Is it possible to de-dome SST & not to suffer Lumen Loss?

ma_sha1

Flashlight Enthusiast
Joined
Jan 10, 2009
Messages
3,042
Location
CT, USA
The benefit of De-dome SST-50 & SST-90 to reflector based lights has been shown many times now. Many posts including some of the lights I made in which that I observed between 30 to 40% increase of lux@ 1 meter using the same Rebel Deep SMO reflector & identical driving current.

It is also known that the de-dome (on any leds) will cause a lose of total lumens for sure, although the % loss is debatable.

With the obvious net gain of throw on reflector based light, I am planning to make transition towards making some de-doomed SST-90 flashlights.

The de-dome obviously send more light sideways, therefore benefit reflector but hurt aspherical applications.

However, there's some questions in my mind: What's the ideal doom-less form for Reflector based flashlight?

Leave it alone?
Coat with another layer of optically clear Silicone?
Silicone + a Flat top doom?
Silicone + a flat thin glass?

With some searching, I found this Report:

http://www.osti.gov/bridge/servlets/purl/963890-7R3mLc/963890.pdf

leddedooom.png


I think it answered quite a few questions, just a summary on what I've learned when it applies to de-dome.

1. If you apply clears silicone without dome (flat top, silicone alone)
The Silicone layer extract more photons out of led, but it's canceled
out by additional internal reflection at silicon/air junction.
As a result, it actually lost total lumen output, although not by much.

2. When glass doom is applied to the silicone, (Silicone & glass has very close defraction index, allows very little lost from silicone to glass), very little light is lost at glass dome to air junction (This is because the light is hitting the doom surface head-on, not at an angle like the flat top, ie. the least internal reflection possible). As a result, the additional light extraction by silicone is able to escape to the air & increase total lumens.

Because adding the silicon alone made no improvement, The net gain of lumens is almost completed resulted from the shape of the doom.

Therefore, it's not possible to prevent lumen loss when de-dome.
Putting additional silcone won't help (Unless you make a silicone doom, then what's the point of de-dome?

Even the factory doom-less SST has a much lower total lumen output than the doom version, nothing we can do about it, come to think about it, if there's something they can do to keep the lumen out put, then the factory would have done it. So my investigations are kind of pointless, I guess I just needed some data to reassure myself.

There might be an small possibility of improvement: If after de-dome, one add silicone & top it with a flat glass lid but with a AR coating on the air side only, perhaps, that may improve a few % points over the de-doom alone?

May be an optical expert can chime in to see if it make sense?
 
Last edited:

bigchelis

Flashlight Enthusiast
Joined
Oct 30, 2008
Messages
3,604
Location
Prunedale, CA
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

I will be testing possibly this weekend a VARA2000 Copper light that has a CBT-90. A CBT-90 is a domeless SST-90 from the factory and all the lumen loss considerations should have been addressed or we hope. The Star is a solid piece of copper from the factory and bigger then the SSR-90. Maybe the manufacturer discovered copper was better then aluminum and brass:naughty:


Here is a sneak peak, but I will open up a thread on it with pictures when I get it. It already does over 90K lux at 1 meter with rebel deep reflector:party:

VP2KST1.jpg
 

ma_sha1

Flashlight Enthusiast
Joined
Jan 10, 2009
Messages
3,042
Location
CT, USA
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

looks awsome! where is the 90k lux number from?
 

gt40

Enlightened
Joined
Oct 13, 2009
Messages
443
Location
bainbridge island
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

Check out the specs from the data sheet:
Absolute Maximum Ratings​
Symbol Values Unit​
Maximum Currents: 18 A

http://www.luminus.com/stuff/conten...3_cbt_90_w_product_datasheet_illumination.pdf

This seems to be a different beast than the sst90. I got blue led of death at 13 amps. They don't show output at higher levels but this may be a significant bump up over sst90...
 

Th232

Flashlight Enthusiast
Joined
Dec 25, 2008
Messages
1,064
Location
Sydney, Australia
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

It's the same die, but the thermal resistance of the package is greatly improved, hence it can take more power before burning out (provided you have it attached to an adequate heatsink, of course).
 

gt40

Enlightened
Joined
Oct 13, 2009
Messages
443
Location
bainbridge island
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

It's the same die, but the thermal resistance of the package is greatly improved, hence it can take more power before burning out (provided you have it attached to an adequate heatsink, of course).

I could only imagine one of these driven to 16-18 amps or ... god forbid, 20 amp range:devil:
 

bigchelis

Flashlight Enthusiast
Joined
Oct 30, 2008
Messages
3,604
Location
Prunedale, CA
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

This new VARA2000 Frankenlight has a gigantic heatsink made out of copper and a aluminum brace used as extention. I dont think it was made for sale for the public, but as an experiment to see how much we can push these SST type LEDS.

