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.

UCC28740: Transformer construction

Part Number: UCC28740
Other Parts Discussed in Thread: CSD19531Q5A

Dear TI Technical Support / others

i have bit problem when build the transformer from WEBENCH, i got the transformer report like bellow picture, but not any factory can make it

the problem is at the secondary winding use the 25AWG by 7 strands, the factory said  the Coil former (EF20) will not enough

so my question is, is posible if i change the size of wire from 25AWG to ?, & what the effect (the output power designed 3.3V 3.3A)

this power supply purpose is use for non stop, so will keep on

Many thanks

Kurnia

  • Kurni,

    I would agree that trying to wind 7 parallel strands of 0.45-mm triple-insulated wire is not practical. And especially with this small core and bobbin size.

    And for 11-W of power, this also does seem to be excessive copper in the secondary for the power level.

    I will get the Webench team to look into this and get back to you.


    Meanwhile, I would suggest trying a reduced number of strands, maybe 3 x 0.5-mm triple-insulated - that should almost fill 1 layer, and still leave room for wire entry/exit.


    Thanks,
    Bernard
  • Kurnia,
    I agree with you and Bernard that this is not right. I will submit request to the webench team to take a look at that and find the errors causing this. Apologies for the confusion.
    I have looked at ti.com/powerlab and internally and was not able to find any reference designs readily available to show you other than this but it is not exactly what you need: http://www.ti.com/tool/pmp5169

    therefore, in the meantime, I suggest you use the UCC28740 excel design calculator here: http://www.ti.com/tool/pmp5169
    by going through this you should obtain primary magnetizing inductance, turns ratios, primary peak and secondary peak current maximums. with this information you should be able to provide it to a magnetics vendor and they can choose appropriate wire gauge and other considerations. If you do not have any magnetics vendor contacts consider to look at this post where we have PDF in there with many:
    e2e.ti.com/.../589215

    regards,
    John
  • Hi Bernard & John

    Many thanks for your suggestion

    basically i need power supply by AC input parameter : Vin: 85VAC - 165 VAC, output 3.3VDC 3.3A, (Vmin 3.15VDC Vmax : 3.55VDC)

    i have try calculate using UCC28740 Design calculator (attached)

    i have bit confuse at sheet "START HERE", row 43 "Vf Dout", because the schematic result from WEBBENCH has no any Diode output, but use the Mosfet CSD19531Q5A (Texas instrument) controlled by the UCC24636DBV,

    so is it correct i entry that parameter by VSD parameter the Diode characteristic at the mosfet (datasheet attached) ?

    i found the result calculation of Peak Secondary Current of Transformer is so big : 18.688 A, is it correct formula ?

     

    Thanks

    Kurnia

    webench_design_4408770_183_339536169.pdf

    transformer_report_4408770_183_822195766.pdf

    sluc487b.xlsx

    m2_csd19531q5a.pdf

  • Kurnia,

    I believe that Webench should support UCC28740 design with diode rectifier and with SR - you seem to have one with SR. This is probably the better one to use if you need best efficiency, the SR will dissipate much less power than diode rectifier.

    In the Excel design spreadsheet, you can use the MOSFET body diode Vf - this will only affect the choice of resistor values on the VS pin for OVP. Since the UCC28740 uses opto feedback, there is no concern about the effect of Vf on the regulation setpoint.

    The secondary peak current will be high, this will be the peak primary current scaled by the turns ratio. With Rcs~0.91, Vcsmax ~ 0.77 V, that’s 0.846 A on the primary, scaled by Np = 22, that’s 18.6 A peak on the secondary.

    I hope this answers all your questions.

    Thanks,
    Bernard.
  • actually the basic my problem is the transformer construction,
    so let me try your first suggestion the secondary winding use 0.5 mm X 3 strands
    hope my transformer partner can build it by this configuration
  • Kurni,

    Sounds good. Let us know if you have further questions.

    Regards,

    John

  • Hi Bernard

    Sorry, i have more question, actually what is the formula to found the secondary winding, i mean can be found 3 X 0.5, or just engineering instinct

    & how to verify if the result is good enough for long life , as the purpose will work for 24 hours non stop (will keep working)
    is can just measure the temperature of secondary winding by full load ? or another way

    many thanks
  • Kurni,

    It is very difficult to choose the optimum size of wire diameter and number of strands.

    You will need to have suitably low DCR (DC resistance) for the secondary winding rms current, but this is not the only consideration. Depending on the diameter, layer count and winding structure, there will also be significant ACR (AC resistance), and this cause losses that are just as important as DCR loss, since the triangular current waveform also has significant high frequency harmonics.

    Usually the wire diameter is chosen based on the switching frequency and the winding structure, then the number of stands is chosen to meet a reasonable DCR loss and/or to give a full layer with good fill.

    It comes down to experience, it's hard to give easy design rules to follow.

    The TI Power Supply Design Seminar “Flyback transformer design considerations …” topic might be useful, it talks about how a 65-W Flyback transformer winding structure was improved to get lower losses (both ACR loss and leakage inductance loss):

    www.ti.com/.../login.shtml


    Thanks,
    Bernard
  • Hi Kurni,

    We are updating the WEBENCH algorithm for Xformer design to limit the number of strands for Triple Insulated Secondary Winding.

    Meanwhile I could create your design with this Transformer which uses 4 TIS strands. Can you see if this will help you in your calculations,

    transformer_report_UCC28740.pdfdesign report_UCC28740.pdf

    Thanks

    sahil