Replies

• 

  Posted by Bugs Bunny at Tue May 15, 2012 6:01 pm

  Hi Tony

  Have a look at http://www.melexis.com/General/General/MLX10803-419.aspx

  Maybe is this information useful for you?

  But what do you mean when you write “at 6.6 amp I need to produce 640ma....”?

  Would be glad to help, but need full information.

  Best regards,

  B.B.

   

• 

  Posted by simmyone at Thu May 17, 2012 2:04 pm

        B.B. the following is a description of the power supply and driver we are trying to develop.
  

   

        The LED will be powered by 6.6A alternating current and may have some harmonic content from the system via an isolation transformer. The rated current will produce 100% intensity per the Photometric Requirements (PR) above. The LED will produce 30% intensity of the PR at a current of 5.5A and 10% intensity of the PR with a current from the system of 4.8A. It is noted that the LED light output is linear with respect to the current from the system and that the intensity spectrum should meet the photometric requirements of 30% to 51% at a current of 5.5A from the system and intensity of 10% to 19% at 4.8A from the system. It is further noted that the output current varies to some 83% (5.5/6.6) and the LED output must be between 30% to 51%. Similarly the LED output must be between 10% and 19% at a current of 73% (4.8/6.6). It is further noted that the current supply wave shape may not be true sinusoidal and will vary at each level of current step, each type of power supply, constant current regulator (CCR), and the value of load placed on each system.

        LED driver, including a power supply (power converter)to convert the 6.6A to the appropriate current and voltage to drive the LED driver, will be designed and used to power the LED at 640 DC mill-amperes (ma) at approximately 3.3VDC.   The LED driver will be designed to dim the LED to 30% intensity range (30% to 51%) at reduced current of 5.5A and dim to 10% (10% to 19% range) with a current of 4.8A. The method of dimming may be power width modulation (PWM) or reducing the current to the LED. The LED chosen will require a duty cycle of 30% to 51% for the 5.5A and 10% to 19% duty cycle for the 4.8A supplied. If the current method is used the value of current will reduce from the 640ma at 6.6A to 150 ma (medium range 40%) at 5.5A from the system, and dim to 10% with a current of 35 ma to the LED (15%) at 4.8A level from the system. The mid-range values are chosen to meet the requirement for all operational conditions. If PWM is used as the dimming method the brightness changes must not cause perceptible flicker to a moving observer throughout the brightness steps.

        The LED driver will be design to be efficient as possible. The driver and power converter is desired to be 90% efficient as a minimum at the 100% intensity.  The true input power factor to the LED driver and power converter shall be a minimum of 0.7 on all intensity settings over frequency bandwidth of 100KHZ. The power factor measurement will not be displacement power factor (cos o). The testing will be performed using a pure sine wave.

        There are some surge supression and noise that also have to be met but this sholuld give you an idea of what we are looking to achieve.
  

  
  

   

   

   

   

   

   

• 

  Posted by Bugs Bunny at Mon May 21, 2012 10:28 am

  Hi Tony!

  I spent hours to try to find solutions to your problem.

  I have some ideas concerning the LED driver.

  Sense: a 100:1 current transformer, correctly loaded. Better than a shunt (isolation). Following: full-wave rectifier, resettable integrator and sample/hold synchronized with the line frequency. This solution overcomes the problems caused by harmonic content, if a peak-detector was used. Output sent to three window-discriminators, each one set for the three current levels. Each output will drive the Melexis switching LED driver to produce the correct light levels. These levels are actually fully adjustable between 0 & 100%. No Pulse Width Modulation (DC drive), because to expensive for driving one (?) LED, although rather easy to implement. All of this is currently 80% ready (schematic diagram, of course). Design free from any flicker. But the efficiencies you mention look like a dream at such low power – less than 3 Watts – especially when the line currents are in the lower ranges.

  Just a question: what are the accuracies of these 6.6 / 5.5 / 4.8 A currents? Needed to fix starting limits for the window-discriminators. For the same reason, the mean expected harmonic content should be helpful, too.

  Next brainstorming: get roughly 3 watts out of the constant current line...

  Your opinion? Any suggestion?

  Kind regards,

  B.B.

  PS: why not an independant power supply, e.g. a small switching Li-Ion battery charger, with sufficient output voltage, ready from the brand? I understand your project as a simple monitoring device?

   

• 

  Posted by Bugs Bunny at Mon May 21, 2012 2:38 pm

  Hi Toni!

  Please forget the PS in my previous post. (And don't laugh). I believe I was completely wrong about your project. Intended to mark out poles and runways, isn't it? I now clearly understand your interest in high efficiency. I will replace the current transformer with a shunt followed by an instrumentation amplifier. Use more efficient IC's (bias currents), to allow higher values for resistors. PWM recovers all his interest, as the Melexis is more efficient with high currents. And a challenge: try to recover the energy lost during the discharge cycle of the integrator, as this latter is the most power consuming part of the driving circuitry.

  Best regards,

  B.B.

• 

  Posted by simmyone at Mon May 21, 2012 3:09 pm

  B.B. 

  Thank you for the response, to answer your accuracy question.  The output currents and their respective limits are set by FAA standards and are quite accurate.  They are as follows:

  6.6 amp - 6.5 to 6.7

  5.5 amp - 5.4 to 5.6

  4.8 amp - 4.7 to 4.9

  You are correct, this application will be for an elevated Taxiway light fixture, thus the need for blue led.  Our intent is to take this design and apply it to all the elevated airfield fixtures we market eventually, but the taxiway is currently the most popular.  Our industry is rapidly changing and the FAA is pushing to replace all incandesant fixtures with alternate light sources, the LED seems to be the best alternative.  We are alot smaller than our competitors and do not have the in house resources that many of them do.  This is why we are turning to people such as yourself for guidance in our efforts to manufacture the most reliable and innovative product to date.

  I thank you again for your interest in this project.

  
  Tony

   

   

• 

  Posted by simmyone at Mon May 21, 2012 3:12 pm

  I have forward all of our communication to our design engineer for his review and again I really appreciate your input on this project.

  Tony