As well as micrometals, mentioned by Warpspeed, there is mag-inc.com, who do a good line in torroids...."magnetics incorporated"
- - - Updated - - -
Ferrite core inductors
If it’s a ferrite core, then the inductance is just L = AL * N^2 * 1E-9
Calculating the B value at current I is also just B = (I * L) / (N * A)
Where
AL = “AL” value quoted in datasheet, its related to the ‘allowance’, the reciprocal of reluctance
A = Area
B = Flux density
L = Inductance
N = number of turns.
Non linear inductor
However, if it’s a non linear powder core inductor, then you will need to use the kind of calculation as attached here.
- - - Updated - - -
The problem for non linear inductors is that at any value of H , you cannot calculate B by simply going H = B/(uo.ur)
This is because for non linear inductors, the B,H curve is not straight. So for non linear inductors, you just accept theat the value “uo.ur” merely gives the gradient of the B,H curve at a certain H value.
You have to make out the B,H curve yourself, and you do this by a kind of “numerical integration”, but really it just involves repeated application of the “y=mx+c” law.
So what you do is you approximate the B,H curve by assuming that it is made up of a series of connected straight lines. You know that your first line starts at the point B=0,H=0. You then make out a series of equally spaced H values, and calculate the value of the gradient “uo.ur” at each H value. So the gradient of the first line, which emerges up from the point B=0,H=0, you make this equal to the gradient at your first non zero H point…it is an approximation, but if you make the H values close enough, then it is sufficient.
It’s a highly time consuming method, but necessary
..................................
so anyway, if you cannot find a big enough ferrite core, then use say an amogreentech core, such as in the datasheet incided with the zip file....if each is not big enough, then dual two torroids together.