Based on your testimonial I tried the soldering iron method myself. I cracked the first ferrite by heating it too long (5 minutes a side). But managed to successfully split the second one by heating it a shorter duration (2 minutes - 2 flips). I used a 100 Watt iron. I Krazy Glued the first one back together. I think it will be OK?

@@1pcfred that would be just fine

Ferrite core can break apart by soldering iron. There is a device called heat gun which can be used as well.

@@curiosity551 low power soldering iron works the best. not everyone has a heatgun and it wastes a lot of power

Well, can confirm hot water method worked 1st time! Now I can take over the world with my evil plans lol

The only time I've seen epoxy used in core assembly was in the case of a gapped core where epoxy was used as a mechanical bridge across magnetic gaps in the external legs, such that the assembly remains integral. Epoxy is both expensive and viscous and rarely used in this type of ferrite transformer manufacturing. I've never encountered it as a vacuum impregnation compound for consumer grade transformer manufacture.

100°C is far below Curie temp. and is below max working temp. of these regular ferrite materials.

So you recommend acetone over other methods shown? Acetone doesn't distort the plastic? Thx

I'm not recommending anything in this litigious world, just reporting my findings. Using acetone I soak a core for a few hours/overnight in a glass jar with a scrap of Al foil sealing between the jar and lid. The bobbin material is typically nylon and I've never had an issue of solvent attack/distortion. I suspect acetone or closely related volatile solvent is likely the vehicle originally in the varnish. The greatest problem I've had was preventing the acetone from being lost to evaporation due to infrequent use. But I'd expect one could otherwise recover quite a few cores with one charge of acetone before it becomes too saturated with varnish to be effective. Sometimes when recovering a core as such I find it has cracked ostensibly due to improper manufacturing or excessive heat stress during use. In this case the core can be repaired with a cyanoacrylate adhesive (aka: crazy glue) and if you don't get the pieces to mate precisely, an acetone dunk will break the bond allowing another try. Epoxy gives you more working time but also needs the better part of a day to fully cure and a bad repair can't be easily undone. Epoxy is also more viscous and may complicate realignment. But both options produce near perfect, magnetically inconsequential repairs with a little care. FWIW acetone can be found in small quantities as fingernail polish remover. Most appears to be dilute acetone but I've also seen 100% acetone at the local dollar store. Even if you boil the bejesus out of a core you're probably gong to need a solvent for clean up of the core's mating surfaces to minimize debris and also to disassemble the bobbin for winding count study and/or potential reuse. Bobbins are essentially a manufacturing convenience which cost quite a bit of window area. So if the application looks to be cramped I'll toss the bobbin and just fabricate minimal overhead bobbin from grocery bag kraft paper and glue using a drill bit/rod as a bobbin form. Plastic wrap on the form beforehand helps release the glued/wound bobbin.

IME it really isn't needed. Acetone has virtually no surface tension and permeates the varnish impregnated transformer readily doing the work for you. It also doesn't mechanically / thermally stress the core, minimizing the probability of damage -- the core just falls apart. Until the acetone flashes off the remaining varnish remains rubbery and can be quickly scrubbed off with an old toothbrush. I expect lacquer thinner may work equally well but haven't verified so.

Yes, acetone is probably the best solvent which do this job really well. I have done over 100 ferrite core disassemble and acetone is best solution. I think that maybe there is one good solution more - acetone and paint remover and/or spirit mixture.

@@uhmgawa6533 Awesome man you really helped me out

Use super glue, and everything will be nice

mine broke too and i used super glue to glue the pieces together again. some say that it works the same

@@f.a3202 super gluee, yep im gonna glue it..

What are you referring to as "this"? The boiling water approach should have the best chance of evenly heating the core, particularly the center leg buried in the bobbin. Even then I'd ramp the water temperature gradually to avoid core thermal gradients causing stress. If you do have a simple break, it most often can be repaired to original magnetic condition using cyanoacrylate or epoxy, with care to get accurate realignment of the pieces. Use of the mating core half/piece helps to achieve this, but may not be necessary if you're going to gap the core in your application. Solvent disassembly avoids the temperature induced stress, although you may infrequently encounter a core which has cracked due to manufacturing or design oversights. But that's quite rare in my experience.

That's alot of work.epcos I.e tdk cores cost $3. I think I was to aggressive with my hot air gun at 450C. One tip I can offer is the optimal switching frequency can be obtained from the rc oscillator on the chip.

I find reclaiming cores far less work and far more immediate than ordering 1~2 cores needed for prototyping. Moreover a cast off PC switcher is a veritable component goldmine for power conversion experiments and can be had for gratis. Granted it is a pot luck component source, but a well stocked scrap bin has been my immediate "go to" source for decades.