1
\$\begingroup\$

I want to wind my own ferrite core inductors for some power conversion. However, I am having trouble finding out when the cores will saturate from manufacturer's data. For example, this material datasheet gives me "Flux density (near saturation)", which is about 300mT (what I expected). Let's call that \$B_{SAT }\$ or something.

Now, sometimes I can't find this spec, but there are a few that are suspiciously close. For example, this page tells me the 61 material has "Flux Density @ Field Strength" of about 250mT when H=15Oe (about 1200A/m) is applied. It seems this is a suggested \$B_{MAX}\$? I don't know. Here's another table with several values for "Flux Density @ Field Strength", and although the maximum flux density seems about right, the field strength values are all suspiciously round.

So, my question is: in cases like this, in which no explicit "field near saturation" parameter is given, what maximum value of the magnetic field should I use in order to calculate the maximum current in the inductor? Do I just rule of thumb 200mT, or is there some way of inferring something from other specs?

\$\endgroup\$
  • 1
    \$\begingroup\$ I'm just a hobbyist, so I don't have to worry about 100k units where a penny would matter. So for ferrite I (1) first-off make sure to use the datasheet specs; and then, (2) almost always try to stay under \$150\: \text{mT}\$ just as a rule. Type 61 (\$\mu_e=125\$) is excellent, by the way, and I use it for about half of what I make by hand. I also use \$\frac{l_e}{N}\ge\frac{\mu_0\:\mu_e\:I_\text{peak}}{B_\text{max}}\$ (path length) and \$f\:A_c\:N\:B_\text{max}\ge\frac{V_\text{ON}\:V_\text{OFF}}{V_\text{ON}+V_\text{OFF}}\$ (area) as discontinuous guidance, as well. \$\endgroup\$ – jonk Jun 13 '19 at 23:20
  • \$\begingroup\$ For freq < = 100kHz type 78 material is better lower winding resistance and much higher permeability fair-rite.com/wp-content/uploads/2015/06/Perm-vs-Freq1.jpg \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Jun 14 '19 at 1:30
  • \$\begingroup\$ Top of page 4 of the data sheet - open your eyes! \$\endgroup\$ – Andy aka Jun 14 '19 at 7:14
2
\$\begingroup\$

Saturation isn't a hard line, but increases as the H field increases. In the graphs below, where would you place the saturation line? It also varies with temperature, which further complicates things. I would place the lower bound at 200mT (if it's the lower bound that matters in your application.)

enter image description here

|improve this answer|||||
\$\endgroup\$
2
\$\begingroup\$

There are two main things that set the maximum field for the magnetic core of an inductor, and we have to respect both of them

a) saturation - leading to reduced inductance
b) power dissipation in the material - leading to excess core temperature

In ferrites, now that we have nice fast switching devices, it's generally dissipation that limits you first, so saturation is rather less important than it would be with, for instance, an iron core operated at mains frequency.

Generally, low loss ferrites have rather soft BH curves, with no 'knee' to put a single number on. The reason your datasheets give you field at a given H is that to get that field, you need to supply that H, and the winding dissipation rises as H^2 (as I^2), so there a large penalty in pushing further round the BH curve.

If you find your design is asking you for 'good' value for saturation field to get your volts per turn up, then try doubling the design frequency, and instantly you get twice the power throughput. When you check the maximum field versus frequency figures, you'll probably find the field has dropped by less than a factor of 2.

|improve this answer|||||
\$\endgroup\$
0
\$\begingroup\$

If you don't have a nice graph to go by, then figure that the highest number that the manufacturer mentions is probably the highest that you want to go.

(This rule of thumb applies to any data sheet, BTW -- not just inductors, or even electronic parts).

|improve this answer|||||
\$\endgroup\$

Your Answer

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

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