Talon


Talon Aerospace  TAE0328-1

Manufacturer: Talon Aerospace

Product Name: TAE0328-1

Description: LED Red Anti-collision light

Category: Patent Application

USPTO Number: 12/183,999 (Click on the number to view the publication)

Technologies: LED, 115V 400Hz Unit with integrated power supply, Switched Mode Power Supply, 

Patent / Product Analysis:

1. Corresponding Product

2. Product Dimensions and Weight

3. Optics

4. Thermal

5. Electrical

6. Market Information

7. Background to the Invention

8. Conclusion

1. Corresponding Product

This patent covers Talon’s product TAE0328-1 (Click to see the product page). This is a red anti-collision light intended for large commercial airplanes, such as those manufactured by Boeing, Airbus, and the erstwhile McDonnel Douglas (Now part of Boeing). This anti collision light flashes at a rate between 45 to 60 flashes a minute. The light is intended for the fuselage of the aircraft. While one unit is installed on the top of the fuselage of the aircraft, another unit is intended to be installed below the fuselage, typically under the belly. TAE0328-1, as per Talon’s Website, is certified to make its way onto the following platforms:

Airbus: A300,A310,A318,A319,A320,A321,A330,A340.

Boeing: 757, 767

McDonnel Douglas: MD10, MD11, DC10

Talon’s RED Anti collision light mounted under the belly of an Airbus A320 aircraft.

2. Product Dimensions and Weight

Length : ~6″

Outermost Diameter : 4.5″

Maximum Diameter facing the airflow: 2.7″

Height of glass in the airflow: 2.5″

Weight: 0.9kg Typical without adapter plates.

Approximate dimesnions of Talon’s TAE0328-1

3. Optics

a. LED

TAE0328-1 Consists of 4 circular printed circuit boards (PCBs), each circular PCB having mounted on its periphery 24 LEDs. Above each board is a solid reflector, whose surface is that of a circularly extruded semi parabola.

Reflector Board with 24 LEDs (Left), ISO view of reflector (centre), Cross section of the reflector-board assembly (Right)

In total, the anti-collision lights consists of 96 LEDs.

Each LED is a Philips Luxeon REBEL colour LED, of part number LXML-PD01-0050 (Red-Orange), which has a luminous flux of 100lumens at a drive current of 700mA, and 50lumens at 350mA. This LED is found in Philips datasheet DS56.

Surprisingly, while the rest of the industry try to squeeze the last lumen from the best LEDs in the market, Talon aerospace has gone the other way by running their LEDs at 250mA.This is 36% of the maximum drive current of the REBEL Red/Red-orange LEDs.

For this LED, at a drive current of 250mA, the luminous flux stands at 0.718 times that at 350mA. This would make the LED give out 0.718X50 = 36lumens at a junction temperature of 25°C.

Flux-Current relationship for the Philips REBEL Red-Orange LED

In total, there are 4 boards in the anti-collision light, each board having 24LEDs. In total, the luminous flux from all the LEDs stands at 3,456lumens. With roughly 30% light from every LED getting reflected, and almost all reflected light passing through the glass, we can assume that the total amount of light given out by the anti-collision light is:

((0.3*3456)*0.85+0.7*3456)*0.9 = 2970 lm (85% reflectivity, 90% transmittance).

b. Reflector

The reflectors described in the patent are so deep that they do not allow light from the LEDs to exit at 75°. The FAA requirement for the 400cd anti-collision light system is to have 20cd of effective intensity between the angles 30° and 75° vertical (above/below the horizontal, as applicable). However, both the types of reflectors used in the invention, prevent light from exiting beyond 60° for the deeper reflector and beyond 63.5° for the truncated version of the deep reflector. If at all light has to exit at those angles, it may be due to refraction from the lens, but that is not described in the patent as it is evidently not part of the intended optic system.

Cross sections of the Light how that no light goes above 60° for one type of reflector and 63.5° for another type of reflector.

c. Performance

On Talon’s Flickr photostream, it is mentioned that the anti-collision light exceeds FAA minimum requirements by 25%. That would result in an effective intensity of 500cd. While this may seem little in front of Xenon flashtube based fuselage lights that have been on aircraft since decades, Talon mentions that the lights are “noticeably brighter”.

With 500cd minimum from the light, the unit’s light output can degrade to 80% before it fails to meet FAA minimum requirements. When one has a look at The Philips datasheet which graphs the light output versus temperature, you will notice that the light output falls, at 50°C to 80% of the value at 25°C. So if the light intensity was measured at a LED thermal pad temperature of 25°C, then on hot days when the unit heats up to 70°C on the ground, the light output will be below FAR minimum specifications.

