Aircraft Navigation Light
Manufacturer: Aero LEDs
Product Name: Pulsar EXP
Description: LED Navigation and Anti-collision Light
Category: Patent Application
USPTO Number: 12/508,450 (Click on the number to view the publication)
Technologies: LED, Thermal Foldback, Tertiary optics, Secondary Optics, Buck/Boost circuitry.
Patent / Product Analysis:
The first thing that stands out when the patent application is inspected is the reference to a 400cd ant-collision light. Truth is, this light has gives out only 150 effective candela (refer installation guide). Secondly, the design is in such a manner that meeting the intensity pattern in compliance with FAR 23.1401 or FAR 25.1401 (as described in the patent application) seems tough. Even more appalling is the fact that the installation manual clearly mentions ” Light output exceeds 70 candela for position lights and 150 effective candela for strobe light, not intended for use on certified aircraft”, and yet, the Aero LEDs website claims that the Pulsar EXP is in accordance with FAR 23.1401 (See the claim here).
The design uses in total 18 LEDs; 2 for the forward position light (red for the left wing and green for the right wing), 2 for the rear position light, and the remaining 14 for the strobe section of this light. The LEDS used are the Philips Luxeon Rebel series, particularly the Luxeon Rebel White for the strobe/rear position light, and the Luxeon Rebel Color for the position lights. This has been determined through inspection of the LED package used in the drawings contained in the patent application.
Aero LEDs’ Pulsar EXP’s installation manual mentions that the strobe consumes ” 2.25A at 14V for 0.33 seconds”. Yet, the video shows a different flash pattern. Maybe the flashing sequence lasts for 0.33 seconds.
Going by the peak current of 2.25A @ 14V, the power during the strobe’s operation is a peak 31.5W. Assuming a 90% driver efficiency (optimistic value), the power reaching the 14 strobe LEDs is 28.35W. This turns out to 2.025W per LED. Inspecting the Rebel Cool White Datasheet (See LED Datasheet here), the Voltage across the LED is typically 3V for a current of 700mA. 3V X 0.7A = 2.1W. This means that Aero LEDs drives 700mA through their LEDs. At 700mA, the luminous flux is 235 lumens, for the brightest cool white LED. Thus, 14LEDs totally produce 3290lumens. Most of this light reflects off a reflector (Secondary Optics; assumed 90% reflectivity), and all the light passes through a diffuser / refractive glass (Tertiary optics, ~80% transmission). In all, the light output may fall to only 2368 lumens, which may be enough for a 150cd light, but not for a 400 cd LED based light with a 180° of horizontal coverage.
The colored position light reflectors seem to be inappropriate in their design. If one examines the reflector closely, one may observe that the design is such that it will prevent colored light from being observed at all angles in vertical planes that are at definite horizontal angles between 0° and 110°.
Electrically, Aero LEDs have been innovative. I particularly like how they have used a single PCB to mount the electronic driver circuit, and the LEDs. Just one double sided board, with an unknown number of copper layers. Although expensive, their idea of having the base devoid of metal wherever the inductors/ capacitors on the lower board are populated is very nice. Also well implemented is the thermal fold back mechanism. It uses the Linear Dimming (LD) pin of the HV9910 IC from Supertex (See datasheet here). Normally, the IC uses a 250mV internal reference to compare with the voltage generated across the current sense resistor attached to the source of the MOSFET, to sense the current through the LEDs. The Linear Dimming pin allows for this reference volateg to be changed, between 0V to 250mV. As a result, the current through the LED reduces. The HV9910 datasheet clearly mentions,” a potentiometer connected between VDD and ground can program the control voltage at the CS pin”. Aero LEDs have placed a voltage divider between VDD and Ground, with the divided voltage sent to the LD pin. The lower resistor of the divider, a 10K (which has been printed in the patent application as 1 OK), has a negative thermal coefficient. As the temperature rises, the resistance drops, and hence the voltage as the LD pin drops. Voltages above 250mV at the LD pin are ignored. However, as the temperature rises and the voltage drops below 250mV, the internal reference reduces accordingly, reducing the current through the LEDs.
But the current sense resistor seems to be 0.27Ω, which means that the current through the LEDs is 250mV/0.27Ω = 0.925A. Observe the patent application’s figure 7 carefully. It shows 4 LEDs, 2 in each string along with a 0.2Ω resistor, each string placed electrically parallel to the other. This would allow 925mA/2 = 462mA through each LED, assuming that the LEDs are identical. Since the “PWMD” pin is connected to a control circuitry, we may be led to believing that the circuit is for the strobe lights. But 4 lights can also lead us to believe that the LEDs are part of the position light system. IN which case, 2 will be the colour (red/green) LEDs, and 2 will be the white LEDs. Paralleling 2 different types of LEDs is not advisable, as the current distribution will be far from equal. However, if each string has one white and one colour, then the situation is more under control. However, this may mean more complex PCB routing.
The schematic in Fig 7 of the patent application shows the timing resistor connected to the gate of the MOSFET. This must have been a drawing error as this IC needs the other end of the timing resistor to be connected to the ground.
Conclusions: Innovative, simple, but does not meet FAR part 23 intensity requirements. Optical design is disappointing.