Here is a quick pix of what I mean by frankenlight:

varast1.jpg
 

gt40

Enlightened
Joined
Oct 13, 2009
Messages
443
Location
bainbridge island
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

This new VARA2000 Frankenlight has a gigantic heatsink made out of copper and a aluminum brace used as extention. I dont think it was made for sale for the public, but as an experiment to see how much we can push these SST type LEDS.

Here is a quick pix of what I mean by frankenlight:

varast1.jpg


Post some beam shots!

BTW, where to buy top bin wr cbt90?
 

bigchelis

Flashlight Enthusiast
Joined
Oct 30, 2008
Messages
3,604
Location
Prunedale, CA
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

Post some beam shots!

BTW, where to buy top bin wr cbt90?


The light was made by lamdalights (kevin) I dont know where he got the emitters and he also helped source the SSR-90 65k tint ones PCC used for our heatsink project.

The light came today, but since I was at work its at the post office for tonight. 6AM I can pick it up and will take as many pictures as possible and do a nice review on this new CBT-90 Copper star, copper heatsink VARA2000.

Underneath the bare Aluminum is the copper slug:eek:
 

Al Combs

Enlightened
Joined
Jul 2, 2007
Messages
872
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

A very interesting pdf on LED domes! I was especially interested in the case of blue LED's you have in the picture. I could understand more readily a white LED where the 460nm peak is converted to a wavelength the eye is more sensitive to by a secondary fluorescence. The lumen by definition favors wavelengths the eye is more sensitive to, 555nm in particular. But the fact that blue LED's are brighter with a dome is a bit of an enigma to me. Quite frankly, I don't get it.:confused: Where did you run across that file?

I know you saw this thread because you post posted in it. The somewhat controversial Anything Brighter thread that introduced the concept of the reflection collar. It does seem to offset the output loss of removing the LED's dome. Of course it only works with aspherics. But that actually works out well since dome removal is most noticeable when using an aspheric. Perhaps something like an upside down IMS27, if you can figure out what piece of it to remove.

I found Techjunkie's Long term effects thread very interesting. Silicon jerky.:laughing: I did not realize there was a gel under the Luminus domes. Is it similar to what's found under Cree's XR-E? Speaking of Cree, their XLamp Reliability pdf rev5 used to say, "The most stressful test for high-power LEDs is the Wet High-Temperature Operating Life (WHTOL) test as defned by JEDEC JESD22 Method A101-B." The current version only says it's a test they subject them to, nothing about the degree of stress it represents. I should say this is another controversial point. But if the LED is left bare after dome and gel removal, it will allow condensation on the die as you turn it off. I have no idea what kind of damage that causes in the long run. Cree never said exactly what the problems associated with condensation are.

Personally I would go with silicon glue as a cover rather than leaving it bare. I'm not sure where you would get the optical grade 2 part silicon glues the LED manufacturers use. Even if you could find it, I'll bet the smallest quantity is gallons and costs hundreds. But as thin as it would take to cover an SST-50 or 90 LED, I'm not sure transparency matters that much. I think the acetic acid found in most silicon glues and its possible corrosive effects might be more of an issue.
 

ma_sha1

Flashlight Enthusiast
Joined
Jan 10, 2009
Messages
3,042
Location
CT, USA
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

But the fact that blue LED's are brighter with a dome is a bit of an enigma to me. Quite frankly, I don't get it.:confused: Where did you run across that file?

I found it by Google for optical Silicone for LED use.
You can see the project here. It's half million dollar grant awarded to Osram Sylvania by the Government to conduct the research & they are obligated to publish the report: http://www1.eere.energy.gov/buildings/ssl/light/printable_versions/white_led.html

I know you saw this thread because you post posted in it. The somewhat controversial Anything Brighter thread that introduced the concept of the reflection collar. It does seem to offset the output loss of removing the LED's dome. Of course it only works with aspherics. But that actually works out well since dome removal is most noticeable when using an aspheric. Perhaps something like an upside down IMS27, if you can figure out what piece of it to remove.

Like another guy pointed out in that thread, the concept is similar to
Incan HIR technology, where HIR uses a special coating to reflect IR back to the light source & make it glow brighter.

However, I am not convinced the reflection collar would make a longer thrower. Michael has shown an image that the collar did make the die brighter, however I doubt that he could use this to make DEFT throw further.

Why? Because the reflection collar takes Visible light to generate brighter spot (The HIR doesn't rub visible light), where the visible light could have been sent forward to being with.
I.e, it take a loss to make the gain where HIR doesn't take the loss of visible light to make the gain.

It is possible to make a small gain buy only reflect the lost light from aspherical set-up (The small angle of light to the side only), but then the reflection collar would need to have a much bigger center opening so not to block ANY light that could reach aspherical lens direct. the picture they shown look like it blocks part of the visible light that could have gone to a bigger aspheric lens to being with.

I don't know how much UV is emitted from LED, but if there's enough UV to worth recycle, then recycle only the UV light to generate into visible would make more sense.

