Glowing Bezel 1

[bvanhiel]

I'm trying to create a glowing effect on an interface panel to signal user interaction. The part that glows is a clear plastic bezel that extends about .25" around a card reader slot.

The prototpye interface used a large (10+) array of bright (10,000mcd) LEDs shining through a diffuser onto a sand blasted acrylic part from the rear. While the brightness was good and the light even, I needed 2-3 inches of space to keep the LEDs from hot-spotting and to give them a line of sight around the card reader mechanism.

Instead of shining directly on the back surface I thought I might extend the clear material of the bezel and fit the LED around the edge. I could array them all pointing inward, or leave them attached to a PCB (preferable) parallel to the bezel and then "pipe" their light towards the viewable area of the bezel with heavy rounds.

I intend to polish all of the surfaces except those on the "backside" of the visible surface, so that the light can be reflected cleanly out the front.

A couple of specific questions:

1) Instead of texturing, will paint work? The advantage to us is that the color of the glow could be changed.

2) Will silver paint or plating on the polished areas have any effect? I'm assuming that they will actually be worse, because you no longer have an air-plastic interface, but it doesn't hurt to ask.

3) Since I want the light from the LED's to make an immediate right turn, am I better off with facets, or a heavy radius fillet.

Thanks,

-b

 

[IRstuff]

See, part of the problem is that you don't specify ALL of the problem. How bright does it need to be? How uniform does it need to be? What color does it need to be? What viewing angle does it need to support?

Why not use an EL panel? Other than the high voltage and possibly insufficient brightness, it runs cooler, and lasts longer.

Why not use cold-cathode backlights like those in PDAs?

The area is so small that a holographic diffuser could probably be used with direct illumination from a CC backlight.

TTFN

 

[patprimmer]

As already mentioned, and from the information provided, acrylic is an excellent material for your part. It can have higher transmission rates and lower haze than glass. It has a higher refractive index than glass, so it can be bent more at the interface.

I know nothing of light sources, so I will keep my comments to light pipes and diffusers.

Keep the surface the light enters the pipe perfectly smooth, polished, flat and perpendicular to the light source.

Keep all other surfaces of the pipe perfectly smooth, polished and flat unless you require the surface to glow.

To change colours, put a transparent coloured light filter between the source and the light pipe. This can also be acrylic.

To get light to emit, you can cut facets or you can frost the surface. Bayer published some literature around 1988 on the subject in relation to a product they called LISA. I still have some data in storage if you cannot find it elsewhere. Most major acrylic suppliers will also have some data.

I do not think paint or plating will help as it will tend to scatter the light and break up the parallel nature of the light in the light pipe.

I would think a 45 deg facet would turn the light 90 deg in a compact structure.

If the light source is monochromatic, you might get slightly better results.

You can make prototypes from acrylic sheet or by casting from monomer and cross linking.

You can injection mould for production if the numbers warrant it.



Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[bvanhiel]

How bright does it need to be? How uniform does it need to be? What color does it need to be? What viewing angle does it need to support?

I don't have a brightness spec, but the the area is meant to indicate an interaction point, so the user should be able to see it flashing in a well lighted room. Uniformity is a goal set forth by ID, but isn't a hard requirement. The viewing angle needs to be wide, as we want to attract the users attention as they approach the unit.

I've looked at EL, but my understanding was that the panel would only last around 2000-4000 hours of continuous use, this unit is expected to run 24/7 for years on end, hence my interest in an LED solution.

The cold cathodes is a possibility. Direct illumination with diffusion was essentially the concept for the prototype, it just seemed to take up a lot of space. The CC's would have an advantage as the light is already somewhat diffused compared to an LED.

My current preference for LEDs comes from the fact that it's easy to package a pcb with the LEDs in that area, and there won't be special power requirements.

patprimmer:

Will a filter work well with a white LED? My thought was to used colored paint only on the back of the glow area that would otherwise be textured. The LED in essense would illuminate the paint.

Initial production is low volume, so I may machine a prototype out of acrylic and then cast copies for the first 25 units.

Thanks,

b

 

[IRstuff]

You might peruse:

http://www.densitron.com/displays/products/?o=1&cat=4&id=17

Their white paper link is a pretty reasonable review of alternatives, although it's really unclear what the LED lifetime might be given the 24/7 duty cycle. Reliability is exponentially proportional to junction temperature. Therefore, a constant-on incurs a substantial reliability penalty.

I would recommend considering holographic diffusers:

http://www.poc.com/lsd/default.asp,

mainly because a true Lambertian diffuser imposes a considerable penalty on transmission, since light is diffused backward as well as forward.

Standard ATMs use just blinking LEDs, which I tend to favor, since the intensity is substantially brighter than with a diffuse source. The blinking allows the LED to run a cooler temperature for more reliability.

TTFN

 

[bvanhiel]

The area will blink to attract attention (or sinusoidally pulse if the ID guys get their way), so I think LEDs should have a long life. I've seen the POC page and have bookmarked it as a possible source of diffusion films, but the Densitron page was new to me. I will order some samples Monday. I probably can't use an off-the-shelf part since I need a hole in the middle, but it may help me design the custom one.

Thanks,

b

 

[patprimmer]

It makes no difference where the light is filtered, as only certain wavelengths will be removed and all others will transmit unaffected.

If you use paint, you run the risk of creating an uneven surface and diffusing the light. You do not want any diffusion until the light is exiting the light pipe.

If you do use paint it must be transparent, not translucent. It must stick to the surface without etching it.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[bvanhiel]

patprimmer,

From what I've read, the rough surface on an edge lit part should be opposite the exit surface, since most of the light will be reflected back through the part. That is where I was considering using a colored opaque paint.

-b

 

[MacGyverS2000]

EL sheets are a no go, IMHO... the sheet itself is cool, but the inverter isn't necessarily so, and the inverter takes up more room than a proper LED solution would. Lifetime is way too short compared to LEDs. You'll run into similar issues with lifetime on CCFLs.

If you wish to use paint, consider UV LEDs and luminescent paints... awesome look, even in the daytime.

White LEDs will only filter well if you use the RGB type. Most white LEDs use blue dice capped with a yellow-emitting phosphour. The spectrum of these LEDs creates some very odd effects with filters... it may work, depending upon the colors you want, but experimentation is in order.

High-efficiency LEDs can put out an amazing amount of light for low currents (<5mA), particularly for low viewing angles, which I suggest using for this application. I could provide better advice if you had a rough sketch of the pipe you want to use and rough dimensions.

A smooth radius bend is better than a 45 degree angle, but angles can have their uses. For example, a 90 degree bend just before the tip of a fiber, followed by the tip being mushroomed creates a nice wide-angle lens. It all depends on what you want to do. I suggest finding pre-made lightpipes for sale and experiment with a few. Some useful ideas may come out of what's already on the marketplace.


Dan - Owner

http://www.Hi-TecDesigns.com

 

[IRstuff]

"rough surface on an edge lit part should be opposite the exit surface, since most of the light will be reflected back through the part"

Not true, if you read Densitron's literature. There were some figures showing various approaches to uniformly spreading the light.

TTFN