Maker Pro
Maker Pro

Super-tiny ferrite rods anywhere?

J

Joerg

Jan 1, 1970
0
Paul said:
Yes, I got that wrong. Better idea - etch a conductive spiral on top of a
ferrite substrate?

Yup, I talked about that with another engineer at a guide dog meeting
yesterday. But the winding is not so much the challenge. We could get
52AWG wire, stuff one probably couldn't even see. The challenge is how
to arrive at a ferrite core of this size.
 
J

Joerg

Jan 1, 1970
0
I don't think anyone's suggested this idea yet, so how about rolling
out the ferrite cores between two flat plates - glass sheets perhaps?
You'd have to absorb or mix ferrite dust into some sort of gummy
medium that would be readily roll-able, but if that could be achieved
then a pretty much uniform 5 thou diameter rod should be easily
achievable, I'd have thought. Just keep rolling out and re-rolling
until the desired diameter is obtained.


That's almost how centerless grinding works. Except now it would be
centerless rolling :)
 
S

Spehro Pefhany

Jan 1, 1970
0
Sorry typo. "Down to" was meant. Now if we can only get 0.005" that
might be ok. 0.006" maybe not so.

You might talk to shops- we have had some custom cores made (very
expensive in small quantities), but nothing really small- this was for
an application that was very, very special.

A note on communication terminology with shops- you'll most likely be
talking about "centerless grinding" rather than "machining" per se.
Maybe try All Diameter Grinding in Orange, CA 714-744-1200 .. if they
can't do it, they ought be able to point you to someone who can.
Grinding diamters down to a couple of thou (half of what you're asking
for) is not unusual. Maybe start with the minimum Fair-Rite can do-
0.5 or 0.75mm, IIRC).

BTW, Tim W., I now have a 0.394" dia x 0.394" diamond grinding bit in
hand (literally) and will hopfully try gapping an E-core over the
holidays. They recommend mist or flood. It would be nice to get close
to the 30K RPM limit, but I don't have a TP grinder <sad face>.
There's a 0.256" one that's rated at 60K RPM (4000+ SFM), 10-20x what
I can get with the 0.394 bit and the spindle of either of my mills. I
suppose I could try and clamp an air die grinder (noisy..).


Best regards,
Spehro Pefhany
 
S

Spehro Pefhany

Jan 1, 1970
0
That's almost how centerless grinding works. Except now it would be
centerless rolling :)

There are spin coated ferrites (binder plus nanoparticles).. if you
search for chips with integrated ferrite inductors you'll find some,
but they're talking pH or likely nH at best, and probably a science
project to boot.



Best regards,
Spehro Pefhany
 
S

Spehro Pefhany

Jan 1, 1970
0
Hey, google "metglas wire." The stuff apparently exists. It's an electrical
insulator, essentially a metallic glass, but has permeability up to 1e6, if you
treat it right.

Is it commercially available? I know some guys who have been following
metglas developments (afaik it's sort of a subset of nanomaterials),
but last I looked the research guys were mostly making their own in
the lab.


Best regards,
Spehro Pefhany
 
S

Spehro Pefhany

Jan 1, 1970
0
Flat stuff, tape, is available. I bought a spool of it on ebay. The tape is
wound on bobbins to make toroids. It's great for wideband CTs and DCCTs and
stuff.

So, useful at 10's or 100's of MHz?


Best regards,
Spehro Pefhany
 
S

Spehro Pefhany

Jan 1, 1970
0
Yup, I talked about that with another engineer at a guide dog meeting
yesterday. But the winding is not so much the challenge. We could get
52AWG wire, stuff one probably couldn't even see. The challenge is how
to arrive at a ferrite core of this size.

I've spoken with a coil winding company (in New England, IIRC) that
can do microscopic coils- they seem to target biomedical anyhow.


Best regards,
Spehro Pefhany
 
W

whit3rd

Jan 1, 1970
0
Jeff Liebermann wrote:
[about very-small-diameter ferrite rods]
That's what I wanted to avoid if possible. Dissolved ferrite never
reaches the performance of sintered ferrite.

