• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 2 answers
  • Subscribers 62 subscribers
  • Views 500 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.

UCC256304: Output Capacitor Ripple Current

Mike Miles52
Prodigy 50 points
Part Number: UCC256304
Other Parts Discussed in Thread: UCC25600, PMP9750

I'm working on a design with the UCC256304 for 14V 40A output and no PFC. This has the usual split-secondary transformer with two rectifiers feeding a capacitor for the dc output. The handy TI spreadsheet shows a ripple current of 22 amps which seems high and difficult/expensive to manage. Our older product (a hard-switching design but with similar secondary rectifier configuration) has an inductor in series with the rectifiers (which also have RC snubbers) and I found in my spice simulations that this inductor ahead of the output capacitor helps a lot with the capacitor ripple current. However the inductance does seem to reflect back to the primary and affect the resonant drive circuit on the primary side. In my simulation I arbitrarily reduced the primary series inductance to compensate.

My spice circuit is attached, with the two versions compared. Vout1 (green) is output voltage without the output inductor (5V p-p ripple). Vout2 (lt blue) is output voltage with inductor (0.25V p-p ripple). But look at the capacitor ripple - I(C1) (yellow) shows about 80A p-p ripple, while I(c2) (red) shows less than 6A p-p ripple.

The reduction of output capacitor ripple current and output ripple voltage seems to be quite dramatic, although there is some voltage drop.

1. Is this a good approach, and is it advisable?

2. How will this affect the overall efficiency?

3. Should the secondary inductor, as reflected through the transformer, be considered part of the primary inductance?

Thanks!

Mike

  • 0
    TI__Mastermind 27145 points
    Hi Mike,

    Thanks for reaching out. This concept is very similar to the CLL resonant converter which gets rid of the primary resonant inductor and uses an output inductor after the rectifying diodes. This output inductor then participates in resonance. It's possible to run CLL above resonance which will make the current in the output inductor continuous which will limit the total peak to peak ripple in the output capacitors.

    Take a look at PMP9750. This is a CLL resonant converter reference design based on UCC25600:http://www.ti.com/tool/PMP9750

    For reference, this app note talks a little bit about CLL resonant converters starting on page 16:www.ti.com/.../slup376.pdf

    Best Regards,
    Ben Lough
  • 0 in reply to Benjamin Lough
    Prodigy 50 points

    Hi Ben,

    That's very interesting; the discussion in slup376 seems right to the point about the trade-offs with secondary inductance, and that seems very beneficial in my design. Is it safe for me simply to move the inductance as calculated by the UCC25630x Design Calculator spreadsheet - or some of it, to balance the design - to the secondary side after the diodes, with a turns-ratio (1/n squared) adjustment? That appears to be what was done in PMP9750. Is the operation of the UCC256304 unaffected by that topology change?

    The reference designs don't seem to reveal the transformer's leakage inductance, although my Webench designs do specify leakage inductance at 0.02 (1/50th) of the primary magnetizing inductance. Is that a safe rule of thumb for the transformer performance? Is it correct to include this (subtract it from the calculated requirement) in the primary series-inductor value determination?

    Also I've read about the transformer potentially providing all the necessary resonant inductance with its leakage and eliminating altogether the need for separate inductors. Is that generally feasible? The EVMs and reference designs I've seen all use discrete inductors; is that because it's challenging to make a transformer with enough leakage inductance? With my cost-sensitive design here, eliminating parts (and reducing ripple current) would be great. Any insight about that aspect of the transformer design would be most appreciated.

    Thanks for your help!
    Mike

  • 0 in reply to Mike Miles52
    TI__Mastermind 27145 points
    Hi Mike,

    Yes, you would use the relationship Lr=(n^2)*Lout to calculate the output inductance for CLL. I do not anticipate any other changes necessary for UCC256304.

    Yes, the primary side leakage would add onto your resonant inductance. 1% or 2% is fine. LLC is more forgiving on leakage inductance than other topologies as it simply adds to the resonant inductance.

    Yes, it is possible to utilize the leakage inductance of the transformer as the total resonant inductance and get rid of the external inductor completely. This approach has been done before successfully with LLC. I would say utilizing the leakage as the resonant inductance makes the transformer design a little more complex but definitely still feasible. You want to have a controlled tolerance on the leakage inductance in order to get reliable performance. Here are a couple reference designs that use the leakage inductance as the resonant inductor.

    www.ti.com/.../tidrxi5.pdf
    www.ti.com/.../tidrt45a.pdf


    Best Regards,
    Ben Lough
  • 0 in reply to Benjamin Lough
    Prodigy 50 points
    Awesome, thanks Ben!
    Mike