2
\$\begingroup\$

We‘re designing a buck converter that should convert 600VDC down to 20V. For that purpose we use the ST VIPer26k.

We’re currently looking for an inductor (L2). The inductance is 1mH, but we‘re not sure about the voltage rating.

Does the inductor really have to be rated for the full 580V? The highest voltage rating I found at Würth was 400V.

enter image description here

(Datasheet, Figure 29, p.21)

\$\endgroup\$
1

3 Answers 3

3
\$\begingroup\$

While they're uncommon, off-the-shelf inductors specified for high voltage operation do exist. (No affiliation with Coilcraft, I've just used their parts and know that they make these high voltage inductors.)

Another option would be to use a different converter topology that uses a transformer instead. Voltage ratings are more frequently specified for transformers than for inductors, both as primary-to-secondary and winding-to-core ratings. This would mean reworking your design, but I believe the VIPER26K can be used for flyback converters, so you can still use the same chip.

And of course there's always the option the other answers suggest, of winding your own inductor. You can get cores from a variety of companies, and if you wind with appropriately high-voltage rated wire instead of common magnet wire, you can make a multi-kV rated inductor fairly easily. The thickness of the wire insulation will mean your inductor will be significantly larger than one wound with magnet wire, however. Make sure you can actually fit all the windings you need into the core you choose! Here's a website that might help with determining that--I've used it for such calculations in the past.

\$\endgroup\$
2
\$\begingroup\$

Yes, or preferably so.

Note that the value given, is almost always winding to core (or exterior surface), not wire-to-wire. There's no way to test (hi-pot) that; it has to be done at frequency, and then real power is needed (to work against the component's impedance). It can really only be done as a matter of design, choosing wire of adequate ratings (grade of enamel).

Having been through this question before, I'm afraid I don't have much help to offer: either flip through a lot of datasheets, or ask manufacturers for comment.

As I recall, the last comments I got were "go ahead, don't worry about it". I wasn't exactly left brimming with confidence...

Though I will note, that was probably with larger inductors than you'll have, and thicker wire tends to have thicker enamel. You may have trouble getting a smaller inductor rated for as much (again, if rated at all). But it's the same problem regardless, and it's not like I've said anything concrete anyway.

If nothing else, a custom design, with tape between layers, or adequate grade of wire (TIW even), isn't terrifically painful. Seems a lot to bother for an otherwise off-the-shelf component, but, if you need design assurances, that's one way to get it.

\$\endgroup\$
1
\$\begingroup\$

I've never seen a voltage rating for an inductor. And I don't know what the voltage rating means for WE's products. For high-voltage applications, there's a risk of arcing i.e. from one terminal to the ferrite core and even maybe to the winding internals. This must be what is meant.

So,

Does the inductor really have to be rated for the full 580V?

to reduce the arcing risk, yes.

If you can't find an inductor from a commercial provider then you can design your own inductor and apply proper insulation, such as

  • keeping the terminals and wire-to-leg solder joints away from the core
  • covering the core with insulating materials such as Kapton tape
  • using double or triple-insulated wires (this is not for wire-to-wire insulation but wire-to-core)
\$\endgroup\$

Not the answer you're looking for? Browse other questions tagged or ask your own question.