Ferrite is an aggregate of carefully sized grains; the ferrite
manufactories DON'T generally sell the powder, because that would
give away their technology. And, grinding a built ferrite object
into powder changes those grain sizes, so there's no hope
of replicating the original material specifications.
 
J

Joerg

Jan 1, 1970
0
whit3rd said:
Jeff Liebermann wrote:
[about very-small-diameter ferrite rods]
That's what I wanted to avoid if possible. Dissolved ferrite never
reaches the performance of sintered ferrite.

Ferrite is an aggregate of carefully sized grains; the ferrite
manufactories DON'T generally sell the powder, because that would
give away their technology. And, grinding a built ferrite object
into powder changes those grain sizes, so there's no hope
of replicating the original material specifications.


That's ok. Even if we only get a 1.5x in inductance increase it'll sure
be worth it. Also, sometimes you can line up deals with NDAs or by
lining up a custom deal. For example, where a ferrite manufacturer has
an in-house department or a trusted companion enterprise provide
epoxy-ferrite mixes to us.
 
J

Joerg

Jan 1, 1970
0
John said:
I sure wish you would/could tell us what you're actually trying to do.

Essentially it's measuring minute changes of a variable capacitor. It's
only 6pF and the variation is +/-0.5pF full scale. The capacitor is
going to be 0.008" width and height. This is wired up as a super-tiny
resonant circuit and, after 5ft of not very ideal cable we'll have to
detect the resonance and track it. IOW, the system has to extract the
FM-modulation and quiescent bias in this signal. In order for this job
not to become too simple or boring, this is all going to be immersed in
fairly conductive fluids.

Sounds all simple but when you have to do all this (less the
electronics) in a metal cylinder of under 0.015" O.D. it becomes
non-trivial. The challenge is that the coil can't be too close to metal
or fluids or the Q and thus the resonance would almost collapse.
 
J

Joerg

Jan 1, 1970
0
Spehro said:
You might talk to shops- we have had some custom cores made (very
expensive in small quantities), but nothing really small- this was for
an application that was very, very special.

A note on communication terminology with shops- you'll most likely be
talking about "centerless grinding" rather than "machining" per se.
Maybe try All Diameter Grinding in Orange, CA 714-744-1200 ..


Thanks! Duly noted and will call them. Theye even have a web site:

http://alldiametergrinding.com/

... if they
can't do it, they ought be able to point you to someone who can.
Grinding diamters down to a couple of thou (half of what you're asking
for) is not unusual. Maybe start with the minimum Fair-Rite can do-
0.5 or 0.75mm, IIRC).

That's one thing I was planning to do.

BTW, Tim W., I now have a 0.394" dia x 0.394" diamond grinding bit in
hand (literally) and will hopfully try gapping an E-core over the
holidays. They recommend mist or flood. It would be nice to get close
to the 30K RPM limit, but I don't have a TP grinder <sad face>.
There's a 0.256" one that's rated at 60K RPM (4000+ SFM), 10-20x what
I can get with the 0.394 bit and the spindle of either of my mills. I
suppose I could try and clamp an air die grinder (noisy..).

I hope your shop is heated :)

Yesterday we had a guide dog meeting in an unheated warehouse. It was a
great event but chilly.
 
S

Spehro Pefhany

Jan 1, 1970
0
Essentially it's measuring minute changes of a variable capacitor. It's
only 6pF and the variation is +/-0.5pF full scale. The capacitor is
going to be 0.008" width and height. This is wired up as a super-tiny
resonant circuit and, after 5ft of not very ideal cable we'll have to
detect the resonance and track it. IOW, the system has to extract the
FM-modulation and quiescent bias in this signal. In order for this job
not to become too simple or boring, this is all going to be immersed in
fairly conductive fluids.

Sounds all simple but when you have to do all this (less the
electronics) in a metal cylinder of under 0.015" O.D. it becomes
non-trivial. The challenge is that the coil can't be too close to metal
or fluids or the Q and thus the resonance would almost collapse.

I really enjoy meeting that kind of "almost-impossible" set of
requirements.. good luck with it.


