Friday, March 1, 2013
System Design: Effective Avionics Backlighting
|Typical turn-key lighted panel design from Anodyne Electronics
Manufacturing (AEM), Kelowna, British Columbia, Canada, includes digital display,
tacticle switches and backlit rubber keyboard.
Backlit panels and controls are a critical part of cockpit design because they enable equipment settings and control positions to be easily determined at night. The idea of projecting light through a panel or knob to make markings visible in the dark is an amazingly clever and simple concept, but it has a very complex physics package and rich history behind it.
Instruments and legends were initially illuminated in aircraft via post lights (tiny masked lights often fed through instrument mounting holes), or via small spot illuminators around the pilot’s head or panel corners. Many tricks were employed; some quite ingenious, such as small UV floods and fluorescent paint markings, to get panel items to illuminate at night. Many of the pioneering early techniques were created and manufactured by Grimes in the 1930s, who also made external strobes, navigation lights and similar illumination inside and outside the aircraft. Grimes is now part of Honeywell, and these lighting products still live on, especially post and map lights.
Panel lamps and indicators, including the earliest “through-panel illuminators” and press-to-test types were mainly provided by Dialco, now called Dialight, and generally were built using ruggedized miniature T-1 ¾ midget flanged lamps like the ubiquitous 327 lamps from GE or Chicago Miniature Lamp (now CML Innovative Technologies). Many of these designs are incredibly sophisticated mechanically, such as dimmable press-to-test indicators, and remain in significant use in military and civil aircraft today. These individual cockpit indicators were eventually consolidated into warning panels and legend light assemblies by makers such as Korry (now Esterline), Master Specialties’ Roto-Tellite (now Eaton) and Aerospace Optics, which pioneered the transition of these indicators to true sunlight readable and light emitting diode (LED) status indicators in very small spaces.
These indicators were complimented by the co-development of two important technologies, trans-illuminated knobs (knobs that light up to show their position), and backlit panels to allow the letting of the legends themselves to illuminate. Lighted panel technology has become very mature and is covered in detail in MIL-P-7788, later MIL-DTL-7788, along with a special standard for Night Vision Compliant panels, MIL-STD-3009. These two standards provide a wealth of useful information for the design and fabrication of lighted panels and general illumination.
Colors have changed considerably over the years. Initial night panel lighting was mainly red (which has the lowest impact on the eye’s night vision externally). There is a famous story about red cockpit lighting that goes as follows: During World War II, British pilots had red cockpit lighting, and German pilots had white lighting. Because the red lighting had no impact on the British pilot’s night vision, their accuracy during night bombing was very good, while the Germans had difficulty resolving landmarks in the dark. This simple design item had a major impact on crew effectiveness. If captured, British pilots were advised to say they ate a lot of carrots to account for their better night vision. Red illumination remains a favorite of infantry, maritime and vehicular designers today, when night vision goggle (NVG) compatibility is not required.
The red illumination eventually lead to “aviation blue-white” to allow all instrument colors to be correctly seen, and to emphasize red as a warning only in the cockpit area. ANVIS/Night Vision compatible green (A or B) is the most recent trend to permit compatibility with Night Vision Goggles to improve the external night vision of the flight crew. In the interim, manufacturers have used many lamp and LED techniques, often covering the entire range of red-yellow-green-white for backlighting and panel indicators as the mood suited them, and as LED technology advanced. The use of yellow and red has been strongly disapproved by FAA and Transport Canada for anything other than warning indications, so the general trend has now moved firmly to white (blue-white) and green for most cockpit and legend illumination.
It is worth noting that colors have specific cockpit meanings, and all designers should be aware of them. Red is for serious danger only (or as one Eurocopter staffer commented during a design meeting I attended, “land now, even if in a schoolyard”) and yellow/amber is for alerts or cautions (land soon, but not in a schoolyard). Green is for normal status indication, blue or white for informational messages. Backlighting should be either aviation blue-white, (some special military tasks are still done in red to protect the eye’s natural night vision), or genuinely compatible night-vision-system green. (Red and IR emission must be suppressed there, not a simple task, even with LEDs). There are no standards for any other colors of backlighting, and equipment with other methods will certainly not blend in well with other instruments and systems in the cockpit.
How do you know if your lighting is even close to being correct, either too bright or too dim or even working over the correct range? As luck would have it, there is a very inexpensive way to quickly test this. Common lighting busses are 28V, 14V and 5V. The full lighting bus (maximum intensity) is that full value, and typical night flying is 50 percent of that value. If you use the ubiquitous 327 lamp as a reference for 28V lighting (330 lamps for 14V and 328 for 5V), and slip a aviation blue white or other AMP/Hexseal Silikrome cover on it, you now have a highly reliable “standard” for illumination, and its intensity should track your system over the bus range.