The collar has to have a bigger diameter than the center opening, otherwise wouldn't be a reflection collar.
For example, If the set-up used a 2" diameter with 1" aspherical center lens, it'll throw further than the 1" Aspherical lens alone. However, it probably won't throw further than a 2" Aspherical lens alone w/o the collar. Until I see a side by side comparsion with Aspherical of the same diameter as the widest part of the collar, the data in that thread remains inconclusive & really didn't prove itself as being the longest throwing set-up.

Personally I would go with silicon glue as a cover rather than leaving it bare. I'm not sure where you would get the optical grade 2 part silicon glues the LED manufacturers use.

I agree, the benefit of the protection is worth it. I couldn't find a place
to buy small qty either. They only sell by the bucket loads. I read this article & requested samples from Down Corning, not sure I'll get it but doesn't hurt to try: :devil:
http://www.dowcorning.com/content/etronics/LED.asp
 
Last edited:

saabluster

Flashlight Enthusiast
Joined
Oct 31, 2006
Messages
3,736
Location
Garland Tx
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

But the fact that blue LED's are brighter with a dome is a bit of an enigma to me. Quite frankly, I don't get it.:confused:
It has to do with the fact that with a dome the escape surfaces are more perpendicular to more of the rays and are therefore more likely to escape than to reflect back and be absorbed. The larger the dome the more light gets through.(absorption losses aside)

However, I am not convinced the reflection collar would make a longer thrower. Michael has shown an image that the collar did make the die brighter, however I doubt that he could use this to make DEFT throw further.
I assure you it makes the DEFT throw even farther.;) Whether you are convinced or not does not change that fact.
 
Last edited:

ma_sha1

Flashlight Enthusiast
Joined
Jan 10, 2009
Messages
3,042
Location
CT, USA
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

I assure you it makes the DEFT throw even farther.;) Whether you are convinced or not does not change that fact.

I am glad to hear that!

Time for me to do some experiments.
 

saabluster

Flashlight Enthusiast
Joined
Oct 31, 2006
Messages
3,736
Location
Garland Tx
Re: Is it possible to de-doom SST & not to suffer Lumen Loss?

I am glad to hear that!

Time for me to do some experiments.
I normally don't worry much with the way people spell but I just thought it was worth mentioning that it is spelled "dome" or "de-domed". Doom is something altogether different. :thumbsup:
 

ma_sha1

Flashlight Enthusiast
Joined
Jan 10, 2009
Messages
3,042
Location
CT, USA
I run some test tonight on some home made collar, by cutting some plastic reflectors & place it up-side down.

I taped the MagDragonn Jr so the focus doesn't move, then open the Bezel & test a plastic DIY reflective collar with various openings:
dscf7324y.jpg


Narrow Opening:
dscf7309.jpg


Medium Opening:
dscf7310z.jpg


Wide opening
dscf7312s.jpg



The numbers below are not optimize for lux, just use it as relative reference to see if the collar makes a difference. :

Plastic OP reflective Collar Test: (various "openings" referring to hole size of the reflective collar opening)

Super narrow opening 29Klux
No Collar: 135K lux
Narrow Opening : 130K lux
Medium opening: 137K lux
Wide Opening:135 K lux
No Collar 134K lux
Wide Opening: 131K lux
No Collar: 135K lux
Wide Opening: 131K lux
No Collar: 131K lux

Average No Collar:................ 133,750 lux @ 1 meter
Average Wide Opening Collar:..132,330 lux @ 1 meter

Darn it, I was hoping for some easy lux here :mad:


Alright, try again. A Collar made of Mag Reflector SMO,
it looks a little shallower & I was hoping for better angle of reflection this time.

It Actually looks pretty cool.
dscf7318.jpg


dscf7322.jpg



I did the Medium opening, as from test #1 above, it's the only one that seemed to have a chance.

No Collar:........... 132K Lux
Medium Opening:..130K lux
No Collar:............130K Lux
Medium Opening:...130K lux

Each test was done twice, again, no luck here :hairpull:

Michael, may be you can show us some more details on your test?
maybe it's the dome that prevent light from penetrating back to the led?
It it only works for de-domed CBT-90, then it won't work for domed R2?
 

Th232

Flashlight Enthusiast
Joined
Dec 25, 2008
Messages
1,064
Location
Sydney, Australia
The collar is supposed to reflect light back onto the die of the LED, I suspect that your reflector isn't doing that, but rather reflecting it somewhere else. Think you need a reflector with a spherical cross-section?
 

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
The collar is supposed to reflect light back onto the die of the LED, I suspect that your reflector isn't doing that, but rather reflecting it somewhere else. Think you need a reflector with a spherical cross-section?
Thats what I was thinking, that at the focus point, a parabolic reflector would just send the light towards the back of the flashlight. Maybe a donut shaped mirror could reflect any light back at the parabolic reflector and to the die, but a spherical shaped reflector with the die at the focus could do the same thing in one step.
 
Top