with no speckle.

 

[saabluster]

How close are you to the bond wires being the limiting factor for how many amps can be applied to the die?

This obviously varies based on the LED. With the XPG2 I am definitely bumping up against the limits of the bond wires. I have dealt with this in varying ways. One of which is supporting the bond wires with a coating of conductive epoxy. This has allowed a much higher ceiling current wise. The XPG2 bond wires stock are good to about 6A. My modified versions go to about 9A.

 

[saabluster]

You've got me excited, Michael. Pure diamond on solid copper!?!? And Wow... that substrate... so thin. Reminds me of the ham sandwiches mom used to feed us kids.

Congratulations to you both!

This is nothing. Working to refine the method and leave no ceramic behind. In fact I want to be able to have the active epi layer in direct contact with diamond. There are quite a few things preventing this ATM but I am working through them.

 

[Offgridled]

Ok I'm asking what you are doing to your house.

 

[saabluster]

Ok I'm asking what you are doing to your house.

Building a loft. With our second kid we lost our dining room. Turned it into a bedroom. The area that was originally intended as a dining room was pathetically small. Right now it houses our computer stations/office. The loft is going to be our new office area and library. I have a relatively small place at 1371Sq feet. This requires efficient use of space. My shop/lab is merely 12'x12'. But I built that as a two story as well so it is really double the basic footprint. Because my house was not designed to be a two story I was a bit space constrained vertically. This required me to engineer a floor very thin. Only 5" thick yet free standing for about 14'. But despite that the floor is super solid. It was down to the way I overlapped panels, good quality wood with tons of fasteners, and adding structural foam in between.

 

[saabluster]

With a copper block that large, why not just.... remove the substrate altogether?

You mean like this? The ceramic has been completely removed.

 

[saabluster]

Yet more extreme LED testing. This is an Oslon Square with the entirety of the ceramic removed. The goal with these tests is to make as low a thermal resistance as possible to be able to pump more current to the chip.

This is an XBH with the ceramic removed and attached direct to copper. I didn't do a good enough job lining this up as a portion of the die is off the copper. This resulted in an early death. Alignment and thin solder bond lines become even more critical without the ceramic layer.

 

[magnum70383]

Another $1200 out the door... lol

 

[saabluster]

Another $1200 out the door... lol

Plus tooling. I wish it was only $1200. Also bought two reels of XHP35 to test on. Yes reels. Knee deep in chemicals to test. I don't know if it will work but I am attempting to remove even the base SiC substrate the epi layer is grown on. One of the projects today was thinning it down an mounting the result to diamond. Many many tricky bits with this.

 

[saabluster]

I think I finally have perfected dedoming the XHP35. What do you think? This is with the newer conformal type phosphor coating instead of the easy to dedome phosphor sheets used on the XPG2.

 

[Offgridled]

Great job you've done there. I had no doubt you would get this figured out. Now what your going to do with it is truly exciting)

 

[saabluster]

Sad to see what Photobucket's greed and ransom demand has done to all my threads. Had they handled this differently they could have had some of my money. Not $400 a year mind you but something reasonable. As is they will be getting nothing from me. I still need to figure out what I'm going to do posting pictures here.

That said that is not why I started this post. Just wanted to update a couple things. In the last hour I "finished"(always more tweaking to do) research into phosphor modification on the XHP35. At 2.2A I have increased performance 5% just from phosphor modification alone. I have also been working on the thermal side of this LED and have significantly increased thermal extraction such that high amperage conditions are netting about a 13% increase in light output just from a decrease in junction temp. This of course will also allow for a slightly higher current ceiling.

 

[Capolini]

Thanks for the update.

Postimage.org works VERY well for me. Been using it about 1 year. FREE, Very fast and NO freezing or sluggishness.

Photobucket became useless. Took forever to upload/down load pics. and would constantly freeze.