EMRFD Message Archive 12876

Message Date From Subject
12876 2016-05-26 15:14:11 Clutter Characterising Unknown Ferrite Beads?
Hi, All-

I have acquired a goodly number of medium size ferrite beads.
The mix is unknown. To get a sense of what I might have, I cut
about 15in of enameled wire and then cut that exactly in half.
After scraping off the insulation on all ends, I wound two near-
identical size coils, one an air coil and one around a bead.

The coils ended up being 3-1/2 turns (air) or 4 turns from a
'toroid perspective'. On my little AADE meter, I get about 0.168uH
for the air coil and about 16uH for the bead coil. Several repeated
measurements have the inductance ratio consistently around
90-95. Looking at tables of data for the various ferrite mixes,
I see materials with the following approximate initial permeabilities:

"68", ui=20
"67", ui=40
"61", ui=125
"33", ui=600
"43", ui=850
(etc.)

My first guess is that I have mix 61 material, ui=125.
Also, the degree of sheen on the material matches well
with the appearance I've seen over the years for known
#61 ,material. What do you guys and gals think? Am I
on the right track?

BTW these beads are 0.375in OD, 0.200in ID, 0.575in L.
They look to be plenty useful for low to medium power
baluns and transformers at HF.

Thanks, 73

David K3KY
12877 2016-05-26 17:18:59 kerrypwr Re: Characterising Unknown Ferrite Beads?
The most reliable way to identify unknown ferrites is to find the frequency at which Q=1.

If you look at the curves of u' and u" -v- frequency in the Fair-Rite or other maker's catalogue, there is a point at which the two curves cross, ie are equal in value; that is the point at which Q=1.

This is the curve for #61;

http://s33.postimg.org/viwn6div3/61_Ferrite.gif

 



This, however, requires an impedance analyser or a VNA.

I haven't thought it through but a Q-measurement setup as described in EMRFD might be applicable; tedious to find the Q=1 point though so perhaps not practicable although, if initial testing suggests #61, some measurements around 40MHz could be made.

Your method is probably the best that can be done in "guesstimating" the mix with limited test equipment; it's more usual to work-out the Asub L and compare that with the figures quoted by the maker for the same core size but your method seems equivalent to that.

Whilst Fair-Rite is probably the company best-known in our world, there are dozens of makers of ferrites so unknown cores may not be directly comparable to Fair-Rite data.

You are right about the colour; in my experience #61 has a dull finish whilst #43 is shiny.

The u' & u" curves for #61 indicate that it makes good inductors up to about 7MHz; in that region u", the loss or resistive component of the complex permeability/impedance, is close to zero so good Q is achieved.

At higher frequencies u" begins to climb whilst u' remains relatively constant; transformer (flux-coupling) operation is good until Snoek's Limit, the ferrimagnetic resonance frequency seen as a little "hump" in the u' curve, is reached.

After that, u' drops until the ferrite "disappears" in a magnetic sense.

Snoek worked-out in the 1930s that there is a fairly simple relationship between the value of u' and the frequency at which it begins to fall; that is why high-permeability materials are only useful at low frequencies, eg power supplies.

http://s33.postimg.org/itpy3lnen/Snoeks_Limit_2.gif

 



PS.  Fair-Rite changed the mix of #43 a few years ago; the current mix has u' of 800 and the curves are a bit different.  Their current catalogue is a mine of information and may be downloaded from their website.

Kerry VK2TIL.
12878 2016-05-26 19:00:08 Clutter Re: Characterising Unknown Ferrite Beads?
Thanks, Kerry- great information! I'm pretty sure my guess is right about
type 61 mix. I'm going to find my long form Fair Rite catalog and see if
I can find a candidate #61 bead in there in the same size. If it gives an
A-sub-L value, I bet I can get close to that figure with my cores.

BTW for inductors I would instinctively turn to powdered iron toroids above
~5MHz or so- but ferrite cores would be great for wide band transformers
and baluns over upper HF especially. Using them as power supply RF
chokes in some cases also springs to mind. At 0.375 dia x 0.575 L, they
should handle substantial current...

I think I'm now well fixed for experimenting, having a bunch of these cores
to play with. I can picture lots of useful transformers using bundles of
2, 4, 6, or even 8 of them, side by side (that's getting up around 2.3in
long)...

73, David K3KY




I haven't thought it through but a Q-measurement setup as described in EMRFD might be applicable; tedious to find the Q=1 point though so perhaps not practicable although, if initial testing suggests #61, some measurements around 40MHz could be made.

Your method is probably the best that can be done in "guesstimating" the mix with limited test equipment; it's more usual to work-out the Asub L and compare that with the figures quoted by the maker for the same core size but your method seems equivalent to that.

Whilst Fair-Rite is probably the company best-known in our world, there are dozens of makers of ferrites so unknown cores may not be directly comparable to Fair-Rite data.

You are right about the colour; in my experience #61 has a dull finish whilst #43 is shiny.



Kerry VK2TIL.
12879 2016-05-26 20:06:43 kerrypwr Re: Characterising Unknown Ferrite Beads?
The dimensions you gave are close to, but not exactly the same as, Fair-Rite 26XX375102 beads.

If the dimensions don't match exactly, within tolerances, they may not be Fair-Rite; that complicates things as other makers' materials are often different to Fair-Rite's materials.

I have catalogues or data from about thirty ferrite-makers; some make materials that are quite similar or identical to Fair-Rite's, others make different materials.

But there is usually a "family resemblance"; eg, a material with ui of about 100-150 usually behaves similarly to #61 for most purposes.

#61 works fine for inductors at low frequencies; for example, AM radio antenna rods are #61 or something similar.

Your beads are rather long for transformers.

The lower frequency of a transmission-line transformer is set by its inductance; the high frequency limit is set by the winding length.

(Note that the "limits" are not set in stone; they are more like a "range" than a fixed frequency).

Good transformer performance comes from toroids or, even better, from beads that are a little longer than a toroid; examples might be an FT50A-61 toroid or its 2x length cousin the FT50B-61.  Binocular cores also work well.

The FT50A or FT-50B shape and the binocular shape give a good balance between high inductance and a short winding; "tubes" such as you have require a longer winding to give the same inductance that would be provided by a shorter winding on an FT-type core.

Conversely, a shorter winding on a "tube" will provide less L at low frequencies.

Try some transformers and see what their frequency range is; if you don't require the widest-possible range, you should be able to use them.

#61 is useful for transformers that are, by their nature,less-than-perfect; whilst its u' is low, it holds-up over a wide frequency range whist its u", which gives loss, is not very high so there is less heating.

For instance, 9:1 autotransformers work well at high power with #61; they would often "cook" if "lossier" #43 were used.

#43 is suited to small transformers such as are used for impedance-matching in small-signal and low-power amplifiers; the u' is about six times that of #61 and this "concentrates" the energy transfer.  The fact that its u" is high is offset by the small amount of net core flux that is a benefit of the high u'.

Philips app notes ECO6907 and ECO7213 are the best RF-transformer references that I know-of; Google will find them.

Kerry VK2TIL.