Sensors are proliferating in vehicle electronics as OEMs automate more systems to improve safety, comfort, and convenience. To address the usual conflicts among available space, styling, functionality, and price, sensor developers are combining previously independent sensor functions into a single device. A prime example of this integration is the way a solar sensor used by a vehicle's climate control system has been married with a twilight sensor that assists in automating a vehicle's lighting systems. Both are linear current sources that correlate to the relative light intensity.

Solar Sensor Design

To maintain passenger comfort, a vehicle's automatic climate control system uses solar sensors to measure the energy from the sun that enters the passenger compartment. Solar radiation can be significant--approximately 2400 W in full sunlight on a clear day--and is generally felt by occupants long before temperature sensors detect any change. Because solar sensors respond immediately to changes in light intensity, their signals enable the climate control system to automatically compensate for the variability of solar radiation due to changing environmental conditions. For example, a cloud occludes the sun or the vehicle enters the shadow cast by a large building.

Figure 1. First-generation solar sensors cannot see the sun at low angles of elevation and consequently are poor models of solar radiation. Second-generation technology corrects this deficiency and enables accurate measurement of solar heating effects in the vehicle, providing the climate control system with the precise input required to maintain maximum occupant comfort.

A solar sensor is essentially a current source with current (output) increasing linearly proportional to the intensity of the light shining on the sensor, but not all varieties are the same. To enable the precise climate compensation that consumers demand, a first-rate solar sensor must offer:

• Linear response function with light intensity
• Horizon-to-horizon visibility
• Customizable angular response
• Tight signal tolerances
• Low device cost
• Visible light spectral response
• Higher current output

As solar sensor technology evolved, distinct generations of devices emerged, all of which are still available.

First-Generation Solar Sensors

First-generation solar sensors are little more than standard component photodiodes positioned on the instrument panel for good exposure to sunlight. The photodiodes were initially placed under a clear plastic cover, but the effect on the vehicle's appearance was less than desirable. A cosmetic improvement was made by housing the photodiodes in plastic packaging that includes a cover made of a material that appears black to the human eye but transmits nearly 100% of all wavelengths above 700 nm. Because standard photodiodes have a peak spectral response near 950 nm, the sensor output is minimally affected by this material and the unit's appearance is dramatically improved.

First-generation devices have significant shortcomings. Because of their fixed angular response, they cannot model the solar energy entering the vehicle. This response function is driven by the half angle of the photodiode and the packaging that contains it. As a result, first-generation solar sensors have a peak sensor output when the sun is overhead and a minimum output when the sun is on the horizon. Figure 1 shows typical angular response curves of first- and second-generation devices.

Actual in-vehicle measurements are essential in determining an ideal sensor response. By measuring the solar energy entering the vehicle for various elevations of the sun and normalizing the data to the maximum energy reading obtained during test, an ideal sensor response characteristic can be developed for a specific vehicle (see Table 1).

Ideal Solar Sensor Response to Actual Heating Effects
	50°	60°	70°	80°
Ideal solar response (Normalized to unity at =50)	1.00	0.97	0.93	0.83
First-generation device	0.5	0.38	0.22	0.16
Deviation from ideal	-50.0%	-60.8%	-76.3%	-80.7%
Second-generation device	1.00	0.96	0.92	0.82
Deviation from ideal	0.0%	-1.0%	-1.1%	-1.2%
=0° is the sun in overhead position

As you can see, first-generation solar sensors deviate greatly from the ideal response for lower elevations of the sun. This is a significant problem because the sun is in the low-elevation region of the sky for the greater part of the day. At the equator, the sun is low for 50% of the daylight hours, but as you move away from the equator, the sun spends a proportionately greater amount of the time in the low-elevation regions.

Second-Generation Solar Sensors

Second-generation devices model the solar heating effects very well because the addition of a diffuser allows them to see the sun at low angles. The diffuser is typically a hemispherical dome made of a material that acts as a neutral density filter that also scatters light. The photodiode is positioned to have a clear view of the inside of the diffuser's dome. When sunlight entering at a low angle strikes the diffuser, the diffused light seen by the photodiode provides an increased output for these angles (see Figure 2).

Figure 2. Solar and twilight sensors are available in five configurations. First-generation technology (A) is little more than a bare photodiode in a cosmetic covering. Second-generation devices (B) add a diffuser to the assembly, improving accuracy in modeling solar heating effects in the vehicle and providing the sensor with a view of the entire sky. This technology can be customized (C), refining the sensor output to match particular vehicle requirements where a reduced output may be needed. (Further definition of this development can be seen in Figure 4). The use of chip-on- board technology (D) reduces package size, lowers material cost, and allows active mechanical calibration, resulting in a third-generation technology with improved precision. The use of a custom ASIC in place of a photodiode (E) allows a single sensor to have multiple functions that save valuable real estate in the vehicle and give OEMs greater design flexibility.

