Big Ferrite HVTransformer Core? / High Voltage / Forums

Forums

4hv.org :: Forums :: High Voltage

« Previous topic | Next topic »

Big Ferrite HVTransformer Core?

Move Thread

LAN_403

LutzH

Tue Aug 18 2009, 05:54AM

Registered Member #1721 Joined: Sat Sept 27 2008, 08:44PM
Location:
Posts: 136

Hello Everyone:

I picked up this big ferrite core recently on ebay, unfortunately I do not know the exact type of Ferrite that it is made from? Its intended application is as a line filter for an extremely large industrial motor inverter. It is a Mitsubishi Line Filter, Type BKO-C2101H01.

The size, appearance, and the metal band mount of the core are an exact match for a TDK UU 79x 129x 31, part PE22 or PC40, which I found in the TDK Ferrite Core Catalog. Beyond the above information, I have no other information on this core.

Please see the attached picture of the unit, if these are suitable for HFHV work, then we have a potential new source for project cores : )

I am hoping that someone reading this post is knowledgeable in this area, and that this person can answer the question as to if this core is suitable for building a big HF, HV transformer?

Thank you so much for sharing your wisdom : )

Sulaiman

Tue Aug 18 2009, 08:57AM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

HV ... certainly
HF ... depends what you mean by HF, if you mean 3 to 30 MHz then no

Almost all ferrites are good to 100 kHz or more
I wouldn't try to make HV at 100 kHz or more, more like 20+ kHz
so that core is good.
At 10's of kHz the main difference between ferrites is relative permeability
emi cores for industrial equipment normally have high (5000 to 10000) relative permeability.

At 20+ kHz I'd be tempted to not use the steel band, find an insulating way of clamping

klugesmith

Tue Aug 18 2009, 06:29PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

That core is a magnificent specimen!

IMHO (not as an experienced designer of HF transformers) the steel clamping band probably isn't a problem.
It does not form a shorted turn. Eddy currents in its cross-section depend on magnetic flux inside the band, which is small because of small area (relative to core), and would be quite negligible if the band is made of nonmagnetic stainless steel.

As a line filter (common mode choke?) the original application needs to be effective at PWM-switching frequencies,
to prevent them from being conducted into the power line.

I wonder if the ferrite formulation might be deliberately lossy at HF?
You can measure the actual core loss using a throw-away primary winding of a few turns, driven at 20 kHz with just enough voltage to explore the approach to saturation.

Rich
[edit] First you can determine the ampere-turns for saturation, using mains frequency at even lower voltages, maybe using the original wound bobbins -- simply chart current vs. voltage.

LutzH

Tue Aug 18 2009, 08:28PM

Registered Member #1721 Joined: Sat Sept 27 2008, 08:44PM
Location:
Posts: 136

Thank you for the replies, yes you are 100% correct I should define HF, ooops. My intended frequency experimentation range is 15 - 100KHz. Mostly at 20-40KHz. Last night I did some more research and it seems that the ferrites used in line filtering / suppression, are almost the same as the ferrites intended for transformers, the filter ones in general seem to have a slightly higher permiability.

I just love the look of this core, my big concern now is if the metal band represents a shorted turn? Since it is electricaly connected it sure seems like it would act as a shorted turn! Maybe I can modify it with some nylon washers, or bushings, to solve this problem. Then it occured to me maybe induction heating could becaome a problem with the band also?

If anyone is interested in these cores they are used as line filters for heavy duty motor control inverters, and you can find them on ebay listed like this. A nice HV transformer built around this core should then perform just fine if I am reading the responses right?

Man, it would shure look sharp to, some serious eye candy in a project. On a related note, I always have to admire the aesthetics of some of the Tesla coils built by members, they look so nice after many coats of varnish etc. Thanks Everyone :)

Sulaiman

Tue Aug 18 2009, 08:53PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3141

For large cores at 20+ kHz saturation is rarely a consideration, core loss is.
For this core size I'd expect around 100 mT (1000 Gauss) at 25 kHz continuous,
more if intermittent or cooled by forced-air or oil etc.
In practice I'd expect more heat from the windings than the core.
Some core heating to around 70 to 80 C reduces core losses,
Just a couple of points to help you along from my 'playing' around.
I've never done cores professionally.

klugesmith

Tue Aug 18 2009, 09:01PM

Registered Member #2099 Joined: Wed Apr 29 2009, 12:22AM
Location: Los Altos, California
Posts: 1716

[edit] the retaining band appears on the next page of the TDK core catalog, followed by instructions that include, literally: â€¢ Do not use the bands for purposes other than to secure ferrite cores. The bands are not edible. Keep them out of the reach of children.

Anyway... if it were a shorted turn, it would be a problem in ANY electromagnetic application of this core. And UU core sets wound from steel tape would not be as popular as they are. The band would become a shorted turn IF installed in a way that passed through the core window, thus linking the magnetic flux path.

It's all about the geometry of linked loops. Current in a conductor (such as a winding on this core) creates a magnetic field which encircles the current (runs along the axis of coil, and returns on the outside). That causes a large magnetic flux within the body of a ferromagnetic core (next to which the flux outside the core is often negligible). Changes in magnetic flux induce a voltage which encircles the flux -- in other words the induced voltage is generally parallel to the current that makes the magnetic field.

If the core were wrapped with conductive tape, inside or outside the main windings, THAT would be a shorted turn. Within the body of the core, the induced voltage tends to circulate within the plane of a cross-section, and that is the way eddy currents flow due to electrical conductivity of the core material. But a conductor which runs parallel to the magnetic flux, even when its ends are connected, experiences no induced voltage so no eddy current.

[edit] An unreviewed calculation from Sulaiman's practical numbers (100 mT, 25 kHz) says you can run at 7 volts/turn (rail-to-rail, bridge drive with square wave) for core loss test.

jpsmith123

Wed Aug 19 2009, 01:57AM

Registered Member #1321 Joined: Sat Feb 16 2008, 03:22AM
Location:
Posts: 843

Hello Lutz,

The one disadvantage I see (as far as HV use is concerned), is the relatively small distance of only 34 mm between the sides of the U.

It seems to me that your other core would be better for HV for that reason.

Regards,
jpsmith123

LutzH

Wed Aug 19 2009, 02:09AM

Registered Member #1721 Joined: Sat Sept 27 2008, 08:44PM
Location:
Posts: 136

Thank You All :)

You have explained it well, if I do use it for very high voltage, I will wind the secondary the long way, and put the primary on the 35mm end, and then plop it in oil. Thanks again.....

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob