Skip navigation

community

Up to Discussions in Lighting
2524 Views 4 Replies Latest reply: Jun 24, 2010 12:22 PM by Jorge_Garcia RSS
Jorge_Garcia Level 4
Jorge_Garcia
4
Currently Being Moderated

Jun 22, 2010 12:38 PM

Suggestions for designing an LLC Resonant Converter for T5

Hi all,

 

If anyone could point me to some good app notes, or share your experiences with Resonant Converters. I will be designing and building my first one and I feel a little unsure. I've already done a lot of research and I understand most of the operating principles but I don't feel like I have enough background to successfully build one. As the saying goes "in theory, practice and theory are the same, in practice they are not." so I would like to hear some real world tips and experiences with these types of converters.

 

Thanks for your help.

 

Best Regards,

Jorge Garcia

Attributes
  • enrico.migchels Level 4
    enrico.migchels
    Currently Being Moderated
    1. Jun 23, 2010 6:58 AM (in response to Jorge_Garcia)
    Re: Suggestions for designing an LLC Resonant Converter for T5

    Hi Jorge,

     

    I find starting up with high loads troublesome. Calculation of the transformer is difficult because the physical turns-ration will not give the correct voltage tuns-ration. Avoid switching below the self resonant point (hard switching / capacitive mode switching). There are controller with internal protections for this.

     

    On T5/LLC

     

    ST Controller L6574

     

    NXP controller: UBA2014

     

    Application note on UBA2014:

     

    http://www.dianyuan.com/blog/u/2009-08/337966_1251106930.pdf

     

    Best regards, Enrico Migchels

  • Jorge_Garcia Level 4
    Jorge_Garcia
    Currently Being Moderated
    2. Jun 23, 2010 4:59 PM (in response to enrico.migchels)
    Re: Suggestions for designing an LLC Resonant Converter for T5

    Thanks Enrico,

     

    I was actually, thinking about running the converter without the transformer. The PFC stage gives me a 400V rail which when driven through a full-bridge or half bridge configuration gives me an 800V Peak to Peak square wave, the T5 Bulb can strike the arc with a voltage lower than that so the converter will probably be running a little above resonance in the steady-state.

     

    At Startup the converter would start with a very high frequency (2 to 3 times resonance) then work it's way back towards resonance causing the voltage to increase until the Arc is struck, once that happens the control system would have to once again pull away from resonance to lower the output voltage to normal operating conditions (As you know fluorescent tubes demonstrate high impedance until the arc is struck, after the tube impedance drops drastically). I know that once the arc is struck the change in load impedance will alter the converter's Q so I'll have to analyze the impact on the system gain to make sure I can keep the converter stable.

     

    I was also thinking to use a dsPIC to run the control loop and to monitor fault conditions such as no bulb.

     

    Let me know what you think, I've seen transformerless resonant converters before so I know it can be done.

     

    Thank you for the info. I really appreciate it.

     

    Best Regards,

  • enrico.migchels Level 4
    enrico.migchels
    Currently Being Moderated
    3. Jun 24, 2010 3:40 AM (in response to Jorge_Garcia)
    Re: Suggestions for designing an LLC Resonant Converter for T5

    Hi Jorge,

     

    I have made such an transformerless application once for an 200W hot cathode fluorescent lamp. The resonance capacitor was located near the lamp itself (in the small starter housing). The input voltage was 600Vdc as it was a 3 phase converter, it had some higher ignition voltages as the tube was more than 2 meters long! A so-called XXL-tube :-)

    I find your approach to digital control challenging. I would advise to build the application with an 'analog' controller and study the preheat-ignition-burn stages, voltages/currents and timing carefully. If the design is finished and running without problems make the step to digital control. I don't know about your programming skills but there are several steps in controlling an smps. step 1 'on/off control' to step 4 'full digital control'. I would advise that if your knowledge on programming is not up to level 4 that it would probably be better to start with lower ambition. There are digital controllers with good analog pherifirals (such as opamps) to have the timing critical regulation loop still analog and the rest digital (example PIC16HV785). If your design is in an research environment there are less risk, if it is a product development (with a target finishing date) there are considerable risks involved in planning.

     

    Best regards, Enrico Migchels

  • Jorge_Garcia Level 4
    Jorge_Garcia
    Currently Being Moderated
    4. Jun 24, 2010 12:22 PM (in response to enrico.migchels)
    Re: Suggestions for designing an LLC Resonant Converter for T5

    Thanks Enrico,

     

    I appreciate the advice and the words of wisdom you've offerred. I think you're right I shouldn't go for a digital control loop right off the bat, but rather try to get the application working with an analog controller first and then (if there's time) make the move to digital control.

     

    Thanks for your help, once again.

     

    Best Regards,

    Jorge Garcia

Related Content


Related Products
Discussions
  • Retrieving data ...

Bookmarked By (0)

Legend

  • Correct Answers - 50 points
  • Helpful Answers - 25 points