Clamping assemblies for large ferrite cores don't exist?

Hello,
We sent an LLC converter transformer manufacturing document to a transformer winding company for a Epcos TDK PM87/70 ferrite core (as on page 299 of ferrite databook below). We asked for the quote for them to make it.
They told us that they would have to charge us extra to make a custom assembly to hold the core halves together. –Since they said that Epcos do not make these for the PM87/70 core.

However, what is the assembly seen on page 288 of the Epcos Ferrite databook? This appears to be a ready-made assembly for the PM87/70 core?

Epcos Ferrite databook:
https://en.tdk.eu/blob/519704/download/2/ferrites-and-accessories-data-book-130501.pdf

That is the assembly they make for some of the PM cores, but if you look at the data for the PM 74/59 on pages 296-298 you will see the variant of this assembly listed for that core size (P298), while the following section on the 87/70 has only data for the core and the bobbin, with no cut sheet for the clamp assembly.

These big cores are often glued in production.

I would note that you must be careful with magnetics, often things exist in the manufacturers books that turn out to only exist as special order items (And how many thousand a month would sir like?), it is annoying.
The transformer winders have better visibility of this, so better to specify the electrical characteristics and then let the manufacturer tell you how best to do it.

Regards, Dan.

These big cores are often glued in production.

Click to expand...

thanks, it must be extremely strong glue for such a big core. It looks like the pm87/70 has a hole through the middle of the centre leg, possibly for a mounting post/
In any case, the core clamping parts on pg 288 are made of brass, and are virtually "in" the ferrite, so they will surely suffer much eddy current heating?

mechanical design is prone to fragility issues from shock and vibe.

treez said:

the core clamping parts on pg 288 are made of brass, and are virtually "in" the ferrite, so they will surely suffer much eddy current heating?

Click to expand...

Even if some heating occur there, should be due to the Foucault currents, which are strongly dependent on the width of any rod or screw placed inside the ferrite hole. Anyway, assuming a leakage magnetic field flowing there, should cross the air gap between both materials, which has quite smaller permeability, so in theory it is expected a negligible portion of the heating due to the clamping part.

What about a drill and dowel scheme with the dowel being
of similar ferrite, and all epoxied together?

The question that could arise with that decision is to keep some mechanical robustness in fastening the core to the board, as well the extra workload required to perform such assembly. I'm still in doubt if the heating at the clamping assemblies is really relevant.

How do you clamp large ferrite core halves together?

Hello,
Do you know how we can clamp together the Epcos TDK ferrite cores? They appear to big to glue together, especially since we have a centre leg gap.

Its on page 286-7 of the “Epcos databook 2013 Ferrites and accessories”…

https://en.tdk.eu/blob/519704/download/2/ferrites-and-accessories-data-book-130501.pdf

There is no clips or yokes offered with this core. Its a vehicle mounted power supply, ,so automotive vibration will be present.

Re: How do you clamp large ferrite core halves together?

treez said:

Do you know how we can clamp together the Epcos TDK ferrite cores?

Click to expand...

wind round by tape and than fix to coil former by compound

wind round by tape and than fix to coil former by compound

Click to expand...

Thanks, that's an idea I hadn't thought of, actually using the former as a base for the glue to hold the core halves together

its common to design your own clamps for larger cores, an Al or brass or copper bolt can be put thru the middle with no ill effect (actually better heat flow with a Cu one), the brass clamp on the outside of the cores suffers no heating unless it passes close to an external gap.