• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 234 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.

UCC25800-Q1: UCC25800 6W LLC converter heat issue

udhay kumar
Intellectual 305 points
Part Number: UCC25800-Q1
Other Parts Discussed in Thread: UCC25800

Tool/software:

Hi,

I need to power an Add-on card in a rack module with UCC25800. I choose this IC because of low inter winding capacitance so the EMI noise of one card will not affect the other cards in the system.

I have designed the circuit and it is working with satisfaction. The IC or the toroidal core is heating within 42°C, but the output diodes are peaking upto 53°C with natural air cooling. This may exceed 60°C+ with enclosure. Is it normal. What should I do to reduce the temperature.

My design parameters are as follows,

Input        -    24V regulated

Output     -    22V to 26V

Power      -    4W (should capable upto 6W)

Frequency    -   550KHz

Toroidal core    -    B64290L0044X49   (AL= 660nH, Primary inductance = 35uH, Primary turns = 7, secondary turns = 8)

I have taken most of the values from EVM schematic and wound the primary coil with inductance of 35uH which is near to the 26uH of reference design. Here I have attached the thermal image of the PCB. The core is in almost ambient temperature at 200mA output current. The IC is 42°C with very minimal thermal PCB track size and it can be improved in actual design. ES1D output diode is selected because of 15nS recovery and 7pF of junction capacitance. The diode has 200V and 1A rating where my output rating is only 24V@ 200mA.

What mistake I made here?. Is it the nature of the circuit or I should select a different diode with different parameters?

Udhay

  

  • 0
    TI__Mastermind 33115 points

    Udhay,

    Your resonant tank is likely not turned to your switching frequency. Try:

    1. Measure the leakage inductance and consider this to be Lres
    2. Cres=C21+C22
    3. Calculate Cres needed to make Fres slightly below Fsw
    4. As a minimum Cout, make sure Cout>10*Cres
      1. Currently your schematic shows Cres=2μF and Cout=4.7μF (Cout=2.35*Cres)
    5. Measure the primary current with current probe and optimize Cres (tune Fres) until you see sinusoidal current waveform.

    Regards,

    Steve

  • 0 in reply to Steven Mappus1
    Intellectual 305 points

    Hi Steve,

    I am using the UCC25800 calculator for the design which I'm attaching here with all the parameter input. Unfortunately the calculated values ar  no way near to your above points, so I'm confused whether I have follow the design sheet or your suggested mathod.

    As per your suggestion, I measured the parameters and given below.

    Primary Inductance - 34.6uH @100KHz

    Leakage Inductance  -  1.5uH @100KHz

    Selected Cres (C21+C22) = 66nF (33nF + 33nF)

    Operating frequency = 510KHz ( RT=51K)

    Resonance frequency = 505.83KHz (for 1.5uH & 66nF)

    The primary current waveform is not a perfect sinusoidal and the screen shot is attached here. The output diode is in 55°C.

    Do you think that the circuit resonate properly? Is it my calculation is Ok? Is there anyway I can improve the circuit?

  • 0 in reply to udhay kumar
    TI__Genius 11350 points

    Hi Uday,

    Steve is OOO until Monday so I hope you can wait until then for a reply from him.

    In the meantime, I'm curious why you selected Cres (C21+C22) = 66nF (33nF + 33nF) while the calculator you shared recommends 2.2uF + 2.2uF. Please check your values against the calculator and share the updated schematic.

    Regards,

    Ray

  • 0 in reply to Ray DiCecco
    Intellectual 305 points

    Hi Ray,

    As per the calculator result, the C21 & C22 is 2.2uF, but Steve suggested to select the Cres value based on the measured leakage inductance value in order to make the resonance value slightly below the actual oscillation value.

    My operating frequency is 510KHz and measured leakage inductance value is 1.5uF, so as per LC tank resonance calculation (1/2π√(LC)) the Cres is selected as 66nF (33nF+33nF). The resultant resonance frequency is 505.83KHz.

    Please correct me if I'm wrong.

  • 0 in reply to udhay kumar
    TI__Mastermind 33115 points

    The calculator values of CB1=2.2μF and CB2=2.2μF form the capacitive divider on the primary and these caps are not part of the resonant tank. You are using secondary side resonance where CR1=CR2=22nF (according to the Excel) and you are using CR1=CR2=33nF. With Lres=1.5μH and Cres=66nF, Fres=506kHz. From the scope plot you attached, your measured Fres is ~526kHz

    "Steve suggested to select the Cres value based on the measured leakage inductance value in order to make the resonance value slightly below the actual oscillation value."

    • Make the resonant frequency slightly below the switching frequency. You have two control variables available to modify, Cres and Fsw.

    Your measured waveform shows you are operating below resonance and looks good according to the operating conditions you have defined. If you want to tune it more, try reducing the dead-time and making Fres~Fsw and you should see further improved THD in the resonant current waveform.

    Steve