0
\$\begingroup\$

every one, I am an automation engineer, not an electrical, but I have some knowledge about electricity (I hope). I am designing a power supply with input voltage of 3,1V and the output voltage of 175V (10mA max). I am using push-pull configuration like on the figure below.

enter image description here

This schematic requires custom transformer. I am pretty much limited with free space within enclosure. As a “real” engineer to make this power supply I first tried to brute force the problem. After few failed attempts I have a working prototype on the breadboard. I winded the transformer from an inductor. It functions approximately at 230 kHz (I know, 230 kHz and breadboard is complete nonsense and stupid). That is why probably I have so much lose and terrible energy conversion efficiency, as well as because the transistors are getting very hot(I believe they are running at their maximum).

In order to make everything properly this time I read a book “TRANSFORMER AND INDUCTOR DESIGN HANDBOOK Third Edition, Revised and Expanded “ by COLONEL WM. T. MCLYMAN. Link to the book Link to the book . If I understand correctly in order to decrease the size I should to increase frequency even more. So I decided to switch to MOSFETs. The schematics with MOSFETs looks as follows.

enter image description here

I have Si7940DP laying around, so I will using it. Also in order to increase energy conversion efficiency I want to wind a proper transformer. I have made calculations (version in google sheets with formulas) for that. Calculations have been made according to the chapter 7 from the book.

Can somebody verify my calculations and say to we that I am not dreaming or vice versa. That I can reach such frequency with MOSFETs and transformer Also there are few sings that I do not understand completely, they are marked with red. There are: 1. How to choose “Operating flux density” and how it affects the design? 2. And how to choose “Magnetic material”, what Permeability (L - 900, R - 2300) suits better?

Hate and critics are welcome. Let me know if additional information required.

Thanks for help, Arttu

\$\endgroup\$
8
  • \$\begingroup\$ Typically you would not use a forward direct transformer for this type of supply. It's more suited to a flyback (stored energy) design. There is lots of information on these, but TI have a reasonable starting design guide: ti.com/lit/ml/slup127/slup127.pdf \$\endgroup\$
    – Jack Creasey
    Commented Apr 6, 2018 at 20:12
  • \$\begingroup\$ OP: Your quote: "If I understand correctly in order to decrease the size I should to increase frequency even more." This is true, however the price you pay is increased core loss, and at 340kHz, you're getting into the range where skin effect can create additional losses. Look at the Magnetics datasheet you referenced for core material R, Loss vs. Frequency. \$\endgroup\$
    – AlmostDone
    Commented Apr 6, 2018 at 22:35
  • \$\begingroup\$ @JackCreasey I give it a read, thanks. What can be an approximate efficiency for it when converting from 3V to 175V? \$\endgroup\$
    – Arttu
    Commented Apr 8, 2018 at 20:23
  • \$\begingroup\$ @AlmostDone I am sorry, not quiet sure I understood you. I referred to core material L (link), and it shows tested conditions at 1000 kHz 30 mT and 3000kHz 10 mT, while core material R (link) test conditions are at 100 kHz 100 mT and 100kHz 200 mT. SO, as far as I can see core material L suits better, since I will not be exiting the test conditions at 340kHz and use it at more higher magnetic flux density of 50mT instead of 30mT from test conditions. Right? \$\endgroup\$
    – Arttu
    Commented Apr 8, 2018 at 20:38
  • \$\begingroup\$ Agree, L looks better than R for core loss spec; similar Bsat. For 300kHz it is the better material. \$\endgroup\$
    – AlmostDone
    Commented Apr 8, 2018 at 21:07

0

Reset to default

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.