Transforner Theory
amdx
Info below from the following site----
http://www.ee.surrey.ac.uk/Workshop/advice/coils/mu/index.html#bhcurve
Unlike electrical conductivity, permeability is often a highly non-linear
quantity. Most coil design formulć, however, pretend that it is a linear
quantity.
=========================================================
My question is-
If I wind a transformer using the specified A sub L and then use that
transformer in a receive antenna where the voltages are very small, wouldn't
I be low on the curve and cause the transformer to function poorly
especially at the lowest frequency of the design?
Mike
PS Thinking about a Flag antenna, which has a small output signal.
Roy Lewallen
Al is usually the value for low flux density. That is, it's the value
you'll have when the flux level is low. Permeability will drop from
there at high flux levels.
If you're making a broadband (untuned) transformer, you only need to
insure that the winding impedance is high enough. If you design it to
have adequate impedance at the lowest frequency, you should be ok for
frequencies above that. If you're making a tuned transformer, you'll
probably be using either powdered iron core or a ferrite core with a big
air gap in the magnetic path like a ferrite rod. Either will withstand
many orders of magnitude of flux density above what a received signal
will produce before there's any noticeable change in permeability.
The assumption of constant permeability is often a reasonable one.
Change in permeability with flux density is certainly nothing you have
to worry about in a receiving application unless you've got a lot of
turns and a lot of DC current in the winding.
Roy Lewallen, W7EL
K7ITM
cores is unlikely to be a problem for you. A while back, I had
occasion to question if some filters we use, made with small powdered
iron toroid core inductors, were causing distortion, so I built up some
similar filters with air-core coils and very good capacitors that I
knew would not distort. The result was "no change". That corresponds
in this case to third order intercepts in excess of +50dBm, which would
be considered at least pretty good by all but the fanatics for use in
receivers. I don't know how much in excess of +50dB, because that was
about the limit of what I could see in that test. Also, I know that
the broadband transformers used in the best H-mode mixers have allowed
those mixers to perform at similarly high--and higher--third order
intercepts. I suppose those transformers are transmission-line types,
wound on ferrite cores.
Cheers,
Tom
Larry Benko
> Al is usually the value for low flux density. That is, it's the value
> you'll have when the flux level is low. Permeability will drop from
> there at high flux levels.
>
Not to nit-pick but the permeability of nearly all powdered iron
formulations actually rises with increasing flux levels (AC) and then
falls off. For #26 material (u=75), the effect is very much exagerated
with the permeability increasing nearly 300% at ~5000 Gauss and then
falling very quickly. However the permeability does drop for any value
of DC bias current and larger DC bias currents produce greater
reductions in permeability.
73, Larry Benko, W0QE
Roy Lewallen
Thanks for the correction. The permeability monotonically drops with
increasing coercive force (H), but rises as you say with increasing flux
density (B) over some range of flux densities. This is true for ferrites
also.
Roy Lewallen, W7EL
amdx
"Roy Lewallen" <w7...@eznec.com> wrote in message
news:1239dl1...@corp.supernews.com...
Please see the following URL Page 6,
http://www.mag-inc.com/pdf/cg-01.pdf
note the graph for the toroid, the inductance decreases by 40 percent
going from 2000 gausse to 10 gausse.
Question 1.
I don't know where on that graph the published permeability would set the
inductance. ( to clarify--How many gausse is used to measure permeability
and set AL?)
Question 2.
Can anyone take a stab at how many gausse in a typical FT140-43 toroid
with 8 turns on the secondary, and 34 or 35 turns on the primary used on a
flag antenna with a low level signal.
Maybe if we have two points on that graph we can have a real number to
see how much inductance changes from published AL at low gausse.
Mike
PS. interesting how pot cores have very little inductance change with
changing gausse.
w8...@akorn.net
amdx wrote:
> If I wind a transformer using the specified A sub L and then use that
> transformer in a receive antenna where the voltages are very small, wouldn't
> I be low on the curve and cause the transformer to function poorly
> especially at the lowest frequency of the design?
>
> Mike
>
> PS Thinking about a Flag antenna, which has a small output signal.
Mike,
I think you are focusing on non-issues and not consider things that are
really important.
First, I would not use a 43 core on a low frequency receive antenna.
This is especially true with an ungrounded antenna that has
exceptionally low signal output, like a Flag.
There are very few antennas in the world that are perfectly UNbalanced
or perfectly balanced. Even what we consider an unbalanced antenna can
cause feed system problems. When the antenna has very low signal
output yet occupies a large spatial area, you are especially looking at
problems.
The flag has low common mode impedance, and fairly high differential
mode impedance. It is neither balaunced nor unbalanced, it is in that
soupy world of something that requires equal and opposite currents at
the feed without perfect voltage balance. It is not a balanced antenna,
and not an unbalanced antenna.
When that is combined with the very low signal output, you have to pay
particular attention to the transformer design.
You really can't use a transmission line transformer because it will
not have enough isolation. You need a primary-secondary transformer
with isolated and slightly seperated windings. You really don't want a
material that requires 30 or 40 turns, because extra wire will increase
stray capacitance from primary to secondary.
You almost certainly want to move into a binocular core with fairly
high permeability at the lowest frequency, like a 73 material. Unless
you have a few volts of RF from a closeby station, flux density is not
an issue.
You want to keep primary/secondary capacitance down near a dozen pF or
less if possible, and have NO direct path for common mode currents.
http://www.w8ji.com/k9ay_flag_pennant_ewe.htm
73 Tom
amdx
<w8...@akorn.net> wrote in message
news:1144326826.7...@g10g2000cwb.googlegroups.com...
Thanks for your input, When I wrote the above I couldn't remember what
people were using on there flags so I went to one site and just copied what
that site said, maybe it was a misprint and should have said 73.
