Actually, these cores are not a ferrite but rather powdered iron with many little gaps spread throughout the material which makes the overall part, a toroid in this case, very difficult to saturate. I doubt if this material has any grain orientation but not sure on this.
What I always found interesting about this material in the various mixes is the change in initial permeability vs flux density. I attempted in the past to utilize this feature for a gain but was unsuccessful.
I've attached the Micrometals data sheet below.
Pm
Hi Partzman,
I was intrigued by the idea of using these positive changes in mu with increasing flux density. This seems to be a sort of 'self-organization' on the part of the domains, which has OU potential. I immediately thought of in-line flux switchers like the old Subieta-Garron patent attached, but considered it would be better to isolate the effect, not have it show up as an increase in current of normal transformer output, as Subieta-Garron claims--or for that matter, increase in flux density in air, like various flux switched motors.
After several days and many designs I came up with the following, attached.
You see an ordinary powdered iron toroidal transformer with primary L1 and secondary L2, with AC source and load R1.
Three neo magnets are attached to the toroid, two outside, and one inside, with (for instance) N poles facing in on the outside of the toroid, and S pole facing in on the inside. They of course can be arranged on the top and bottom of the toroid as well.
Two counter wound coils L3 and L4 intercept the flux from the two 'wings' of PM flux, which are naturally in opposite directions. These are connected in series to a second load R2.
Normal transformer operation creates a changing net flux, and thus causes a variation in mu of the core. We are using only the change in mu in this device.
If the transformer flux is rising, it causes an increase in mu, and thus in flux that L3 and L4 see from their respective permanent magnet pairs. Each coil generates current to oppose the rise in flux, and this current adds in series to go to load R2. The same is of course true if transformer flux/mu is dropping.
Because L3 and L4 are counterwound, transformer induction cancels and doesn't go to load R2. Because the fluxes that L3 and L4 generate are always in opposite directions, their flux doesn't induce on the transformer. There's complete decoupling between the two sets of coils.
As an added advantage, the opposition between L3 and L4 fluxes reduces the reactance of the output circuit.
I've shown only three magnets but of course they can be spaced around the entire core with outputs adding in series. The net flux from this coil set remains zero at all times.
L1/L2 are shown here as spaced apart but to reduce flux leakage they can be bifilar wound around the whole toroid-- it changes nothing about the operating principle.
More PMs can be added to the stack, as long as the flux in any 'wing' of the three magnet array remains around the steepest part of the flux vs. mu curve-- around 2000 Gauss in the chart you sent. At this bias, even a small primary A/turns in the transformer will cause a wide variation in mu, leading to an increased output to R2.
Since the transformer part can be made pretty efficient all on its own, the additional output dissipated across R2 could well take the whole device into OU territory.
Patent/project references: Subieta-Garron, Tupper, Jensen, Cobb, Magnetic amplifiers, etc.
Fred
Author
Topic: Parametric/inductive transformer concept (Read 4466 times)