Degradation of light output of the Red-Orange LED with temperature

d. Life

DS68, the latest Philips datasheet for the REBEL color LEDs states that, “Red, red-orange and amber LUXEON Rebel color products will also deliver, on average, 70% lumen maintenance (B50, L70) at 50,000 hours of operation at a forward current of 350 mA and is based on constant current operation with junction temperature maintained at or below 110°C.”

As claimed by Talon, the unit is qualified to meet FAR requirements following 20,000hrs of life. However, it must be borne in mind that as the unit ages in operation, the light output degrades, and the light will fail to meet the FAR requirements at temperatures between 25°C and 50°C.

4. Thermal

The fist glaring description in the invention is the temperature operating range : The power supply is designed to operate between -50°F to 130°F (Patent application paragraph [0039]). This would result in an operating temperature between -10°C and 54°C. This would mean that on a hot day the power supply would fail (a unit can easily heat up to 70°C on a hot day, and conduct the heat down into the electronics compartment). This would also mean that when flying at 35,000ft, where the temperatures drop to -55°C, the power supply will fail. In short, no guaranteed anti-collision light operation on the ground and in the air.

Talon has kept the drive current through the LEDs extremely low. With each LED dropping 2.022V at 250mA at 25°C, the peak power dissipated by each LED turns out to be 0.5W. Multiplying this by 24 LEDs, each board dissipates 12W peak. LEDs are flashed for 250mS, and when running at a flash rate of 60fpm, the average power dissipation per board is only 3W. With the reflectors upon which the LED boards sit serving a heat sinks, it may seem that the heat from the LED boards is extracted. However, there is no described connection between the different reflectors in the stacks, from a thermal perspective. A reflector placed on an FR4 board is effectively isolated from the other boards, allowing for no heat transfer to the rest of the anti-collision light components.

5. Electrical

Circuitry. Note how one MOSFET drives 2 strings. IN total there are 8 strings, requiring 4 MOSFETS. Each string has a resistor between the anode and ground.

115V AC 400Hz passes through an EMI filter before entering a rectifier. This rectifier converts 115V AC to ~115V DC. This smoothened 115V DC is applied to a 100kHz transformer, through a switch that turns on and off at 100kHz. The secondary side of the 100kHz step down transformer is rectified and smoothed by a capacitor to deliver 40VDC. How the 40VDC is maintained is not mentioned, but one possibility is through PWM control  on the switching side.

Each circular board has 24 LEDs, which are divided into two strings of 12 LEDs each. In all, the anti-collision light has 8 strings of LEDs. Each string has a current limiting resistor on the anode end and connected to ground on the cathode end. Current flows through the LED string when the gate to the MOSEFT is provided with sufficient voltage. The MOSFET essentially shorts the cathode end of the string to the circuit ground, thereby allowing the 40VDC to be applied across the 12 LED + resistor string.

But what’s surprising here is the location of the MOSFET, and the control voltage applied at the gate. The Microcontroller, which feeds the control to the MOSFET, can deliver at max 5 V and at minimum 0V. However, a MOSFET on the high end can only be a P channel MOSFET, which will need a low potential to turn on and a high potential to turn off. This High potential must be close to 40VDC, which the microcontroller is incapable of producing. Talon can best answer this.

a. Efficiency

While the whole world recommends driving LEDs from a constant current source, and while Philips itself recommends the same, Talon has resorted to a constant voltage approach with a current limiting resistor.

It is described that every 12 LEDs are placed in series, in a  string. These 12 LEDs are run off a 40V constant voltage, with a current limiting resistor.

String voltage = 12 X 2.022 = 24.264V

Current Limiting Resistor Value = (15.74V)/(0.25A) = 63Ω

Peak Power dissipated in resistor = 4W

Efficiency of each string = (Power Across LEDs)/(Power across String)  = (2.022*0.25*12)/(40*.25) = 60%

Assuming that the switched mode power supply has an efficncey of between 80-90%, the total electrcial efficency of the anti collision light is between 48% and 54%.

b. LED current variation

Because each LED string is driven off a constant voltage source with a current limiting resistor, and because the voltage of LEDs varies with temperature, there is bound to be a variation in the current through the LEDs.

LED voltage variation = Upto -4mV/°C

At a LED thermal pad temperature of 110°C, the LED voltage = 2.022+(110-25)*-0.004 = 1.682V

String Voltage = 1.882 * 12 = 20.184V

Voltage across Resistor = 40-20.184 = 19.816V

Current through resistors (and hence LEDs) = 314mA

At a LED thermal pad temperature of -20°C, the LED voltage = 2.022+(-20-25)*-0.004 = 2.202V

String Voltage = 2.202 * 12 = 26.424V

Voltage across Resistor = 40-20.184 = 13.576V

Current through resistors (and hence LEDs) = 215mA

The variation in LED current (assuming an invariant resistance with temperature) is such that the LED builds up more heat when hot and produces less heat when cold. This can be remedied by a resistance with a  positive thermal co-efficient.

c. Power Factor Correction

The invention first rectifies the 115V 400Hz AC to ~115V DC using rectifiers. Because this is a non linear switched device (power rectifier), there will be a distortion induced in the 115V AC line. While this distortion results in a power factor less than 1, it may be remedied by the use of filters, which may be resolved by the EMI filter described in the invention. However, no reference is made to the apparent power factor of the device, which is important for aircraft AC buses.

