Ferrite bead saturation

They have it but don't like to give the information out. That reminds me, though, the local Murata rep said she'd get me datasheets with this information.

Saw a tip in the crapacitors thread from a couple of months back that Laird actually puts those data in their catalogue, and what do you know, they do.

Kudos and POs to Laird!

Cheers

Phil Hobbs

Virtually all of the ferrites ( at least those used for beads, transformers, inductors, etc, saturate at around 0.2 Tesla ( 2000 gauss ). That would be a good starting point.

IIRC, FairRite (and possibly others) lists the materials on their beads, and their properties, and gives enough info elsewhere to do the math yourself.

But having a number on a data sheet is _definitely_ better.

Don't they tell you which ferrite the beads are made of?

The ferrite material data sheets always seemed to include magnetisation curves.

Usually they do. The rated current is typically way less than what would cause heat issues. Operating above can result in saturation.

I'd go for a 0805 size here. Laird has graphs, most other's don't but if you get one with a 1A rating you should be ok.

I have done measurements but those were decades ago. It's almost like measuring the saltiness of salt.

Sure, I can make a guess by assuming that the bead is all ferrite, and uniformly magnetized, and that the internal coil has negligible volume, and that the fringing fields are zero, and that my math is right....

Or the morons can just come clean about it--it's not like they don't have the data.

Kudos again to Laird.

Cheers

Phil Hobbs

Sorry, I was engaged in a brain-fart. I was thinking "ferrite bead" in the old "little dark gray thingie that slips over a wire" sense, not the new "little surface mount inductor with a possibly misleading name" sense.

That's because they're saturable reactors sold under a different name!

They certainly aren't inductors. :^)

If you need inductors, shop for inductors!

As you noted already, Laird absolutely!

But offhand, I don't think you'll find anything that can deliver that much, even at 1812 size.

What do you need 1kohm for, in a power supply? That's equivalent to about

about

er

I can't always get what I want, but if I try sometimes I just might find I get what I need. ;)

That sounds a bit over the top. Below the Z peak, the R appears in series w ith the L. I'd believe 40%, but not 10x.

RF

There's one there already. The most recent iteration uses a 2SD2704K, which is an amazing part except for being slow. That's where the bead comes in. I can get beads cheaper than swoopy capacitors, and since they make an open circuit, there's less high frequency current to radiate than if I rely sol ely on bypass caps.

Mostly I want the truth about what the parts can actually do.

Cheers

Phil Hobbs

We stock some 6 amp rated 1206 beads, 33 ohm, Murata. I sure hope they don't saturate at 400 mA!

The other thing most bead people don't specify is low-frequency inductance. You have to figure it out from the impedance curves (if any.)

In parallel, that is.

Cheers

Phil Hobbs

I still use those ones on dead bug protos. It's a lot easier than putting 0603s in mid air.

Cheers

Phil Hobbs

If you refer to the Fair-Rite catalog (14ed), you'll findcharts of effective impedance vs frequency at various DC bias settings, for simple bead-on-lead products. These rarely achieve impedances above

In the same catalog, you'll find similar data for SMD and multi-aperture parts. Again, 1K at 100MHz is a little optomistic.

The good news is that DC bias up to 2A seldom reduces the lossy impedances by more than 2:1 in multi-aperture parts - so you know what to lay out for......

RL

You should also be able to see from comparison of mechanically similar parts, that lower-frequency, lower temperature, higher permeability materials are more sensitive to DC bias effects. Unfortunately, these are more commonly available.

Avoid anything claiming significant (design) impedance at 10MHz or lower.

RL

I broke a ferrite bead in half and it looks like this:

Isn't it just great? Murata doesn't report it, so how would you know? ;-)

IIRC, a comparable Laird part begins rolling off around there. "Saturation" being defined as some useful fraction, like 70% of zero-bias impedance.

And sure, it keeps going beyond there, but it drops off faster and faster; by rated current, it's literally no better than the length of wire within.

Yeah, but that's the fault of the material -- inductance isn't a useful metric for a device with such a low Q at all frequencies.

I should build a DC-DC converter, using only ferrite beads, some time. Just as a gimmick. Most of the heat will be dissipated in the ferrite beads, and that's given that I choose the lowest loss frequency!

I've also measured some Steward/Laird ferrite beads, #28 material I think, as being truly horrible (meaning: truly wonderful as a ferrite!). Motivation was balun / pulse transformer duty. You can't do much better than having the shunt impedance a modest multiple of the RF port impedance.

Tim