Stray Side Emissions of Domed LEDs

[UnknownVT]

While doing Xnova AA vs Ultra-G, ArcAAA, Dorcy 1AAA -
I noted the very strong stray side emission of the Xnova AA

with strong "spikes" at approx the 150/30deg markings on the protractor.

If the light is held close to a surface a noticable halo/ring of light can be seen

This of course is not all that uncommon to 5mm domed LEDs -
what was surprising was that the 5 LEDs are completely recessed in the (holed) head (one cannot feel them when pressing against the head) and there is even a slight lip/hood to the front bezel

Other LED lights also exhibit this stray side emission and therefore the halo/ring of light

Of course what's surprising here is that the bare exposed LED of the Photon 2 seems to give less side stray emissions and hence the halo/ring of light than the well recessed LED of the 20 for $20 LEDs from CountyComm.

Even in the image of the coin-cell lights where I left them switched on - one can just make out those side-spikes on all three lights - despite 2 of the lights had well recessed LEDs

My speculation/guess is that these side stray emissions are due to the internal refraction/reflection in the clear plastic dome - and the light emanates from near the tip of the dome so any recess needs to be quite deep to mask off the stray side emissions.

The Nichia(?) LED used in the Photon 2 shows less severe stray side emissions than the (probably Chinese) LEDs used in the 20 for $20 and the Xnova AA - and by the characteristics and tint I would hazard a pure guess that these may be the same LEDs.

Just as an experiment I wanted to see how much masking/hooding it would take to eliminate the stray side emission of the Xnova AA.

I simply rolled some (heavy gauge) paper (a leaflet) snugly around the light to form a tube, and secured it with a rubberband

This allowed me to slide the light in and out in the tube to vary the hood/masking effect

fully masked/hooded to eliminate the stray side emission, hence halo/ring of light (at depth - about 5/16" or 8mm)

Not enough mask/hooding - but shows an interesting clover-like effect (at depth about 5/32" or 4mm)

a sort of compromise - perhaps "optimum" depth of about 1/4" or 6mm

Please let me know what you think -
and perhaps your thoughts on the causes of these stray side emissions?

 

[Lurker]

I think it is caused by the geometry of the epoxy lens/housing that encases the LED. The function of the domed top is to focus the light coming from the die and form it into a coherent beam of light. It does a pretty good job of that. Its limitation is that it is positioned and shaped to create that beam using only the light coming directly from the die and from no other sources. Some of the light probably leaves the die at a wide angle and hits the cylindrical sides of the epoxy enclosure and reflects up into the domed top. Approaching the domed lens from an odd angle like that results in an odd exit angle for these photons. I think that is where the ring of light comes from. Maybe some LEDs have more precicesly shaped epoxy enclosures or dies that minimize the light ring.

This would explain why it seems to be coming from the very tip of the LED, but it's just a theory.

 

[UnknownVT]

[ QUOTE ]
Lurker said:
I think it is caused by the geometry of the epoxy lens/housing that encases the LED.
<snip>Maybe some LEDs have more precicesly shaped epoxy enclosures or dies that minimize the light ring.

[/ QUOTE ]

Good food for thought - thank you.

Simple lenses are designed normally for a point source of light -
a theoretical point source of light would emit light equally in all directions.

A LED die may approximate closely to a point source -
but it is not, therefore there is light OUTside of the point.

The simple lens therefore cannot "perfectly" focus the light from the die -
but I wouldn't have thought the source of light was that wildly deviant from a point to cause light to want to go out so dramatically sideways from the lens?

 

[The_LED_Museum]

My "Xnova AA" does not have this obvious ring, so it could be from a different manufacturer than your "Xnova AA".

 

[UnknownVT]

[ QUOTE ]
The_LED_Museum said:
My "Xnova AA" does not have this obvious ring, so it could be from a different manufacturer than your "Xnova AA".

[/ QUOTE ]

cool - thanks Craig.....

you sure it wasn't your dog flooding it that stopped the side emissions? /ubbthreads/images/graemlins/icon15.gif /ubbthreads/images/graemlins/grin.gif

Yes, we are probably talking about different (but similar looking) lights - the runtime plot from Roy is quite significantly different - that it couldn't just be sample variation or batteries -
and there are enough cosmetic differences -
I've posted a pic of my Xnova AA and hotlinked your pic next to each other for comparison in this thread -
Run-Time Plot - Xnova AA

Stray Side Emissions -

It could be different LEDs which may have less side emissions - or perhaps deeper recess of the LEDs that have masked the stray light.

Some lights do need to be held very close (less than 1 inch) from a surface to see that halo/ring of light easily.

Most 5mm domed LEDs need to be recessed quite deeply to mask off these stray side emissions.

 

[RonM]

If you take a full exposed 5mm LED and mask the sides with electrical tape, but not the dome, do you still see the side emissions?

 

[UnknownVT]

[ QUOTE ]
RonM said:
If you take a full exposed 5mm LED and mask the sides with electrical tape, but not the dome, do you still see the side emissions?

[/ QUOTE ]

Definitely -
the dome needs to be well recessed/masked.

You can experiment yourself - by shining a LED like a Photon at 90deg to your line of sight - now slowly turn the light toward yourself - with not much a turn - you'll see the dome tip become a lot brighter than before - continue turning and that brightness diminishes until it's fully turned toward yourself........ That first bright spot is the stray side emission - halo/ring of light we're discussing.