I used a 3F3 4229 pot core with 3 turns and 13 turns
on my flag, I cut a styrofoam clamshell to go between the sec and pri to
limit capacitance. I also put a grounded electrostatic shield between
windings.
I was happy with my nulls, (whatever that means;-) but wonder how efficient
my transformer was.
My questions still stand,
Question 1
How many gausse is used to measure permeability and set AL?
Question 2.
Can anyone take a stab at how many gausse in a typical FT140-43 toroid
with 8 turns on the secondary, and 34 or 35 turns on the primary used on a
flag antenna with a low level signal.
On the second question the material can be modified to reflect the material
and turns as needed.
Thanks
Mike
Roy Lewallen
> . . .
> My questions still stand,
> Question 1
> How many gausse is used to measure permeability and set AL?
Essentially zero.
> Question 2.
> Can anyone take a stab at how many gausse in a typical FT140-43 toroid
> with 8 turns on the secondary, and 34 or 35 turns on the primary used on a
> flag antenna with a low level signal.
I won't bother to calculate it because the change in permeability would
be so small you wouldn't be able to measure it. This is a non-problem;
you're wasting your time worrying about it.
> On the second question the material can be modified to reflect the material
> and turns as needed.
If your circuit is sensitive to a change in a few parts per million of
permeability, it has serious problems. The permeability will change
several orders of magnitude more than that with modest changes in
temperature.
Roy Lewallen, W7EL
Richard Clark
wrote:
>Question 1
> How many gausse is used to measure permeability and set AL?
Hi OM,
You will never in your lifetime escape the bare minimum of ½ Gauss
presented by the Earth's magnetic field.
73's
Richard Clark, KB7QHC
amdx
> amdx wrote:
> > . . .
> > My questions still stand,
> > Question 1
> > How many gausse is used to measure permeability and set AL?
>
> Essentially zero.
Ok, So I'll use the figure of 1 gausse as where permeability is measured and
from there I can assume the inductance increases 40 percent at 2000 gausse
for the toroid specified on the Magnetics webpage. The only info I have is
from
Ferroxcube Soft Ferrites and accessories 2000 data book, and it simply says
"The initial permeabilty is measured------ at a very low field strength."
> > Question 2.
> > Can anyone take a stab at how many gausse in a typical FT140-43 toroid
> > with 8 turns on the secondary, and 34 or 35 turns on the primary used on
a
> > flag antenna with a low level signal.
>
> I won't bother to calculate it because the change in permeability would
> be so small you wouldn't be able to measure it. This is a non-problem;
> you're wasting your time worrying about it.
I'm not worrying, just curious, since I've been using a large potcore to
deliver microwatts that at one time I used at near a kilowatt. Just
wondered
if we were losing some low frequency response because of a change in
permeability. It seems as though the permeabilty measurement is made nearer
the power levels of our receive antenna signals.
>
> > On the second question the material can be modified to reflect the
material
> > and turns as needed.
>
> If your circuit is sensitive to a change in a few parts per million of
> permeability, it has serious problems. The permeability will change
> several orders of magnitude more than that with modest changes in
> temperature.
>
> Roy Lewallen, W7EL
Thanks Roy,
I appreciate the discussion and information.
Mike
Peter O. Brackett
[snip]
> PS. interesting how pot cores have very little inductance change with
> changing gausse.
[snip]
Most, if not all, pot cores used for "precision" inductors and transformers
come in matched pairs with a very accurately ground narrow air gap between
the two center posts. The Al (or alpha from some mfgs) is accurately set by
the width of the air gap as ground during manufacture. The air gap also
forms a large part of the overall magnetic path (air having a much larger
reluctance than the ferrite). Such pot cores for precision inductors
usually have an adjustable slug (set with a non magnetic screwdriver) to
allow the finally assembled inductor to be set to an "exact" value.
And so... pot cores intended for use in making precision inductors (as in
filters or delay equalizers) or precision transformer applications exhibit
little change in inductance over a wide range of conditions simply because
of the air gap.
--
Pete k1po
Indialantic By-the-Sea, FL
Roy Lewallen
time back I learned a bit about the ferrite manufacturing process. It
turns out that the green ceramic shrinks by something like 20 or 25
percent during the firing process, with poor control over the amount of
shrinkage. The gaps in gapped pot cores, I found, were only nominally
the specified gap width -- each pot core was individually ground to meet
the Al specification, not some specification on air gap size. This isn't
important in using them, but I found it interesting.
What you get with an air gap in trade for effective permeability, is
much greater independence of Al from the core characteristics (material
permeability and physical size) and therefore much greater stability
with regard to temperature and flux density, and much greater flux
density capability.
Roy Lewallen, W7EL
Peter O. Brackett
Yes, that's why "gapped" cores must be used for precision [inductance] work.
The properties of the basic core materials are too difficult to control
during manufacture and hence result in manufactured pieces of wide
variability. In addition the basic core material properties exhibit a wide
variation under environmental variations such as temperature, pressure,
etc...
Inserting an accurately calibrated air gap in the magnetic path, by grinding
the pot core center posts to a specific Al or alpha, accurately regulates
the overall reluctance of the magnetic path and overcomes both of these
variable effects, it alsow and allows the engineering of high performance
"precision" inductors. In addition the air gap also mitigates a lot of the
non-linear effects noted at higher levels of flux density.
Gapped cores are "de riguer" for precision work. Ungapped cores are only
used for "sloppy" inductance work. This includes many applications of
transformers as well.
--
Pete k1po
Indialantic By-the-Sea, FL
"Roy Lewallen" <w7...@eznec.com> wrote in message
news:123jk90...@corp.supernews.com...