Post Go back to editing

LT3825 Flyback Transformer Inductance - Eval Schematic

Hi All,

I'm working on an LT3825 design for a 60W 12V output PoE PD. I'm using the 894C-C eval schematic as a reference. The schematic uses a Pulse PA1736NL transformer, which has a 20.5uH primary inductance. Using the LT3825 datasheet calculations this seems a very low value and would result in a very high ripple current when operating at the 200KHz. My calculation comes up with something around 170uH. 

I'm not that familiar with flyback converters, but I would think that the primary side of the transformer should work pretty much like a straight buck converter. Can anyone help me out here - the Eval board obviously works OK, so I must be missing something basic about how the transformer works?

Any help much appreciated.


  • Hello Julian, There are multiple solutions for a Flyback. The best advise is on page 15 of the the LT3825 Data Sheet which suggests "A simple spreadsheet program is useful for looking at trade-offs." For a transformer, N * A * B = L* I. In this case, 170uH would require 5 to 6 times more turns (N) or core area (A). Everyone likes (their own) rules of thumb. For a Flyback, I check the ratio of the peak secondary current to the output current. In this case, with 20uH, this is about 15A / 5A ( = 3:1) which isn't too high. Good Luck.

  • Hi Kurk,

    Thanks for your reply. On your advice I have started to build a spreadsheet for all the calculations and it does shed light on things. I have also downloaded a complete LTSPICE simulation and played around with that looking at transformer currents etc. Really useful for understanding these things and remarkably accurate.

    For me it's a new area, I'm very happy with boost and buck converters, I just need to get more experience and as you say get some rules of thumb for the isolated converters.