Light intensity tracking is also a critical element, and on a 28V dimmer bus, full control and illumination is required below 9V, which is a common NVG setting. Care is required in the design of lighting, especially when LEDs are series-connected, to make certain that series LED strings do not abruptly cut off visually before this threshold. This is a very common design error in LED illuminated backlighting, when driven right from the dimmer bus.
Many companies offer turn-key panel backlighting services to design and fabricate lighted panels, such as Canada-based Anodyne Electronics Manufacturing (AEM); Astronics, of East Aurora, N.Y.; Korry (Esterline), of Everett, Wash.; IDD Aerospace, of Redmond, Wash.; and others, and they can provide full capability for you. They can also include illuminated keys and keypads, switches and displays as part of the complete design.
Many cockpit lighting applications do not require full blown mil-spec panel fabrication, and often the task you have is to simply make a few buttons, knobs or legends illuminate on a piece of avionics equipment, this is what we are going to examine here. In our test case, we want to make some panel indicators and legends light up, and perhaps the position of a level control.
Should we use lamps or LEDs? This is not a trivial question, and both techniques have their own perils and benefits. The origin of all cockpit lighting is incandescent lighting, so “standards” all mimic this performance, and it is useful to understand the implications of this method. “Aviation blue-white” lamp illumination is typically the same as the 6500-8000 K white LED range, slightly blue in tint, or cold-white.
First, lamps have one enormous advantage, they are SPHERICAL emitters, meaning light is emitted essentially in every direction. This makes flooding the back of a lighted panel very simple, and reflected light reaches all areas easily, especially if the panel is coated white on the inside surface for reflection. In contrast, all LEDs are CONICAL emitters, outputting a relatively small cone of light, all basically forward, with almost no side pattern emission at all. This distinction means LEDs usually must directly flood the specific legend to be lit, and very little light can be counted on to reflect and fill in any other areas. This is a major mechanical issue in the panel design. LEDs emit near-coherent light, with all radiation forward, and even special concave lens designs provide mediocre fill results when compared to lamps. LED light is also very difficult to diffuse, and optical surface techniques that work well with lamps, usually fail with LEDs when trying to make even legend illumination with a single source.
In addition, the LED’s cone diameter of emission is directly proportional to the distance from the front surface (the farther away, the bigger the illuminated disk), so this means very little front area will be lit if panels with the standard 0.25 inch total thickness are made. Retracting the LED behind the front panel can offer a much wider lit area with the same number of LEDs, as even 0.1-inch of additional depth has a major impact on the flood diameter. In addition, many 3mm and 5mm conventional LEDs often have a black spot in the center when projected to a flat surface at short distances. This often shocks first time users, and has given rise to the term “dark-emitting diodes” in some circles.
Here’s a great tool that is almost free for checking your planned backlight scheme:
1. Position a piece of common printer paper at the font surface height above your LED pcb array (about 0.2 inch), the exact (and usually very disappointing) flood illumination pattern will appear on the paper.
2. Move the paper up or down to see the immediate impact on height. Note that small wide angle chip LEDs can actually illuminate a bigger area than larger Topled cavity reflector style parts, simply because their thinner size puts them farther away from the projection surface.
LED intensity is often a mathematical game for device makers, they can provide very high intensity (>5,000 millicandela), but often over a very tiny angle, sometimes as narrow as 10 degrees. For backlighting you want the highest possible angle (>140 degrees), and good intensity (>400 millicandela) to allow operation at reasonable currents (<20mA). Nothing beats physical testing however, and be aware that LEDs are “binned” in groups, and you need to be sure both color and intensity work out for you in the bin value you select. White is especially problematic in this regard, and even with careful device selection, you will probably have to re-work any large arrays to remove the odd outlier part that is too dim or is the wrong tint in a finished array. It is not uncommon for a panel array to contain a hundred LEDs for full illumination, which can be done just as well with only 2-4 lamps.
Lamps have their own failings, for example, they shift orange and red as voltage is reduced, which is why color-correcting pale blue filters are often used to keep the cool white appearance. They also have limited shock resistance and operating life, so unless you have carefully selected your lamps, and de-rated them a bit, you can get very brief lifetime, some types as little as 1,000 hours. The ubiquitous 327 lamp is only rated for 4,000 hours at 28VDC. Here is where de-rating is critical, reducing the lamp voltage to just 94 percent of full rated voltage can double operating life, with only a 16 percent loss of intensity according to the Standard Handbook of Electronic Engineering. A small amount of protective circuitry (even just a series diode or two) can yield important lifespan benefits in the cockpit.
