DIY Powdered Core Progress Report

[valveitude]

Well, I finished my Ball Mill to pulverize my break chips, and its..its..*sniff*..beautiful man *sniff*....



...OK, it won't win any prizes in the looks dept.,but it works exceedingly well. In the picture it’s humming along at around 60 rpm beating the chips into submission. I used 8 cast iron “T” fittings at about 2 lbs. apiece for the “balls”.

This is a picture of the break chips, clinging to a small pocket magnet used to retrieve screws, before processing ..



The first thing I did was to heat the chips to around 1400 deg F (red heat) burning off any oils/breakfluid contaminants, and then cooling them rapidly by pouring through the air into a cast iron pan. The rapid cooling causes the chips to become brittle from the stress. I then ran them in the mill for around 6 hours, this was the result…



Two thing jump out, how densely packed the powder is around the flux lines, and how black it is. Grey iron is peppered with little pockets of graphite, this is the form the carbon takes in the metal. It is along these pockets that the metal breaks when struck. As the iron is reduced by the mill, the graphite is freed from its metal prison.

The next step is to separate the graphite from the iron powder. This is done by float separating. I put about a cup or so of iron in a glass jar then poured lacquer thinner to about the 2 cup level. I then stirred it hard enough to suspend the iron. After it settles, a layer of graphite forms on top of the iron. I then stir it just hard enough to suspend the graphite. While the graphite is still swirling, I pour it off into a filtered funnel, cleaning the thinner for the next batch. The iron is then spread out in a pan OUTSIDE to dry. After doing 2-3 cups of powder, (it was still black, telling me I need to repeat this a few times) I got the graphite you see here…



It is graphite, rubbing between your finger and thumb, it feels like oil. First unexpected benny, a supply of graphite lubricant . The mass of the powder before and after cleaning went up 7% . That’s more than I expected. I am re-running the powder through the mill for 30-40 hours to reduce/clean it as much as I can. Right now its loose-filled mass is around 67% of solid iron.

After I clean the powder that’s in the mill now, I am going to ballpark its relative permeability by burying a coil in the powder.

Stay tuned…

Casey Brown

[valveitude]

Pics seemed ok when I posted, but when I checked again they didn't show up.

Is anybody having a problem seeing them ?

[Stocker]

Pics look good. Can't you get your hands on any of the right type of iron without having to break down the reinvented wheel yourself?

[valveitude]

Pics look good. Can't you get your hands on any of the right type of iron without having to break down the reinvented wheel yourself?

Nope...see this thread for details...

DIY Powdered Iron Tranny Core

Casey

[smoking-amp]

Hey, impressive start! I think a bunch of us are still waiting sceptically to see the Mu results though. I'll bet some 6AL5s on the Mu comes out under 20. Any other bets?

I tried some crushed ferrite core powder once. Original Mu was 5000. The powder, even packed, came out around 15. But the iron "stuff" has the advantage that its malleable, so maybe can do better.

Don

[valveitude]

Hey, impressive start!

Thank you

I think a bunch of us are still waiting sceptically to see the Mu results though.

Sceptisism is a good thing.

I'll bet some 6AL5s on the Mu comes out under 20. Any other bets?

Do you want my snail mail address now or later ? I don't have any 6AL5s, but I got a pair of 26s I'll put up.

I tried some crushed ferrite core powder once. Original Mu was 5000. The powder, even packed, came out around 15. But the iron "stuff" has the advantage that its malleable, so maybe can do better.

Ah..but at what frequency was the Mu speced at? I have seen the specs for dozens of ferrite cores, the lowest cut off I've seen on any of them was 500 hz, and if memory serves, I haven't seen that high of Mu on any core with a cut off lower than 100khz.
Powdered iron ,without silicon, has a Mu of 5000 down to dc. The fact that its malleable is also an advantage, I think the contact point grain to grain will be larger, resulting in a more linear B-H curve.

My minimum target is a Mu of 200

Casey

[smoking-amp]

my latest trip to the scrap metal yard, $1.00 a pound

The toroids are 1KW M6 , got a half dozen, and the extra long E -I cores must be a couple of KW, M6 with MgO2 coating, bought a big box of them. I figure, the bigger the core the fewer the turns to wind for an OT and the better the bass, also the toroids are big enough that I can get the wire spool thru the hole.

[valveitude]

Don,

The toroids are 1KW M6 , got a half dozen, and the extra long E -I cores must be a couple of KW, M6 with MgO2 coating, bought a big box of them.

Holy magnetic induction Batman!!

A PP 833 in your future perhaps ?

I want YOUR salvage yard..mine is good for parts off of apple sorters ect.

Nice score.

Casey

[smoking-amp]

Those were the ones that were small enough to carry out. Should seen the other ones that got away!

Try around Portland, with Tektronix there, must be something. I've heard the garbage they through out around Santa Clara, Ca is gold plated. Of course, there's always Ebay, but the shipping on iron will kill ya.

[smoking-amp]

Here's a nice long E transformer, part of a Kepco ATE --- , 500 Watt linear power supply I got on Ebay. Cost me $9.99 and $50+ for shipping. Was supposed to be a working power supply, but arrived with the heatsink/power transistors missing. The usual Ebay -------! It came from Washington state too! The core is really thick, so can probably get two cores out of it. But it happens to have a nice HV winding on it already, so may keep as a power xfmr. or I may just fix the supply. Its also an enameled core, so a bit of a headache to get apart. Haven't checked to see if its M6 yet. But I would guess that the other Kepco ATE --- 500 Watt models have the same xfmr size in them. Can find them cheap occasionally. Low voltage models are usually cheap.

Its not easy to find long E laminations, even new. So this is a nice find. The long E lams have low leakage inductance compared to the usual scrapless ones. Was the secret behind the best AcroSound output xfmrs.