EMRFD Message Archive 11633

Message Date From Subject
11633 2015-09-13 12:10:28 rspanbauer Amidon/Micrometals AL value derivation

My son is starting out on his BSEE, and we were working on winding/measuring toroids.  He wanted to connect his recently acquired EM knowledge, and we set off down the rabbit hole trying to figure out exactly how the various toroid mfg compute their AL values that appear in the design sheets.  Specs on mean magnetic path, cross-sectional area and such are easy to compute.  But there seems to be some other process that is baked into AL values.  If I think about the various approximations to toroid self inductance, the u value is what differs between the value mixes.  So one might expect AL to be proportional to the ratio between the mixes, for any given core size, but looking at the spec sheets, it seems easy enough to find examples where this isn't true.


We're not quite throwing in the towel yet, but has anyone in the EMFRD group worked through how Amidon comes up with their published AL values?


I suppose the key question is whether this is a matter of getting the right approximation, or if the toroid mfg base their AL on observed, statistical characteristics of samples coming off the mfg line.


Tx

Rick Spanbauer


11634 2015-09-13 14:57:19 Graham / KE9H Re: Amidon/Micrometals AL value derivation
Rick:

I am doing this from memory, but you can download the engineering applications handbook for Micrometals from their website.
I think they empirically measure the permeability of their cores with ten turns of wire, and report that. 
The tolerance on the permeability of the materials is typically plus/minus ten percent, except for -10 material which is plus/minus five percent.
There are capacitive parasitic effects which keep the turns equations from being exact, and if you do a multiple layer
winding, the capacitive parasitics almost throw the turns equations out the window.

If you wind a toroid, you can squeeze and spread turns to trim the inductance value.  The equations take none of that into account.

There is also a small amount of permeability shift with frequency for powdered iron.  Severe for ferrite.

I have found that you use the equations to get you into the ballpark for values.
Then measure the inductance AT THE FREQUENCY YOU WILL BE USING IT AT.
(The value you get from the 100 kHz LRC meters is accurate at 100 kHz, but only an estimate of the value at HF.)
Then calculate an effective permeability at that frequency, and correct to the closest value to desired you can get to with the integer turns restrictions.

Then wind multiples in that configuration, and measure the spread in inductance values, as a variance input to your design.

--- Graham / KE9H

==

11635 2015-09-13 19:35:42 kb1gmx Re: Amidon/Micrometals AL value derivation
They wrap wire and measure them.  

AL varies with a lot of parameters for the same physical core.
But AL is fairly closely linked to u-effective.  Higher the u (mu)
the higher the inductance per turn (AL).


Allison/Kb1GMX
11636 2015-09-13 20:20:55 kerrypwr Re: Amidon/Micrometals AL value derivation

The Fair-Rite catalogue is an excellent reference; it may be downloaded from their web site.

There are also a number of application notes on that site.

A great article by Carole Parker of Fair-Rite is here;

incompliancemag.com/article/using-ferrites-to-suppress-emi/

and a similar but not, I think, identical article is here;

disti-assets.s3.amazonaws.com/pscelex/files/datasheets/2007.pdf/

 

There are, in those articles, equations that tie the various elements together.

 

The Fair-Rite catalogue says that they measure Al at 10 kHz; the fixture is not described.  I would expect it to be more sophisticated than a few turns on a toroid but it's not possible to be sure.


I expect that they would use a co-axial fixture for impedance measurements; perhaps the same fixture is used for Al measurements.

 

Initial permeability, ui, is approximately equal to u’, the real component of the complex permeability, over the low part of the frequency spectrum; it is u’ that “creates” the reactive/inductive part of the complex impedance of a winding on a toroid.

 

B-H curves describe the magnetic behaviour of ferrite (indeed all) magnetic materials.

 

Initial permeability is the ratio of B to H (equivalent to the slope of the magnetisation curve) when the magnetising current is first applied; once magnetisation begins the permeability is the ratio of B & H at different points because of hysteresis; there is no guarantee that that ratio/slope is exactly the same as at the zero point of the B-H axes; it is only approximately so.

 

The imaginary component, u”, is a measure of the loss that is expressed in the complex impedance of the winding as a resistance.

 

This may seem “back-to-front” to the way that we usually think of complex impedance but the world of magnetics is like that.  J

 

Here is the complex permeability data for #43 material;

 

 http://s13.postimg.org/x3qnvzlkn/43_Ferrite_New_Version_u_800.gif

 

The plot does not extend down to 10 kHz so we do not know the value of u’ at that frequency.

 

I think that a tolerance of +/-20% in the value of Al is usual; another reason that it’s difficult to accurately relate ui/u’ and Al.

 

Kerry VK2TIL.

 

 

11638 2015-09-14 05:43:59 rspanbauer Re: Amidon/Micrometals AL value derivation
Kerry - thanks for those links, and your commentary.  Helps a lot -- Rick