By customizing the sensor output, you can achieve an even more accurate model for solar heating effects. This step is regularly required because a sensor's ideal response is influenced by the vehicle in which the device will be used. Some of the most common ways to customize the sensor output are to:

• Vary the aspect ratio of the diffuser
• Vary the diffuser wall thickness
• Direct light away from some portions of the diffuser

The third option is sometimes achieved by adding a part to the assembly, but the most cost-effective solution is to modify an existing component (see Figure 2C). The modified sensor cap produces a response curve similar to that in Figure 3, generally desirable for vehicles with a large roof area. When the sun is overhead, the roof shades the passenger compartment and solar energy entering the vehicle is reduced. For precision in climate system compensation, the sensor output must be reduced accordingly.

Third-Generation Solar Sensors

First- and second-generation solar sensors generally have a part-to-part accuracy of ±20%, derived by summing the tolerances of the components. To achieve this tolerance, the parts are generally screened with less than perfect yield in the manufacturing process. Third-generation technology adds active calibration for tighter sensor signal tolerances, at the same time reducing size and costs.

To calibrate the sensor output, a design was developed that permits a relative movement between the diffuser surface and the photodiode. As the diffuser is moved closer to the photodiode, the sensor signals increase for a fixed level of illumination on the sensor; move the diffuser farther away and the signals decrease. This adjustment is made in the manufacturing process and the photodiode/diffuser assembly is held securely in position by the sensor packaging after calibration. The results are improved accuracy and higher production yields. Third-generation devices can be specified to an accuracy of ±10%.

Figure 3.Second-generation solar sensor technology allows the sensor's angular response to be customized to match the solar heating effects in a particular vehicle. Localized reduction of sensor response has been found advantageous for vehicles with large roof areas that shade the occupants from the sun when it is directly overhead. This reduced response is a function of reflectors molded into the sensor cap as shown in Figure 2C.

The photodiodes of first- and second-generation solar sensors are prepackaged silicon photodiodes. Prepackaging usually takes the form of either a metal can or molded plastic with leads extending axially for traditional through-hole mounting on a PCB. Because of available package sizes, costs, and attachment methods, however, prepackaging is not conducive to active mechanical calibration. Third-generation designs directly mount the silicon photodiode chip, wire-bonding it to the sensor's PCB. This specialized chip-on-board (COB) packaging makes for lower raw material costs and smaller package sizes while facilitating the calibration process. The resulting reduced package size addresses the design and styling constraints of the OEMs.

Third-generation technology is widely deployed and is considered the best available today for most applications. Some OEMs are finding it useful to have a separate solar sensor signal for various regions of the passenger compartment: left- and right-side independent climate control or front and rear comfort zones. Third-generation technology is easily applied to these multizone climate systems because it provides multiple solar sensor outputs from a single, customizable sensor package.

Fourth-Generation Solar Sensors

Fourth-generation technology will soon replace the three previous iterations as advances in automotive glass continue. OEMs are beginning to use glass that reduces--by as much as 50%--the transmission of IR wavelengths from 650 nm and above. The signal from conventional solar sensors will be significantly attenuated because these devices use photodiodes with a peak response in the near-IR region. There is no similar effect on visible spectrum transmission, however. The advantage of making a solar radiation detector sensitive to visible light is therefore apparent. The peak wavelengths entering a vehicle are in the visible spectrum. Maximizing the sensor signal calls for a peak sensor response that matches the peak spectrum present in the environment. With a solar sensor having visible-light sensitivity, the climate control system remains unaffected by the various glass options that may be available for a particular vehicle. Some would argue that a solar sensor should respond mostly to IR light, holding that IR wavelengths are the cause of heating. However, all wavelengths entering a vehicle help create a "greenhouse effect." Light energy absorbed by objects inside the vehicle is re-radiated at longer wavelengths that remain trapped in the vehicle.

Figure 4.Twilight sensors incorporating a diffuser made from proprietary band-pass optical filter resin and a custom ASIC photodiode provide a response that closely matches that of the human eye. The result is a sensor that enables precision automation of vehicle lighting decisions including control of headlamp, convenience lighting, and illuminated gauge display functions.

Fourth-generation solar sensors are relatively new and are an extension of twilight sensor technology that makes visible-light spectral response and higher sensor currents possible for solar sensors.

Twilight Sensors

Automatic lighting systems use a twilight sensor's signal to control a vehicle's interior and exterior lighting. Common uses include automatically switching headlamps at dawn and dusk and automatic illumination control for instrument panels and back-lighted gauge displays.

