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.

TIDA-00367: TIDA-00367 flyback transformer design

Part Number: TIDA-00367

Hi,

Related to the TIDA-00367, is the flyback transformer (T1) design available ?

Thanks,

Manu

  • Hi,

    The transformer details are available in Wurth Electrik (WE) website. Please search for the WE part number: 750342946 in https://www.we-online.com/

    The same can be seen at the page:
    katalog.we-online.de/.../UCC28630-DB-TIDA-00367-Rev-E2_Flyback_1

    Thanks,
    Manu Balakrishnan
  • Hi Manu,

    Ah this is quite confusing. :-)

    I did notice that Wurth has the transformer with them. Actually, we have a transformer manufacturer doing work for us, located here and hence the query related to the transformer design.

    Thanks,

    Manu

  • Any thoughts on the transformer design ? At least any thoughts on where to start with the design of the transformer ?

    Thanks,
    Manu
  • Hi Manu,

    The transformer design is done by Wurth, specifically for this design. The part datasheet from Wurth has the basic details like size (from that, the core size can be found out), turns ratios, isolation requirement between windings, primary inductance, dc resistance and leakage inductance. I think the missing parameter is the number of turns, which you may have to calculate using the standard flyback transformer design. I will not be able to share more details on the transformer, as the transformer design is owned by Wurth.

    Please note that, for this design, the most challenging and important part is to minimize the leakage inductance. The leakage inductance is very crucial as the power level is as high as 150W. Higher leakage inductance will cause challenge in taking care of the snubber losses.

    Please use the design calculator of UCC28630 for detailed design:
    www.ti.com/.../getliterature.tsp

    Please use the below training material to understand more on the transformer design.
    training.ti.com/transformer-designer-isolated-high-voltage-power-design

    Hope this will help you to find the missing parameters like number of turns. Please validate the designed values with the Wurth transformer parameters.

    Please let us know if you need further support.

    Thanks & Regards,
    Manu Balakrishnan
  • Hi Manu,

    Thanks for your reply, but without knowing the core material, vendor etc, it would be quite hard to continue, unless the design is started from scratch, I guess! :-( Looking through past posts on the forum, I feel that many others also feel the same. The whole point of a reference design is to have a reference to something. If someone is holding that hostage, then that doesn't make any sense. How I wish TI would provide the transformer design like how Power Integrations does for their reference designs!

    Thanks,
    Manu
  • Hi Manu,

    We do provide transformer design for many of our designs. Below are some designs where we have transformer design available for a 65W converter using the same part family UCC28630.
    www.ti.com/.../ucc28630evm-572
    http://www.ti.com/tool/PMP9643

    For some reference designs, we directly take the custom transformer parts from Wurth, so that it is easy to support worldwide customer requirements. This will help the worldwide designers to directly get the sample from Wurth for quick prototyping.

    Let me direct this E2E thread to my colleague Ramkumar, he will be able to help you with the specs for the transformer.

    Thanks,
    Manu Balakrishnan
  • Hi Manu,

    I had a look at PMP9643, but that has a slightly lower power than I required, everything else is great. If I can simply increase the output power of PMP9643 from 65W to about 100W, that would exactly fit the requirement.

    I follow very much what you are saying.

    Since we have transformers being manufactured here, importing the same transformer, the levies are much higher. For prototyping/testing sourcing from Wurth might work, but in reality for production when volumes are the intent, then the Wurth road does not seem to work in addition to their their very high price quotes for the transformer.

    It's their product they can determine it's price, but in the end to be sold in the market, it's price should be justified. I could've taken the samples from them and asked manufacturers to simply copy that design, but that's not a comfortable thing for me to do.

    Hence looked towards a design that TI provided.

    Thanks,
    Manu
  • TIDA-00367 Transformer spec.docxHi Manu,

    Please find atached a spec sheet for the transformer. This should work for the 100W output you are looiking for. Please take care of minimizing leakage inductance to <4.5uH.

