Creating "custom ferrites" by modifying off-the-shelf ferrite cores possible?

[RandomUser]

Hey folks,

this is my first post on the EEVblog Forum, so bear with me.

Introducing myself:
Maybe a little introduction first. I'm a "freshly" graduated EE (roughly one year ago) from Germany. In my bachelor's, I studied EE with focus on communication electronics. I switched to power electronics in my master with a specific interest in 48V/12V automotive dc-dc converters. I'm working in the industry since May 2019.

Problem/Idea to solve:
For my private investigations of 48V/12V multi-phase dc-dc converters, I use coupled inductors as output filter inductance. Unfortunately, those coupled inductors often use custom ferrite cores and hence, are very expensive to produce if the required quantity is low.
My idea is to use off-the-shelf ferrite cores, e.g. ELP/EILP, to build some coupled inductors. But for my purposes I would have to modify those cores, e.g. grind off some µm off the core or even "cut" the core in half to obtain specific properties for my coupled inductor.
The problem is, that ferrite is some kind of ceramic-like material that can break very quickly and is very difficult to work on. I know that manufacturers of ferrite cores like TDK, Sumida, etc. use CNC machines for prototyping. So maybe some similar but less expensive tool could use for that as well. My hope is that we have some experts here in the EEVblog Forum who know how I could modify the cores to build my own custom magnetics from off-the-shelf components.
What kind of tool would be required to grind/cut ferrite cores?


Thanks in advance, greetings from Germany and stay healthy

[TimNJ]

Hi. For non-toroidal ferrite cores, gap modification is quite easy to do on your own. Recommend getting an approximately 10mm flat diamond file. Set the file down on a flat work surface. Then, firmly (but gently) pass the ferrite over the file while applying a little pressure until enough core has ben ground off to achieve the correct gap length.

Design and wind the coil on a bobbin/former first. Then, as you grind down the core, every so often put the coil in the core and measure the inductance until it’s right. Professional inductor manufactures usually use a “standard test coil” which is usually 100 turns, I believe. Instead of tuning inductance value, they tune AL value.

I know you’re in Germany but here’s the variety I use.

https://www.mcmaster.com/Diamond-Files/rectangular-diamond-files

[wizard69]

If this is for non production experimenting it might be better to consider off the shelf components.   If you needed a transistor would you fabricate the silicon yourself.   I realize that inductors are a different story but cores really are the same thing as transistors in this context.

Given that this isn't entirely impossible though I do have some suggestions.   If you have a lathe or mill it might be possible to machine some core materials there.   The trick is work holding and frankly this applies to manual approaches also.

There are a number of ways to approach workholding would be to mount the material, likely a ceramic, is to take an example from the optics industry and use a mounting wax.   One example of such material comes from: https://strongholdwax.com/.   Depending upon what you are trying to accomplish these waxes can be very effective.

A second approach is to use super glue to mount the form to a backer.   Like waxes this can be effective and possibly a more aggressive way to mount your form.   In both cases you can release the part with heat or solvents.

In any event these sorts of mounting approaches can help support the more fragile forms.   then you have to consider how to actually machine the materials.   Here you may need diamond tooling for that tooling to last any length of time.   However some materials may machine OK with carbide mill cutters or burrs.   Note that much of this info comes from the machining of odd stuff at work.   The problem is ceramics can differ a lot from one type to another when it comes to machining.

[RandomUser]

Thanks for the quick replies!

@TimNJ:
I will take a look where to buy a not too expensive diamond file. Are there any suggestions on quality tools? Also, what type of surface finish is recommended for the file? Which file surface finish would be best? Coarse to get off more material?


Have you ever tried diomand cutting discs for a dremel to cut ferrites? To grind of material maybe one could use something like this?
https://www.dremeleurope.com/at/de/siliziumkarbid-schleifstein19-8mm-151-ocs-p/
How long would such a "bit" last? Has someone experience with this?

[TimNJ]

Thanks for the quick replies!

@TimNJ:
I will take a look where to buy a not too expensive diamond file. Are there any suggestions on quality tools? Also, what type of surface finish is recommended for the file? Which file surface finish would be best? Coarse to get off more material?


Have you ever tried diomand cutting discs for a dremel to cut ferrites? To grind of material maybe one could use something like this?
https://www.dremeleurope.com/at/de/siliziumkarbid-schleifstein19-8mm-151-ocs-p/
How long would such a "bit" last? Has someone experience with this?

Hmm. I'm not sure what grit our file is. I'd recommend the coarsest though. In general, the surface finish (smoothness) does not matter. The only downside of a coarser file is maybe a higher likelihood of catching the edge of the ferrite and ripping off a chunk by accident. If you're careful, not really a problem.

You read my mind about the diamond drum/cutting disc. I was about to leave a second reply/suggestion last night, but never got around to it. For big gap dimensions, let's say greater than 1mm, we tend to start with a Dremel + diamond drum, and then finish off with the hand file. In my experience, it's easier to control with the rotating axis 90-degreesto the surface. For the wheel you linked to, you have to come straight down on the ferrite, which has been tricky in my experience. Something like is probably better: https://www.dremel.com/en_US/products/-/show-product/accessories/83702-silicon-carbide-grinding-stone

Useful life may be limited with silicon carbide. We have a diamond rotary drum, but it's a questionable quality one from Amazon, and the diamond is basically gone now. We've  gapped maybe 20 cores with the rotary attachment, and 50 cores+ with the file. The file's still in perfect shape, and the drum's basically shot as I mentioned.