Saturated ferrite

Yeah, the common yellow-white and green-blue ones are slightly inductive resistors. That's a little unfair, but really, there are a lot of applications where a Q of 10 or 15 is simply ludicrous.

This is a hunk of #26, at 1MHz, ungapped even.

Heats up quite a bit slower, especially initially, but once it's there, it's especially lossy, little more than slightly porous steel.

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
Website: http://seventransistorlabs.com

It's interesting because most materials, the saturation is still gradual; ferrite might start to wheeze at 300mT, but after putting a few thousand A/m on it, it still keeps on going to the 400 or 420mT the datasheet says. Not that you'd normally ever be going near that level of field strength; copper's not conductive enough.

Alnico is one of those materials. Which is partly why it's a somewhat cantankerous material. You need to "keep" it, prevent bumping (mechanical shock; not sure what the physical reason is, perhaps microcracks or strained/deformed crystallites?), prevent demagnetization (excessive current pulses on the motor/speaker/etc.), and all that. Considering how old it is, it's a pretty good alloy -- Bsat is up there with NdFeB (1.2T I think??), it's just not as "hard" (less coercive force --> less B*H product).

NdFeB is one of those "square" materials. It saturates all suddenly at once, and hard; little permeability is left. The only difference between it and, say, supermalloy or orthonol, is instead of teensy coercive force (< 1 A/m), it's MASSIVE, so it's very hard to demagnetize (and also gets rather hot, from eddy currents and sheer magnetic hysteresis loss, in the process, or during magnetization just as well). (Okay, and Bmax, because those examples are in the 0.6-1.2T range, not 1.5ish. But they're some of the lowest B*H product, square loop, materials. Since then, of course, nanocrystalline has taken over, with higher Bsat and lower losses.)

Playing with mag amp cores is interesting. You can connect a square wave from a function generator, to a core with oh, a hundred turns or so, to a load resistor, all in series. Measure the voltage on the resistor. Rather than a properly "square" wave, it will have dead time during which no current flows. Then all of a sudden, BAP, the core saturates, and the source and load are connected, and you get an up-pulse. Then it discharges flux, then you get a down-pulse, and so on. The switching often occurs fast enough to excite parasitics, i.e., you see a little overshoot and ringing.

On a rather higher power level, these things are used for "magnetic pulse compressors" for particle accelerators (driving charge plates, magnetic or electric deflectors, etc.). Usually goes something like, whack a huge IGBT (t_r ~ 500ns) into a stack of VITROPERM toroids, wait for it to saturate, then wham, out goes

Well, that doesn't quite work how I expected. If I set up the compass near the choke and fiddled with the angles so that you might expect the the coil's field to swing the north-pointing needle through 135 degrees, what happens is the needle moves as you increase the current, then stays more or less stationary above some current.

As it happens, this is roughly the calculated saturation current.

So I guessing that the needle is responding to the external field which reaches some limit at saturation. I was expecting the field to 'burst through' but it didn't.

A tip: if you use a 'Dremel' to grind away the centre post of a ferrite core, a close by rare-earth magnet will suck up the fine dust this produces.

Cheers

--
Syd

The magnetic field doesn't spill all over the place when the core is saturated (on the flat part of the BH curve). Increasing the Ampere- turns (H) won't change the flux density (B) very much.

** Another tip: if you gap a ferrite pot core with insulation material, make sure to put some on the centre post as well as making a ring for the outside rim.

Otherwise, when you tighten the centre bolt...

.... Phil