6. Market Information

Talon starts off their patent very well. Their description of the lights on an aircraft reflects a passionate understanding of their need in aviation. They even dive into the market details, quoting US$870 for a single flashtube, although they don’t really mention which flashtube they’re referring to.

7. Background to the Invention

In paragraph [0004], the Talon inventors state that “…in the case of a xenon flashtube, use hot filaments at each end of the flash tube to initiate electrical discharge through the flash tube. As such, take off and landing shocks, in addition to in flight vibration, causes all of these lamps to fail frequently”. The truth is, flashtubes do not use a hot filament of any form. There is a trigger filament in some cases, but that is on the outside of the glass tube to initiate the discharge. This trigger coil/filament may be done away with, if the flashtube is in close proximity to a metal reflector. This proximate metallic reflector serves to ionize the gas, as is the case with most aircraft anti-collision light designs employing flashtubes. These flashtubes can last for more than a month, typically running into months.

In paragraph [0006], Talon talks of a Honeywell awarded patent, US Pat. No. 6,483,254. This patent was filed in July 2001, when high power LEDs were just introduced. Yet, the design seems to have incorporated LEDs with a P4 package (this square like package consists of 4 leads, 2 of which are internally connected as the anode, and the other two are internally connected as the cathode). The P4 packaged LEDs give out a low luminous flux, but because they direct this low flux in narrow angles, they produce high luminous intensities of around 4-10cd. As such, with low currents of 20mA to 80mA, they do not dissipate much heat, and rely only on the leads of the P4 package to dissipate whatever little heat they produce. As detailed in the 2001 filed US patent, these LEDs are mounted on anodized aluminium disks, which are more than enough to extract heat from low power, strobe LEDs.

From an optical perspective, Talon is right; there has been no reference to the intensity pattern, or how that is met.

Reference has been made in paragraphs [0008] and [0009] to different light intensity requirements for “small” and “large” aircraft. However, certified fixed wing aircraft, be it large or small, fall into only one of 2 CFRs: Part 23 or Part 25. The anti-collision light system as required to be installed on Part 23 and Part 25 aircraft are identical, in intensity, colour requirements, and the intensity pattern.

In paragraph [0010], the Talon inventors refer to US Pat. No. 7,236,105, wherein they observe, “Problems with this device are that no provisions are made for heat sinking”. This patent is owned by Emteq, which has now taken over Flight Components AG. Understanding their anti-collision light involves a deeper understanding of LEDs, in particular their thermal management. While this shall be talked about in necessary detail in the section that covers Emteq’s patents for a 150cd light and a 400cd light, it must be touched upon in brief in this analysis. Commercial Off the Shelf LEDs, such as those manufactured by CREE and Philips Lighting, have a substrate made of ceramic. The LED die is placed on one side of this ceramic, while right underneath the die, on the other side of this ceramic, is the metallic heat sink tab. As Philips has very clearly shown, in their Application Brief #32 (LUXEON® Rebel and LUXEON® Rebel ES Assembly and Handling Information), the heat generated by the LED is thermally conducted to the heat sink tab through the ceramic substrate.

Cutaway of a Luxeon REBEL LED

What Brenner et al have very intelligently done is to do away with the LED’s substrate, and directly mount the LED die on the circuit board. As a result, the heat generated by the die flows directly to the printed circuit board, instead of having to pass through the ceramic substrate before reaching the board. This is a practice of advanced thermal management.

When Talon filed this patent application, another advanced patent of Flight Components AG was not yet available in the public domain. This is US Pat. No. 7794110, which is an entire reference to European Patent awarded to the Flight Components AG, on the 13th of June, 2007. This patent talks of a light from Emteq, which is available in the market as a 400cd Red Anti-collision light for the fuselage and 400cd white anti collision light intended to be used at the wingtips of aircraft.

8. Conclusion

Talon’s red anti-collision light is well made, and aesthetically appealing, and probably aerodynamically well performing. But when it comes to the LED’s intensity, and its proximity to 400cd at room temperature, the light may fall below the minimum levels at elevated operating temperatures. Besides, with the electrical circuitry designed to not operate at temperatures that are normal for most airliners, the design is disappointing. Moreover, the electrical efficiency of the whole unit is a meagre 50%.

What Talon does differently is the constant voltage circuitry and the low current through LEDs. Electrical analysis as shown above proves that the current doesn’t vary too much, but it is enough to change the intensities.

Their background information, and the technical observations contained therein are flawed, and a result of a shallow study.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s