Wide-band RF filtering for mains - some interim conclusions
The 1.8mm iron wire arrived and it did indeed have a 1.2mm internal diameter which is perfect for winding small chokes. I built a second box with steel chokes in but the third box I tried with the iron wire coils. Having insulated wire means the coils are smaller and I was able to fit the second tier filter into the same box.
I wound a couple of NiZn toroids with iron wire, but not as common mode chokes. When I did the sums I found that for the currents I'm running at (under 3A total draw) the toroidal core isn't close to saturation. Hence I decided on separate inductors to gain some differential mode filtering on top of the common mode.
The next stage of development has been trying to wind multilayer iron wire air-cored chokes. For this I've had trouble finding formers which allow more than 4 layers. The reason for wanting more layers is - this makes the chokes much more lossy. Proximity effect helps concentrate the higher frequency currents only in areas where the windings are adjacent to each other. This can be seen on the LCR meter - with 4 layers I was getting almost 8X as much loss at 100kHz as at 50Hz. Since nobody makes formers which lend themselves to such lossy chokes, I had to do a little lateral thinking. PTFE tape is sold on plastic reels which are a promising line of enquiry, but after unwinding the PTFE I found the depth didn't allow more than 4 layers of the 1.8mm diameter wire.
So I ordered more wire, of 1.5mm diameter to see if I could squeeze in 5 layers. This wire has turned out to be a pain to wind - seems its not really iron but steel in disguise. Its far too springy! I may return to this if I find a more promising coil former or a source of real iron wire in roughly 0.8mm conductor diameter.
For now this project's on the back burner but the three filters I've so far built are in daily use. No formal listening tests though.
I wound a couple of NiZn toroids with iron wire, but not as common mode chokes. When I did the sums I found that for the currents I'm running at (under 3A total draw) the toroidal core isn't close to saturation. Hence I decided on separate inductors to gain some differential mode filtering on top of the common mode.
The next stage of development has been trying to wind multilayer iron wire air-cored chokes. For this I've had trouble finding formers which allow more than 4 layers. The reason for wanting more layers is - this makes the chokes much more lossy. Proximity effect helps concentrate the higher frequency currents only in areas where the windings are adjacent to each other. This can be seen on the LCR meter - with 4 layers I was getting almost 8X as much loss at 100kHz as at 50Hz. Since nobody makes formers which lend themselves to such lossy chokes, I had to do a little lateral thinking. PTFE tape is sold on plastic reels which are a promising line of enquiry, but after unwinding the PTFE I found the depth didn't allow more than 4 layers of the 1.8mm diameter wire.
So I ordered more wire, of 1.5mm diameter to see if I could squeeze in 5 layers. This wire has turned out to be a pain to wind - seems its not really iron but steel in disguise. Its far too springy! I may return to this if I find a more promising coil former or a source of real iron wire in roughly 0.8mm conductor diameter.
For now this project's on the back burner but the three filters I've so far built are in daily use. No formal listening tests though.
Total Comments 0