    Best Regards,

    Ram

  • Hi Ram,

    Thanks for the help. In the construction, W3 seems to be missing ?
    What needs be done to minimize the leakage inductance, as per your suggestion ?


    Thanks,
    Manu
  • Also, have 3 queries on the topic.

    1. W4 is the winding for the Aux power for the UCC28630. But that winding seems to have triple insulation, rather than output that would need that isolation.

    2. W2 and W3 the outputs that require isolation uses standard 26AWG ?

    I guess that's by accident ?

    3. If I need to reduce the output voltage to about 21V instead of the calculated 24V, can I simply assume that 24/(5+4) = 2.66V/turn and thus deduct 1 turn from W2, thus making it 4 turns ?

    Thanks,
    Manu
  • Any thoughts/suggestions ?

    Thanks,
    Manu
  • Hi Manu,

    1. Please use tripple insulated wire for W1 and W4. This will be easier than for W2 and W3 as they are high current outputs.


    3. To reduce output voltage from 24v to 21V, you dont need to reduce the turns as such. Keeping the same turns and adjusting the resistors R11,R20 to get 21V at output. Keeping the same turns just reduces the reflected secondary voltage and hence te snubber voltage.
    Of course you can reduce the secondary turns too little bit if you want.

    Reducing leakage is a more complext thing to answer in a generic sense. In this case, my advice would be to make sure you use the entire length of the bobbin and sandwitch the secondary between primary winding.

    Best Regards,
    Ram
  • Hi Ram,

    Thanks for the explanation.
    Can you please clarify on the construction ?

    The following you provided me:


    Winding Procedure:
    -------------------------------------------------
    * Wind (W1) in three layers,
    * First wind two layers of W1
    * Reinforced Insulation
    * Wind 5 turns of (W2) starting at pin 8 and ending at pin 9.
    * Reinforced Insulation
    * Wind one layers of W1
    * Reinforced Insulation
    * Wind 4 turns of (W2) starting at pin 9 and ending at pin 10.
    * Use triple insulated wire to wind 6 turns of W4
    * Gap core suitably to get required primary inductance
    * Bond the core to avoid audible noise
    * Vacuum impregnate with varnish

    W3 appears to be missing in this procedure. Can you please fix the Winding procedure ?
    Can you please verify the procedure again ?

    Thanks,
    Manu
  • Hi Manu,

    I have discussed with Ram. There was small typo in the line 8 - "Wind 4 turns of (W2) starting at pin 9 and ending at pin 10". This is W3. This was correctly mentioned in the table.

    Thanks,

    Manu Balakrishnan

  • Hi Manu,

    Thanks for the confirmation. That was one thing that puzzled me, wanted to confirm with Ram on that, before jumping into it.
    Additionally, shouldn't it be: "• Wind 4 turns of (W3) starting at pin (10) and ending at pin (11)" instead ?


    Also, should the very first 2 layers of W1 consist half the turns (20) or should it be whatever that can be would in 2 layers ?
    (ie the secondary sandwiched between exactly half the primary turns. If there is a difference between the 2 sandwich ends, would it result in some sort of an imbalance ?)

    One more query that I have:
    W2 and W3 comprise of 6 strands twisted together. Is a litz wire recommended with a specific number of twists per feet, or is it sufficient to have regular wires twisted together ?


    Thanks,
    Manu
  • Hi Manu,

    With pin 9 & 10 connected together, W3 will be between pin 10 and 11.

    For winding pattern, wind 20 turns (half) of primary and then secondary and then remaining 20 turns to get better coupling. If you have more space in first 2 layers, please spread the wires. The other options could be tried out but not sure about the effectiveness of coupling, which you may have to try out.

    You can go for regular wire twisted together.

    Please work with the vendor to achieve the best coupling and thermal performance.

    Thanks,
    Manu
  • Hi Manu,

    Will do that. Much appreciated.

    Thanks,
    Manu