Twilight sensors interact with a vehicle's lighting system in much the same way that solar sensors interact with the climate control system. They provide a linear output current proportional to the intensity of the light detected by the sensor but their range of functionality is much narrower. Twilight sensors function in low-light conditions with a linear output spanning the range from total darkness to ~2000 lux. (For comparison, the sun's full intensity on a clear day is near 130 K lux.)

Early twilight sensors were based on a cadmium-sulfide technology. The cost was low but there were problems. Their response was not truly linear and their hysteresis resulted in a sensor with two different response functions, one for increasing illumination conditions and a separate one for decreasing light. Add a high temperature coefficient, and you had a device with a very wide tolerance--too wide, in fact, to provide the precise sensor input that modern automatic lighting systems require. In addition, cadmium is a substance that manufacturers and OEMs would rather avoid.

The first attempt at a better solution was to use the same approach as a solar sensor. These twilight sensors were little more than amplified photodiodes positioned on the instrument panel surface where they received good sunlight exposure. Because of the small range of light intensity over which a twilight sensor must function, designers added discrete component amplifiers to produce a signal in the target illumination range. The higher current output device resulting from the amplified signal helped prevent the connector contact corrosion seen with some low-current solar sensor applications. The drawbacks of adding the amplifiers were increased costs and package size.

Mimicking the Human Eye

The best twilight sensor incorporates all of the advances of the third-generation solar sensors. It uses the linear response provided by photodiode technology, the horizon-to-horizon visibility provided by a diffuser (molded from a proprietary resin that transmits only the visible-light wavelengths while absorbing all other light), tight signal tolerances provided by individual mechanical calibration, and small package size provided by COB assembly. But for precise control of vehicle lighting systems, a twilight sensor needs yet more.

Figure 5.The custom ASIC uses an amplified photodiode designed to have a spectral response shifted more to the visible spectrum than a conventional photodiode provides. This characteristic is present in both the twilight and the combined solar and twilight function ASICs. Visible light sensitivity enables peak sensor signals to remain unaffected by the use of glass that blocks a large amount of IR radiation.

The goal of lighting control systems is to make adjustments in much the same way as a human operator would. Because human eyesight is based on visible-light sensitivity, the ideal sensor would be responsive to the same range of wavelengths as the human eye. A model of human eyesight developed by the The International Commission on Illumination (CIE), based in France, has become the international standard. The human eye, it posits, has a peak sensitivity near 550 nm with a range of ~400 nm to 700 nm, and is the target spectral response for the twilight sensor (see Figure 4). Actually making this sensor required:

• An engineered resin that transmits only the visible-light wavelengths
• A custom photodiode ASIC with a peak response shifted toward the visible spectrum

Sensor developers used the engineered resin in the diffuser and substituted a custom ASIC for the third-generation sensor's photodiode. As a part of the ASIC development, higher current outputs became possible without the use of external amplification, saving package space and cost. The ASIC incorporates a series of integral amplifiers that produce a peak signal current of 2.5 mA at a light level of ~2000 lux.

The new twilight sensor has seven key features:

• Visible light spectral response
• Linear response at ambient light levels
• Horizon-to-horizon visibility
• Tight signal tolerances
• Low cost
• Smaller package size
• Higher current output

This list is nearly identical to that for an ideal solar sensor. Combining solar and twilight technology into a single sensor yielded the best of both.

Combined Functions

The twilight sensor ASIC led to an improved chip that features a single photosensitive area with a peak performance shifted toward the visible spectrum (see Figure 5). The amplification is broken down into two distinct stages. The first provides the range of sensitivity a solar sensor requires; the second provides the twilight range. A signal output is provided after each amplification stage. Because a single photosensitive area is used for two distinct signals, size reduction is significant.

The combination of solar and twilight sensor functions into a single device has many benefits for OEMs, the most significant being improved solar sensor function for use with high-IR-blocking glass. The unit occupies approximately half the real estate required by two separate sensors, meaning greater flexibility for interior design and styling. OEMs no longer need to be concerned about sourcing, mounting, styling, wiring, and connecting multiple devices--which translates to reduced costs.

Further Integration

The technology can integrate other functions as well. For example, a multiple-zone solar sensor can be combined with a twilight function in a single package. Other candidates for integration include:

• An LED to indicate that an alarm system is armed or to visually confirm that the doors are locked
• Temperature-sensing functions
• A humidity function or other optical-based moisture-sensing technology

As combinations proliferate and multifunction sensing becomes more mature, sensors will incorporate microprocessors and be connected via serial communication buses with the systems they serve. Software will enable further refinement in signal tolerances.

But even in these future developments, the task remains the same: to preserve the original sensors' individual functionality and create a synergy that moves sensing technology forward in response to OEMs' more exacting requirements.

Home | About Us | Site Index | Privacy Policy © 2006 Questex Media Group, Inc.. All rights reserved. Reproduction in whole or in part is prohibited. Please send any technical comments or questions to our webmaster.