I think you will understand the problem better,
if I generate a small table.
60hz 400hz values
1.125 .75 core size needed in inches
441 150 turns needed for gauss
4.6 .91 avg resistance of wire
98 96 precent of use of bobbin area
3.245 .973 total weight (includes wire)
9.250 1.9 wire loss watts
2.016 10.3 watts loss in the mag. material
14.1k 13.9k gauss
The only items I spec were fc (60/400),
voltage (115), current (1A), and after the computer
pick was made, I had to change the number of turns to get
the flux denisity about the same.
This material should be run at about 12k gauss at
this woltage level. Mil spec ,I would try to do
10k or 11k to be able to go 30% higher in voltage.
This is for EI cores with a sguare stack,
(center leg .75 wide times .75 high).
Geting back to real designs, I would not
use either of these basic designes. too much
loss (temp rise, and too near the Bsat limit).
To answer the statement I made about
peak current losses, from the table you can
see that there is about a 4 to 1 ratio
in resistance (for the winding (copper loss))
which means a 4 to 1 ratio in
watts loss for the winding. This applies to rms as well
as pk currents.
I need to inject this, it may not flow
but it is revelant nowdays ( always with mil spec).
In a ps that converts ac to dc, there is a
concept called "power factor".
the best expresion is "real power/ apparent power".
For simple terms, a current conduction of 60% is
"NOT" a power factor of .6 (60%) in
a single phase system.
Some other losses get into the mix
as the currents and or HZ get larger.
If you are doing, a power loss (efficency), or
a temp related design, the "normal" design,
is to make wire losses equal mag losses.
The only way I can address this is to say
that it is in general correct for low fc designs.
As you can start to see, even a "simple"
60Hz or 400Hz transformer starts to get complex.
Once upon a time there were courses
you could take to understand "analog"
design. I have had sr. (20 yrs) design
know how, not understand the basic problems,
in a transformer, and power supply design.
As far as I know, less than ten schools teach ps,
and say only 4, have a complete spectrum.
Example (newbi from school) no concept of
how to cal a caps energy in a given system.
No general concept of temp rise in a component.
I used him to do "mean time between failure"
reports which forced him to learn the above
concepts and much more. That assignment
coupled with my guidence and other eng help
enabled hin to understand the real math side.
When he was ready we next started him doing
pc card layouts and making it work.
(You canot do 20amps with a 6 inch loop at 10ns.)
The only real problem he has remaining is magnetics.
I have been trying to teach him, it will happen soon.
He understands the magnetics it just will not fall
into place (click). Once this happens he needs
to unify all of this into a overall understanding.
Once he gets there he will be able to do any design.
Note: he should surpass me and all the designers
that I know, some are considered the best
in the U. S. Than he needs to update his technology.
We are about ten yrs behine.
When he gets there I hope he stays with us.
This has been a 4 yr teaching.
As eng mag I can only do job assignments and
show how they should be done.
As a person I try to connect with him and understand his problems, and also show him all of the aspects of
coping with companies such as BO and RA.
At this stage of my life "62" sharing knowledge
(while it has always been a large part of it)
has taken on a different aspect, (this is my third
teaching, and must be my last).
I debated to myself about putting this last
segment in, but after readin the context
I decited that I should. It is a flow that just
came out, and I think gives the reason we are
all here. We all want to share "know how".
Personal note, I do thank you for your
understanding, and again sorry for the
stupid reaction.
