Re: Low loss core material

rote:

e:

te:

efficiency

ncreases

ncy increases

max =3D VT/

e max B field

result is a

owerloss, if

/pc45_pc46.pdf

ss at 100kHz,

ttp://www.a...

at low B?

s?- Skjul tekst i anf=F8rselstegn -

oid. We

actor to

ects the

vity

info on

er

A/m no

ubtracts

a roll

/10 any

pretty

ability

ve eddy

d,

ishi,

ame

th

n is

an

all

um

ike

as

ut

it

=F8rselstegn -

Oops, memory is shot: that cost was for the torroid core, unwound. TOTAL cost per interface in 1000 quantities, including PCB area was around $0.87, and under the budgeted $1 [and promised to my client's customer] The design I replaced was created by the client, cost around $3 ea plus required installation labor! to match electrical characteristics to what was being measured - whether high voltage, or low voltage.

regarding how to center

align itself)..." that is first pass solution, NOT the best way to check. Key word there is 'assumption'

Was that design a BAD idea carried out WELLl? Or, a GOOD idea carried out BADLY?

From reading your posts, I know I'm 'preaching to the choir' here, but as intriguing as it is to pursue the BEST approach, keep in mind your performance SPECS. When you meet those specs YOU'RE DONE! Else we'll be here all year. :)

Regards, Robert

Surprisingly, I've observed materials that retain DC bias. I've played around with orthonol before (essentially fancy-pants Si steel, maybe a bit of nickel, plus whatever makes it really high permeability and square loop). Circuit: +/-15V square wave generator at variable frequency, low output resistance (a few ohms), series circuit of 15 ohm current limiting resistor, coupling capacitor (4.7uF electrolytic) and current transformer. Bmax roughly 0.05 to 0.4T in the "drifty" range, on up to saturation obviously (over different test windings and frequencies under test).

Results: by dropping the frequency (to a few kHz for most size windings used), the core saturates symmetrically, resetting flux consistently. When frequency is increased, saturation ceases and flux remains momentarily balanced. Then it visibly drifts back to one side of the B-H curve with a time constant of about 0.5 second. The side it drifts to is fairly random. That it drifts at all suggests a positive feedback mechanism, a "waist" shaped B-H curve even.

Saturation can be induced with a small bias current (such as by loading the cap-coupled end with a 100 ohm resistor to either supply), constituting an amp-turn or so of bias. Around 7mA (less than 0.1 At) was the smallest "nudge" that would move it out of bumping-its-head-against-saturation.

Tim