The 387 is the long-life version of the 327, offering 10,000 hours at the same voltage and current. There is a slight drop in brightness, 0.30mscp versus 0.34mscp (mean spherical candle power). One of my personal favorites for panel lighting design is the (CM)7387 lamp, basically the same 28V bulb in an easy to attach bi-pin base, but with an amazing 25,000 hours of life. And the price is almost identical. Highly recommended by me if you wish to use lamps in your design.
Lamps can be a nuisance for NVG operation, as they are massive red and infra-red emitter. Use of some type of (costly) ANVIS filter is usually required such as those from WAMCO or Korry, to block the unwanted emission, and change to color to green. Unfortunately, military ANVIS filters are ITAR restricted items, which can be a big logistical and legal problem for many manufacturers to control. Only dual use or civil designated materials used for the lamp conversion are easily used.
LEDs are often used to create “NVG Compatible” green backlighting, and usually fail miserably during acceptance testing. Why? It is not well understood, but almost all green LEDs (other than InGaN based designs) are still significant IR emitters, thus causing the LEDs to instantly fail ANVIS testing. Don’t forget, NVG goggles amplify red-IR emission about 30-50,000 times, so what seems minor to you on the spec sheet is often catastrophic in final test. Supplemental IR and ANVIS filters can eliminate this easily if the correct emitters are selected.
Lighting those panel knobs is the final task. Only a few makers still offer them, and they are often hard to search for and locate, such as Korry (Esterline), and Electronic Hardware Corp. (EHC), of Farmingdale, N.Y. EHC has an especially wide range of small trans-illuminated control types that I often need.
Generally, the traditional way knobs (with a clear bottom) are backlit is to have a white island under the knob that is backlit from the lamp/led system below. This can be quite effective, and coupling efficiency into the knob bottom can be high as long as enough light can be projected to the white island. This can be hard to achieve with LEDs if the knob shaft is surrounded by mounting hardware at the chassis side, which precludes any LEDs from being fitted close to the shaft. Lamps can easily project into this area from far away, but not LEDs. Also, stacked or concentric knobs usually suffer from high coupling loss into the top knob because of physical obstructions. Beware of over-sized attachment hardware and set screws, as they can severely interfere with light propagation. I have been unpleasantly surprised to see how many serious optical errors exist in stock trans-illuminated designs. Always TEST any potential parts with the shaft present. Sometimes, you are just going to be very unhappy with the results.
Some knobs (slim pencil types), which may be pushed or pulled on and off, may pass right through the lighted panel, and are actually lighted by 90 degree reflected side light (horizontal light) from the panel. This works reasonably well with lamps, but is not especially good with forward facing LEDs. The use of special side LED or bottom flood knob designs may be required to get satisfactory lighting of this type of knob.
All of this technology lives behind the front panel surface itself, which is typically matte black, tan or gray, generally painted on top of a white underlay, and all made from a plexiglass/acrylic core. Don’t forget that any white area intended to be illuminated must be translucent, not opaque. This is a common error made with polycarbonate overlay techniques, most overlay vendors will naturally use opaque inks, but it must be translucent white, or nothing will actually light up. It is also very important to insure that black areas are truly opaque black, and free of pinholes or thin areas. Double striking the black is a good practice to insure this. Painted panels are then laser etched, chemically etched or engraved to reveal the white illuminated surface below.
So, a few key things to keep in mind to get you to the result you want in your design:
1. Even “green” LEDs emit IR and red light, so be careful in NVG designs.
2. Series LED strings need to still work down to 9V (at 28V).
3. Use a suitable range incandescent lamp as a quick reference to check any LED designs.
4. Always check a design with a paper test sheet at the intended height first to see if your areas are correctly flooded.
These lighted designs are complicated to implement well, and some experimentation and adjustment will usually be needed to give the even and well-lit appearance you want. Don’t be afraid to work with large series-parallel LED strings (which also have series resistors, and possibly diodes), they can work very well, and can give attractive illumination at low current if the parts are selected and spaced correctly. The panel of your equipment is all anyone usually sees, so it’s worth that extra effort to make it memorable in the best possible way.
Walter Shawlee 2 is the president of Sphere Research Corp. in West Kelowna, British Columbia, Canada, and a senior designer at Technisonic Industries. He can be reached at firstname.lastname@example.org.
To see a complete archive of Walter Shawlee 2’s System Design columns, visit www.aviationtoday.com/shawlee.