Best regards,
Spehro Pefhany
 
J

Joerg

Jan 1, 1970
0
John said:
I got involved in tomographic atom probes, interesting and profitless. In one
case, we needed tiny (as in micron diameter) vertical samples of silicon wafers
to stick in the machine to analyze. They got pretty good results by ion-beam
milling arrays of posts (remove everything that doesn't look like a post) and
then breaking them off. Ion milling would probably be too slow for production.

Whether this particular venture will be profitable isn't sure, of
course. That's what we engineers are for, to push the envelope and try
to make the impossible not just possible but turn it into a product.

Hey, google "metglas wire." The stuff apparently exists. It's an electrical
insulator, essentially a metallic glass, but has permeability up to 1e6, if you
treat it right.

I only found links like this:

http://scholar.lib.vt.edu/theses/public/etd-11198-54532/materials/ch3.pdf

125um sounds just right, about 0.005" would be ok. Got to call Metglas
after New Year.
 
J

Joerg

Jan 1, 1970
0
Spehro said:
I've spoken with a coil winding company (in New England, IIRC) that
can do microscopic coils- they seem to target biomedical anyhow.

If the name of it jumps at you please let me know. I only know New
England Wire but they don't do inductors and are in New Hampshire.
 
J

Joerg

Jan 1, 1970
0
John said:
Why resonate? Why not measure the 3T capacitance directly? fF resolution isn't
hard, aF has been done with some accuracy.

That's because there's roughly 5ft of cable between the place of
measurement and the system. It moves around a lot. The unintended
variations in the capacitance would suffocate any measured information
unless we make sure there is a distinct local resonance that can be seen
from afar.

My old Boonton 72 would measure 6+-0.5 pF easily at the end of some long cables.

In a not too noisy environment. We have to put this into a fairly quiet
frequency band and remain very selective in the system filters.

I'd be concerned about variation in cable capacitance changing the resonant
frequency. A 3T cap measurement doesn't much care about cable capacitance.

It works, I've tried it on the bench here. But with slightly larger
inductors of about 0.060" O.D., that's the smallest I have right now.

Or use one of Phil's fiberoptic interferance things. Presumably capacitance is
just a surrogate for displacement.

Can't use that for various reasons, one of them being IP.
 
S

Spehro Pefhany

Jan 1, 1970
0
I hope your shop is heated :)
Yes.

Yesterday we had a guide dog meeting in an unheated warehouse. It was a
great event but chilly.

Still well above freezing.. I suppose we'll get some cold days soon
and the rain will turn into something more inconvenient.

I've got a bunch of hard cider sitting in the garage- it's about at
the perfect drinking temperature. Saves room in the dual-zone wine
fridge for better stuff. ;-) The cold cellar is about right for reds.



Best regards,
Spehro Pefhany
 
J

Joerg

Jan 1, 1970
0
Spehro said:
Still well above freezing.. I suppose we'll get some cold days soon
and the rain will turn into something more inconvenient.

I've got a bunch of hard cider sitting in the garage- it's about at
the perfect drinking temperature. Saves room in the dual-zone wine
fridge for better stuff. ;-) The cold cellar is about right for reds.

We've got some "La Fin du Monde" in the fridge, from your neck of the
woods. Very good stuff. It seems it should be on sale now, considering
that the world did not end on Dec-21.

Oh, and we also have "Delirium Noel" from Belgium. Can't wait.
 
S

Spehro Pefhany

Jan 1, 1970
0
If the name of it jumps at you please let me know. I only know New
England Wire but they don't do inductors and are in New Hampshire.

It's not NE wire... we do deal with them regularly.

A quick look, can't find them (they showed up at a trade show and I
talked with them a couple of times, but we ended up designing our own
coil winding machine). 8-(


Best regards,
Spehro Pefhany
 
S

Spehro Pefhany

Jan 1, 1970
0
That's because there's roughly 5ft of cable between the place of
measurement and the system. It moves around a lot. The unintended
variations in the capacitance would suffocate any measured information
unless we make sure there is a distinct local resonance that can be seen
from afar.

You could probably do it if you could run a few coax or triax lines to
the place of measurement, but that sounds.. challenging.


Best regards,
Spehro Pefhany
 
Top