Re: Physics of magnetic flux concentrators
Posted by Bert Hickman on 4/8/2009, 9:50 am, in reply to "Physics of magnetic flux concentrators"
--Previous Message-- |
: I need to build a current sensor and am using a coil wrapped ferrite ring.
: The question I have is how to focus the flux concentration best on the
: sensor. I have several questions I need more knowledge on.
: 1) If I have a larger ferrite band, that means i will have a larger
: concentration of flux within the ferrite band?
: In order to place the sensor within the field, I need to embed it in the
: ferrite. so given (A) a ferrite band of 2in diameter with a thickness of
: 1/4in and a height of 1/2in and (B) a small transistor sized hall effect
: sensor, which of the following will yield the highest field density
: crossing the sensor?
: a) drilling a tiny hole in the middle of the ferrite band
: b) cutting a slit cleanly through the ferrite band
: c) cutting a slit as in (b) but shaping the ends in an angular fashion to
: focus the ferrite ends at the sensor
: Thank you in advance!
The larger the cross sectional area of the ferrite, the greater the amount of flux that can be handled within the core without saturation and (potentially) the greater the degree of flux concentration. The cross sectional area of the core should be at least as large as the sensor and up to about 2X the area of the sensor if tapering the core. Example A (above) should not be used since most of the flux will bypass the sensor and you'll also get highly nonlinear performance.
Because ferrite is quite brittle and hard to machine, most ferrite-based hall effect sensors simply use a slotted ferrite core (your example B above). If you really need to further concentrate flux, then the ends of the ferrite should be tapered to channel the flux down to the area occupied by the sensor - your example C above. You can use a simple conical taper or a more complex taper. An example of a more complex geometry can be seen here (using a steel alloy, not a ferrite):
The degree of further flux concentration by shaping the slot faces is limited by increased flux leakage in the tapered region. For linear characteristics, the reduced core area (adjacent to the sensor) should be no less than half that of the main core area. Following is some additional information that may help you with your design:
Allegro also has a number of other publications dealing with Hall Effect applications that may be useful for you: