1

Initially I need an inductor to be used in cheap low-power AC-DC power supply. I'd like to use LNK304 solution:

enter image description here

When I reached 1mH coil I found Bourns 77F coils: very cheap (only 8 cents at some quantities), quite small (11 mm in length). But I need to take into account that this coil should withstand peak AC voltage which can be as great as 350 V (if I consider +10% ACV tolerance). There is a Dielectric Strength parameter in it's specification which is rated for 500 Vrms.

However there are dedicated high voltage solutions like Wurth Electronics HV series. But it is bigger and significantly more expensive (about 2 USD).

Main question: Can I be sure that this cheap coil will have a long and happy live in my schematic?

Community
  • 1
Roman Matveev
  • 2,942
  • 7
  • 32
  • 75

1 Answers1

2

If you go to the Power Integrations site, (specifically the LinkSwitch-TN) there are several design examples given. Each design has a Bill of Material. I checked one and it listed the parts recommended for all the components in the circuit including all the inductors in the design: -

enter image description here

Andy aka
  • 434,556
  • 28
  • 351
  • 777
  • The question was mostly about inductor coil voltage strength, not about LNK304 implementation. However thank you anyway for pointing me to this huge source of reference designs - I didn't know that. – Roman Matveev Jun 02 '15 at 14:21
  • @RomanMatveev Each Bill of material will give you a bigger number of options for both (all) inductors used in your specific design. I recommend that you sift thru a few of these and get a feel for this (these) components. I felt that information that the manufacturer recommends for these inductors is better than trying to engineer a philosophy on what is right or wrong. – Andy aka Jun 02 '15 at 14:24
  • If you want to know how long the coil will last there is the exceelant document entitled MIL-HDBK-417F. You can download it freely and look up the reliability of inductors. The formulas in the document are quite accessible and you will be able to derive an answer of MTBF that is based on literally thousands of failure events. Give yourself 4 hours to digest the methodology though because you will likely be facing a couple of concepts you haven't encountered before. – Andy aka Nov 20 '15 at 11:28
  • Thank you! Are you sure that it is MIL-HDBK-**4**17F not [MIL-HDBK-**2**17F](http://www.sre.org/pubs/Mil-Hdbk-217F.pdf). 217 is actually about the thing you're talking about. 417 I wasn't able to find :( – Roman Matveev Nov 20 '15 at 13:26
  • yeah it's 217. Silly me. – Andy aka Nov 20 '15 at 13:49
  • I never fail to be impressed about how relevant it is despite it's age (a bit like me LOL). Were you able to get a failure rate figure dude. – Andy aka Nov 20 '15 at 16:35