• State
  • Locked Locked
  • Replies 11 replies
  • Subscribers 66 subscribers
  • Views 10018 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

UCC25600 Based 3KW LLC DC-DC converter

Cristi Vasilache
Prodigy 10 points
Other Parts Discussed in Thread: UCC28070, UCC25600, UCC28950, UCC28061, UCC29950

I have to design and build a AC-DC converter which will have to deliver 25V DC at 120A and the efficiency must be above 90% including the PFC stage, so  the DC-DC converter itself must be at least 92.5 to 94% efficient. I was thinking to build the PFC using the UCC28070 and for the DC-DC converter i had to choose between a Phase-Shift FB and a LLC Converter. I want to approach an LLC Full-Bridge converter design and from all the IC's which are available, i would like to use the UCC25600 since it has adjustable DT and no gate drives, which offers me flexibility in driving the power stage. For the rectification i will consider Synchronous rectification. The primary DC voltage i also want to increase at 420-440V DC, to take advantage of the lower switching losses and minimize the conduction losses. I plan to use Infineon Cool MOS transistors.

Now the question is, did someone used this IC or at least the LLC Full-Bridge topology at this power level ? What kind of difficulties i should expect ? Would be practical to use this topology at this power level ? And most important, Reliable ??? I do know that there are some issues regarding MOS-FET's failure rate in LLC converter at low load or no load due to the body diode conduction. 

Please advise.

Thank you very much.

  • 0
    Prodigy 150 points

    For 3kW full bridge is a must, since  >1kW application draws the capabilities of the half-bridge topology.

    I managed to use half-bridge LLC converters up to 1kW, but over that, the gap in the core, needed to give appropiate magnetising inductances get bigger. The high magnetising current also causes high idle losses, so these are the difficulties.

    With a UCC28950 an SMPS with your parameters is realisable.ű

    Hope this helped.

  • 0 in reply to Laszlo Lorantfy
    Expert 1920 points

    I've used the UCC25600 up to 550W (designed to 800W). The LLC is great below 1kW but has the disadvantages as described above IF you're regulating the output. I designed mine as a chopper at the resonance frequency (unregulated), which gave me around 97% efficiency at 520W (for LLC alone).

    Input 400VDC (from PFC UCC28061).

    Output: +/- 82VDC (165V at light load and 162V at 520W).

    These specs gave me almost a 1:1 turns ratio of the transformer (one ETD39).

    Read the ST notes regarding LLC. Of course a LLC at 3kW is possible, but I would recommend the UCC28950 for these 3kW power levels.

  • 0 in reply to CWilson
    Prodigy 70 points

    Hello (again) Mr. Wilson,

     I wanna build also a +/- 80Vdc SMPS with this resonant controller.

    Can you tell me how did you wounded (what cross section for Cu, haw many turns )the ETD core ?

    Did you used an external Lm or you tried to intergrate it into ETD core? 

    Knowing that the excell sheet doesn't work so good, can you detail tour design (in therms of values for the resonant components and sensing also)?

    Thank you!

    Regards,

    eng. ALBU Gheorghe.

  • 0 in reply to albu george
    Expert 1920 points

    Hi A. Gheorghe,


    Frequency: 100kHz.

    Wire diameter: 0.5mm (matching 100kHz at around 80 degrees Celcius).

    As I recall, it was 16 turns on the primary and 14 on the secondary.

    I used an external Lr (resonance inductance). The size of this is Lr = Lr (theory) minus the leakage inductance of the primary.
    In my case I should have around 2.7uH and the leakage was 1.3uH (measured). Then the Lr should be: Lr = 2.7uh - 1.3uH = 1.4uH.

    I didn't have any overcurrent sensing since my LLC was unregulated.

    U can see some info here: http://e2e.ti.com/support/power_management/isolated_controllers/f/188/t/162389.aspx 

    I'd suggest that you read the ST note regarding LLC.. There is two.. One of 32 pages and one of 64 pages.. Start with the 32 pages edition to understand the LLC principle.

    Good luck :o) 

  • 0 in reply to CWilson
    Prodigy 70 points

    Helllo Mr. Wilson,

    Thank you for the info and the docs.

    I'lltry to read all of it and then try to start the design.

    Though I saw that your ref were some ST products I think I'll stick to TI UCC25600. It seem's more simpler.

    By the way, at full load what was the output voltage diference refered to no load condition?

    Thank you again!

    As soon as possible (don't know what "soon" means) I'll get back to you with the rezults.

    Regards,

    George.

     

  • 0 in reply to albu george
    Expert 1920 points

    The ST notes is just to learn the theory of operation of the LLC.. Read also Hong Huang's paper/walkthrough I linked to.

    I used the UCC25600 :-D


    My output was +/- (not GND to "+" since my SMPS was for a Class D audio amplifier). Measured from - to + it was:

    Vout no load: 165VDC.
    Vout 515W out: 162VDC. 

  • 0 in reply to CWilson
    Prodigy 70 points

    CWilson, I suppose you didn't used any kind of feedback, correct?

    Just the simple schematic in the app note, but without any regulation, yes?

     

  • 0 in reply to albu george
    Expert 1920 points

    Correct.

    No feedback since my output voltage dip of 3V wasn't aproblem for the following Class D amplifier.

  • 0
    Prodigy 140 points

    hi

     how about your 3KW Project now?

    i think Full brigde LLC is fit for 3kW..

    but the controller of this FB LLC is difficult to select,

    switching frequency :200~250KHz

     

  • 0 in reply to Jack.Leung
    Prodigy 85 points

    And how about a bridgeless pfc?

    I'm currently designin one and will post some results after it works...

  • 0
    TI__Expert 3095 points

    Hi Cristi Vasilache,

    Please take a look at this application note:

    http://www.ti.com/lit/pdf/slua733

    It is intended for a different controller (UCC29950), but outlines a different approach to LLC converter analysis/design. It should be applicable to any LLC converter including one controlled by UCC 25600.

     I hope you find it useful.

    Joe Leisten