In my last column we discussed the struggle among avionics designers to make cockpit displays clear and fully readable, even when harsh sunlight shines on them. We talked of such problems as washed-out displays, legends disappearing in reflective glare, and the inevitable bad feedback from pilots who cannot read their graphical displays. This month we will examine an additional technique to help make sure pilots actually can read all of their displays: optical filtering.
If you have a text- or numerical-style light-emitting diode (LED) display, the visual problems can be complex. First, you need to find a bright enough display. This means giving it a quick "sunlight" test--shining an incandescent lamp directly onto the display--to ensure it is even suitable. The display probably will require heat sinking to support enough current for sunlight readability. You can plan for that by attaching a heat spreader directly to the package or board traces. It should be thermally ducted to the case by the shortest possible distance or internal heat buildup can become unacceptable after only brief operation. Heat is the deadly enemy of displays: it reduces light output in the short term and leads to premature failure in the long term, so plan for it early in the design.
The next critical step is to design a suitable optical filter. If the display is a single color, the task is much easier. All LEDs are essentially monochromatic by design but may have a production wavelength distribution that is as wide as 10 nanometers. This makes matching yellow and green displays difficult in a larger array because the range between the two colors is only 10 nanometers. Vendors will "bin" displays by wavelength, so you should use such a service to achieve uniform brightness and color wavelength.
You can then get a display filter that is centered on the correct wavelength and the ideal target value. Such a filter will prevent light from getting inside the display's front surface and inner structure to create disturbing reflections, and it will return the maximum possible light from the display itself. The result: a huge increase in contrast and minimal light loss.
Be careful not to spec the filter so that it is too narrow in optical bandwidth. Otherwise, considerable optical energy will be lost in transmission. You will need the best possible, lowest transmission loss value you can get at your LED wavelength.
Many people are astonished at how dramatically a suitable filter will improve display viewing. Removing light to improve visibility seems intuitively wrong. But as I pointed out in my last column, the key is to improve contrast and reduce glare, not always to add brightness. The eye sees by contrast and perceives a high-contrast display best, even though its optical energy may be reduced.
The optical filter also can include a circular polarizer, or special anti-glare treatment, on one or both sides. Polarizers keep surface glare and spurious reflections to a minimum. Unfortunately, many filter vendors have dropped out of small-volume markets, including aviation, making sourcing a real problem. The most impressive materials I have seen come from Astra Products (www.astraproducts.com/i_aremifil.htm), which makes an optical filter material called Clarex. I have used this material with excellent results, finding it at times so low in glare that the filter appears invisible.
If the display includes a wide range of colors--so that a tinted and optically selective filter is not possible--the only useful choice is a clear, neutral gray filter or a circular polarizer, coupled with the best possible anti-glare treatment. Additional treatments include scratch-resistant coatings (desirable) and conductive coatings to suppress radio frequency and electromagnetic interference (EMI). Note that EMI suppression coatings are only effective if they are uniformly edge-grounded--not a simple installation task to achieve. Fine black metallic mesh also is effective, and is widely used in the test equipment world.
If you use LEDs for either switch or legend illumination, try to avoid packages with an integral current limiting resistor and use wide-angle styles, rather than bright but narrow emitters. (Limiting resistors typically are used in series with the LED to keep the current to a safe value, but they generate considerable heat at high voltages.) The waste heat in such a package, especially at 28 volts DC, can be very high. This, in turn, reduces the light, making the quest for adequate illumination self-defeating.
If you have specific sunlight problems, drop me a line. I will try and suggest workable solutions.
Walter Shawlee 2 can be reached by e-mail at firstname.lastname@example.org.
This, regrettably, will be Walter's last column for the indefinite future. After many years of writing for Avionics Magazine, he is taking a break from his word processor--Editor.