• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 241 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UCD3138ALLCEVM150: LLC converter (40V…60V) to (40V…60V)

Sergi Lopez Cano
Prodigy 105 points
Part Number: UCD3138ALLCEVM150

In the Project we are developing we are designing two converter:

  1. LLC converter 400V to (40…60V) with the resonant tank already calculated and already tested @ lower Power.
  2. LLC converter (40V…60V) to (40V…60V) with the resonant tank still not calculated.

This second converter design is not finished because following the design guide from TI (slup263_TI.pdf), the calculated Lm is VERY LOW (around 0,5uH) in half brigde topology. If we move to full bidge, the LM value is a little bit higher but not so much. These values are not manufacturable. Using an EE core, if you remove one of both magnetic parts (E) of the core, the primary magnetizing inductance value is still higher than the one calculated.

We want to know if there are other resonant topologies better than LLC for that converter. 

Thank you.

  • 0
    TI__Genius 17090 points

    I've assigned this thread to an LLC expert.  I expect you will get a reply on Monday.  

  • 0 in reply to Ian Bower
    TI__Expert 3370 points

    Hello,

    phase shift Full bridge is a other resonant topology. But PSFB's advantage is for High input to Low output application. LLC may is good selection for your application I think. To getting higher Lm you have to do. using core with higher Ae or selecting higher transformer turns is effective way to increase the inductance.  

    thanks 

  • 0 in reply to Bliss Zhou
    Prodigy 105 points

    Hello Bliss Zhou,

    First, thank you for your reply.

    We understand that we can do some actions to increase the Lm. However, we must match the value of the Lm with the calculated one (following the design guide form TI slup263_TI.pdf). I think we can’t increase the Lm if it should be lower according to our calculations, because it will provoke modifications in the resonant tank.

    Let me share some of the values that we have obtained. We can share with you an Excel with the details of our calculations:

    • Vin= 40 to 60 V due the specifications.
    • Vout= 40 to 60 V due the specifications.
    • Pout= 500 W due the specifications.
    • Mg_min= 0,64 due the maximum input voltage and the minimum output voltage.
    • Mg_max= 1,66 due the minimum input voltage and the maximum output voltage.
    • n= 0,5 due the 48V of the nominal input voltage and the nominal output voltage
    • Re = 0,93 Ω due the nominal output voltage and the nominal power output.
    • Ln= 2 and Qe= 0,2 due the Mg_min and the Mg_max.
    • Cr = 8,52 µF due the Re and Qe.
    • Lr = 0,30 µH due the Cr.
    • Lm = 0,59 µH due the Lr and the Ln.

    For these reasons, we have obtained this value of Lm.

    In summary, we should know how increase the Lm trough the calculations or how to get a Lm as small as the calculated.

    Let me know if you want our Excel with all the calculations.

    Thank you.

  • 0 in reply to Sergi Lopez Cano
    TI__Expert 3370 points

    Hello Sergi,

    You don't need share me the excel you calculated. LLC is hard to cover wide range input/output.  It seems that single stage LLC is hard to match you application. Do you prefer to pre-Boost is placed before LLC? The system architecture is Boost + LLC.

    thanks 

  • 0 in reply to Bliss Zhou
    Prodigy 105 points

    Hello,

    What do you think about using a half bridge not resonant topology? Do you think it could be suitable fot our application? Or maybe you can advise to us about another topology (not resonant) suitable for our application.

    Thank you.

  • 0 in reply to Sergi Lopez Cano
    TI__Expert 3370 points

    Hello,

    That way may not good efficiency. Boost + LLC is probably good selection.

    thanks