There are online designs that use a simple ferrite bolt 50 turn primary and 500 turn seccondary powering one 5 watt CFL. I attempt to customize it with:

Specs: TDK Core



  • Vin primary= 6V DC

  • Core material PC40

  • Saturation at 380 mT


  • Power output 5,5 Watt

  • Nominal voltage = 34V

  • Current nominal = 0.18A

  • Breakdown voltage: Unknown

Transistors: Arbitrary

Question 1: Is just taking half of the values, so 26 turns primary and 260 turn seccondary adequate or bad practice for my PC40 material core? Or propper approximation? Is your calculus more efficient?

Because my two seperate calculations innitially yielded 4 and 5 turns for primary. I thought this can't be right. And the variables have to be estimated guessed, and I'm pretty bad at that.

Question 2: Can you come up with a better more efficient design by calculus? Or is 4 or 5 turns primary pretty close to optimal here or should I go with 26?, or YOUR value? Can your calculus yield a more efficient N1/N2 for the given specs?

I know that primary inductance leakage plays much less of a role in cfls than in other circuits, but:

Question 3: Is it the primary inductance leakage which causes the voltage spike on the secondary when the load circuit side opens by a glow starter, or even without a starter, when it's normally closed? Or what inductance specifially cause the voltage spike to overcome breakdown voltage?

Thanks for taking a look at this community! I believe together we can solve many electrical problems!

  • \$\begingroup\$ Welcome to EE.SE! "estimated guessed" That won't work. You need to know your switch frequency, max on-time at max input voltage and core area to get started. Calculate according to Vt=NAB. Also, please show your schematic. \$\endgroup\$ – winny Jul 2 at 11:24
  • \$\begingroup\$ Thank you winny for editing the messy original post! duty cycle =40%. The Ae is 51,8mm², and I strongly suspect core is overkill for a 5,5W load capable of max. 70W forward mode? And that could be one reason why I get 4 or 5 turns prim. with the formulas, because of very high cross Area of 51,8²? with a very low inductance primary 0,604uH, secondary 43,671uH. The Fsw (switching freq.) I chose to be 300khz, but it's purely copycatted, edmucautiously guesstimated. I might as well choose 25khz and by chance get more turns? Though I did put 300khz into the fromula as exampled in the guide. \$\endgroup\$ – rantoniusw Jul 2 at 11:52
  • \$\begingroup\$ Hello winny, thank you for your answer. Here is how I got the 5 turns, t=Vout/2/(Ae in m²)/300khz/0.38T; t=6V/2/0.00000518m²/300khz/0.38T; t= 5,08 I chose maxiumum field 2x because of a flyback transfomer. What is incorrect here winny? I chose 300khz because I have to guess a freuency, I have no ide at what freuency a flyback like this would swing at. I'd like to hear your professional advice if you want to help me. I will try to give you a schematic to have a look at, since you want one, but I don't know how it works here. \$\endgroup\$ – rantoniusw Jul 2 at 12:09
  • \$\begingroup\$ Hello winny, I added the rquested schematic of you to the original post. Here again: duckduckgo.com/… \$\endgroup\$ – rantoniusw Jul 2 at 12:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.