Using the Jug-shot of the Photon 2 (in opening post)- one can see the peak side (spike) emissions are at approx 20deg markings on the protractor - which means the LED has to be fully recessed below at least
2.5mm * Tan(20) = 0.9mm - close to 1mm = ~1/64".

The Chinese LEDs on the 20 for $20 are side peaked at about 30deg - so the depth of recess has to be at least -
2.5mm * Tan(30) = ~1.5mm = ~1/16"

Those are for a tube that's tightly fit to the 5mm LED - wider tubes will obviously need more depth as my experiment with the 5LED Xnova AA showed.

That's why both the 20 for $20 LEDs and the Inova Microlights despite being recessed still show this stray side emission. The 20 for $20 LED requires more masking - and they just show the tip of the dome so that's why the stray side emissions are so severe.

On the Xnova AA - despite the 5 LEDs being fully recessed - the depth is just not enough - hence the very severe side-emission.....
a design that obviously was NOT tested out in practice.

 

[Lurker]

The stray side emissions are not necessarily a bad thing. For instance, if you are using the light to walk through a dark room with the light pointing out in front of you, the side emissions may put some light on the ground immediately in front of your feet and also illuminate the ceiling and walls which will contribute to the overall illumination of the room.

I would rather have the side emissions reflected forward into the primary beam, of course, but maybe they are not such a bad thing (depending on the situation).

 

[idleprocess]

In a recessed-emitter light such as the X5-lookalike, odds are that the rings at the edge of the beam are reflections from the bit of sidespill that strikes the recess.

 

[UnknownVT]

[ QUOTE ]
idleprocess said:
In a recessed-emitter light such as the X5-lookalike, odds are that the rings at the edge of the beam are reflections from the bit of sidespill that strikes the recess.

[/ QUOTE ]

That's a good thought (I also thought of this)

But on my Xnova AA (X5 lookalike) I can definitely see the LEDs' side emission directly and not via any reflection from the recess.

The photo of the Xnova AA with the paper rolled tube kind of shows that - those LEDs are very bright and rival the close flash of the camera - because they are turned just at probably the worst angle.

I don't disagree with Lurker about the side emission being still usable light - I said as much in my comparison Xnova AA vs Ultra-G, ArcAAA, Dorcy 1AAA

But those side emissions are way out on the perimeter of the overall side-spill (around 120-130degs) which is much wider than one's normal view even with the eyes' normal moving/scanning.

The very bright side-emission is distracting and will attract attention - neither of which are desirable - and these disadvantages far out-weigh any advantage of extra usable light way out at the perimeter - for me.

The masked versions (illustrated in the opening post) either fully masked or what I called "optimum" would work a lot better for me.

I may well think about building myself a more durable 6mm depth hood for the light.

Of course, someone may actually like the strong side -emissions - I don't -

as always, YMMV

 

[Negeltu]

I looked at some 5mm leds while lit up... and sure enough if you look at the led from the side... and slowly turn it towards you while looking at the dome only...you will see the dome get very bright way before it's pointed directly at you. Some 3mm leds I looked at had even brighter side spill coming from the tip of the dome.

 

[Roy]

Yes....on the Xnova AA if you view from the edge you see no spill.....roate a bit and you'ww get a flash from the tips....rotate some more and the light dims....then as you start getting head on, you get the main beam. The spill light on mine is such a wide angle, you can hardly see it.

 

[idleprocess]

Sounds like it could be internal reflection.

The reflectors and "optics" of through-hole LEDs aren't usually of exceptional quality.

 

[UnknownVT]

[ QUOTE ]
idleprocess said:
Sounds like it could be internal reflection.

The reflectors and "optics" of through-hole LEDs aren't usually of exceptional quality.

[/ QUOTE ]

yes... internal reflection/refraction was also my guess in the opening post.

There are NO separate "optics" used in this Xnova AA -
just the domes themselves......
and it's the tip of the domes that emit this stray side light.

However I was very impressed on how the 5 LEDs were aligned - so that one could not tell there were separate/discrete LEDs making the smooth and bright central hot-spot - as the beam shot hopfully shows -
see more beamshots and photos at -

Xnova AA vs Ultra-G, ArcAAA, Dorcy 1AAA

 

[greenlight]

Is it possible that over-driving the leds causes the sidespill?
--------------------
reducing infinity glare

 

[UnknownVT]

[ QUOTE ]
greenlight said:
Is it possible that over-driving the leds causes the sidespill?

[/ QUOTE ]

Nope.......

The stray extraneous emissions are from the tip of the dome - which I believe are caused by internal reflections/refractions by the plastic dome, and not directly from the emitter itself.

Overdriving the LEDs (or not) should not make a difference -other than the obvious proportional change in intensity.

 

[Niteowl]

This is the first time in my short time here that I've seen this refraction phenomenom discussed. I myself do not care for the halo effect, it's distracting.

I have a Peak 3LED with the knurled round head(proto). The LEDs are flush with the tip and have the HALO effect big time. Love the light but for that. It's almost enough for me to get rid of it. A Peak 1LED McKinley was returned for that reason. My Infinities have it to a much lesser extent, the LEDs are recessed further.

Mark