Ferrite core conection

Posted by: Viktor Horbunov on 09/30/2024

Hi Colleagues.

As seen in the DER-717, PI Expert and other sources, the ferrite core can be used as shielding, by connecting the wire to high voltage ground.
Some cores have this option (like RM cores with), some didnt.

And now the question. How to connect the ferrite core with wire, to obtain long-time stable connection? In mass-production?
-simply press the wire by the glue tape, like in DER-717/screenshot attached?
-use copper foil with conductive glue (my samples of such tape have non-conductive glue by the fact)
-other options?

Files

Screenshot 2024-09-30 164247.png (225.18 KB)

Comments

Hi Viktor,

Good day. Thank you for using Power Integrations solutions.

Thank you for reviewing our reference design. As you've observed in our published reference designs, we usually use separation tape in the windings to secure the bus wire around the transformer core. This method is the most practical for us, as we construct the transformer manually and prefer to use minimal materials; the tape we use also helps fix the cores to the bobbin. Regarding other methods for manufacturing in mass production, we could also consider soldering the ends of a copper sheet or foil around the core for a snug fit instead of using conductive glue. I've attached samples of another method for your review.

I truly hope this helps. If you have any questions or need more support, please feel free to reach out.

Attachment	Size
Core Termination.png (259.93 KB)	259.93 KB

Hi PI-Hueka,

So, for my better understanding, You are used a peace of foil without any glue? Photo You shown is 1:1 what I have now, but my foil is glued, and fully isolated from core...

Hi Viktor,

Yes, the piece of foil doesn’t need any glue between the cores. This method functions like grounding the core; as you noticed in your previous example, bus wires were soldered to one of the primary pins while the rest of the body rested on the laminated core without insulation. Then, tape was wrapped around the laminated core with the bus wire. If the foil or wire is fully isolated from the core, it won’t work effectively as a filter.

Thanks!

Just for FYI. In my case, works even isolated copper foil ( non-conductive glue ). This connection capacitance foil-core have a +/- same order as a snubber capacitance, and by result 1-2dB noise decreasing from sample to sample.

Hi Viktor,

Thank you so much for your prompt response.

You’re absolutely right in your observation about how it behaves like a snubber capacitance, given that the foil-to-core construction creates capacitance with the glue material. This approach differs from the methods in the reference designs, as we’re terminating the core to a quiet node, while your method effectively introduces filtering capacitance.