Hi Sean, It is a simple switching ferrite transformer (full bridge, PSFB to be precise), Primary 12T, Secondary 200T The core is ETD type, similar to EE but the central section is round, which makes things slightly more easier, operational frequency is between 95kHz to 100kHz The transformer is an interleaved one, the windings interleaved to improve coupling and to reduce leakage inductance. As you said, the flux is rather small and I have kept the flux still smaller, hence the larger no. turns eg: 12T and 200T otherwise it would've been still lesser. This has been done with the intent, to reduce core loss, but Cu losses increases inversely. But that is not a big issue, as the Cu is rated at 2x the current. So, Cu losses is also somewhat reduced/handled there. Yes, I agree completely to what you say and hence divided the primary into 2 parts and the secondary also into 2 parts, for interleaving; 2 exactly, because the flux needs to be balanced between the different interleaved segments. I have the unitrode documents at hand and have pondered over them again and again. One thing that confuses me always is the direction of the winding, during interleaving. Somehow, this beast seems to keep attacking me all the time. ;-) But I think, having the primary closest to the core would be more ideal as Jim says in a previous post, which makes sense. Thanks, Manu On Mon, Mar 18, 2019 at 12:48 PM Sean Breheny wrote: > > Manu, > > I am having trouble picturing this transformer. What is the core material > and shape? Are all the windings on the same part of the core, or is it an > E-I core or E-E core and some windings are one one "leg" and others on > another "leg"? > > Given the low turns count, I'm guessing that this is a transformer for a > switching power supply. Power transformers in the 10s or 100s of kHz use > very different winding techniques than 50/60Hz transformers. It is very > common to use several layers of turns in parallel for each winding and to > interleave them. The reason is not mainly to get better coupling - it is = to > reduce losses. Think of it this way - the principal path which the flux > takes is through the core material, but for a transformer, the flux is > rather small because most of the flux from the primary is cancelled by th= e > flux from the secondary. The remainder is only what is needed to produce > the voltage across the transformer terminals. So, in the space between th= e > primary and secondary windings, there is higher flux than elsewhere - eac= h > layer of primary winding increases the flux and then the next layer of > secondary decreases it. If you put all the primary together and THEN all > the secondary, you would have a higher flux in the region where the prima= ry > layers end and the secondary layers begin. This would create larger eddy > currents in the copper of those nearly winding layers (eddy currents in t= he > copper). > > The TI-Unitrode seminars (available online in PDF form) are a very good > guide to this, although I found some technical errors in them. The ones > most relevant to this are the PDFs named: > > slup126.pdf > slup171.pdf > slup197.pdf > slup200.pdf > slup205.pdf > > You can just search for these filenames and download them directly from T= I > > Sean > > > On Sun, Mar 17, 2019 at 10:44 AM Manu Abraham > wrote: > > > Replying to my own post, > > > > On second thoughts the interleaved layering would cause an imbalance, > > since not all layers are linked withe same amount of flux, which > > brings me to: > > > > L #1 100T Sec > > L #2 6T Pri > > L #3 100T Sec > > L #4 6T Pri > > > > That said, 2 questions again: > > #1. Should the Primary be in Layer #1, or is it better to start with > > the Secondary ? > > #2. Both the Primary and Secondary should be would in the Clockwise > > direction itself ? (I hope so) > > > > Thanks, > > Manu > > > > > > On Sun, Mar 17, 2019 at 3:45 PM Manu Abraham > > wrote: > > > > > > Hi, > > > > > > Looking at a transformer, primary 12T secondary 200T, step-up > > > > > > The primary is split in two, the secondary in 3, to reduce leakage > > inductance. > > > > > > ie, > > > 1. L1. Secondary 50T > > > 2. L2. Primary 6T > > > 3. L3. Secondary 50T > > > 4. L4. Secondary 50T > > > 5. L5. Primary 6T > > > 6. L6. Secondary 50T > > > > > > > > > L1 and L3 (secondary) are at opposite sides of the secondary. > > > > > > In which case isn't the direction of the flux changing ? > > > So, when L1 is wound clockwise, shouldn't L3 be wound in a > > > counter-clockwise fashion ? > > > > > > Am I thinking right, or thoughts in the wrong direction ? > > > > > > Apart from the original questions, any thoughts on a better > > > methodology to reduce leakage inductance ? > > > > > > Any thoughts ? > > > > > > Thanks, > > > Manu > > -- > > http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive > > View/change your membership options at > > http://mailman.mit.edu/mailman/listinfo/piclist > > > -- > http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .