Wide-band RF mains filtering - part II
Posted 14th February 2011 at 05:24 AM by abraxalito
Updated 2nd March 2011 at 12:10 PM by abraxalito
Updated 2nd March 2011 at 12:10 PM by abraxalito
I've been actively researching via Google on this project since the last posting and found out a few interesting things which I'll share here.
Firstly I checked out piano wire as the material for making the L1 & L2 input side chokes in my original schematic. I couldn't find any consistent figure for what the permeability of piano wire is - it seems steel comes in so many variants that its hard to tell. So I continued to look for alternative materials which have reasonably high permeability. Remember, its the high permeability which gives rise to skinny skin depths and hence higher losses as the frequency climbs. Normally this is really undesirable, but for our purposes its exactly what we're looking for.
Other elements which have high permeability include nickel, cobalt and iron. Nickel wire is available, but its expensive and not widely sourced. Cobalt is even more expensive with even fewer sources. Iron is the cheapest option and it turns out that iron wire is fairly easy to get hold of, though not from normal electrical vendors. The last option I looked at was mumetal, an alloy developed for extreme permeability. Mumetal wire is available, though not widely, and its not massively expensive in the lengths we'd need. The primary disadvantage is its need for an annealing process after its been bent to ensure its amazing properties. This process is pretty hard to carry out for diyers as it involves a reducing atmosphere and very high temperatures. For those who have access to the right process, mumetal is the gold standard for this application. However for the rest of us, I believe iron can be made to work well and is considerably cheaper.
Iron wire I found is used in gardening, to create attractive shapes to grow and drape flowers through. It comes insulated in green PVC but I'm not going to trust that insulation alone as its there to protect it from rust rather than people from shocks. I'm going to order some from Taobao, the price is 3.5rmb for 40m of 1mm diameter. That's really peanuts. I reckon I'll need 2m or so for each choke. Given that its so cheap I might run two chokes in series for even better noise absorption. I'll also get hold of some plastic boxes to put the chokes in.
My original suggestion of 0.5mm wire for these chokes turns out to be misguided. To get the best noise absorption, a larger diameter is better as that results in a wider band of frequencies where the losses are climbing at 3dB/octave. Hence my choice of 1mm diameter, which is turns out is probably the smallest I can get it in easily.
In part III I'll talk about the L3 choke.
Firstly I checked out piano wire as the material for making the L1 & L2 input side chokes in my original schematic. I couldn't find any consistent figure for what the permeability of piano wire is - it seems steel comes in so many variants that its hard to tell. So I continued to look for alternative materials which have reasonably high permeability. Remember, its the high permeability which gives rise to skinny skin depths and hence higher losses as the frequency climbs. Normally this is really undesirable, but for our purposes its exactly what we're looking for.
Other elements which have high permeability include nickel, cobalt and iron. Nickel wire is available, but its expensive and not widely sourced. Cobalt is even more expensive with even fewer sources. Iron is the cheapest option and it turns out that iron wire is fairly easy to get hold of, though not from normal electrical vendors. The last option I looked at was mumetal, an alloy developed for extreme permeability. Mumetal wire is available, though not widely, and its not massively expensive in the lengths we'd need. The primary disadvantage is its need for an annealing process after its been bent to ensure its amazing properties. This process is pretty hard to carry out for diyers as it involves a reducing atmosphere and very high temperatures. For those who have access to the right process, mumetal is the gold standard for this application. However for the rest of us, I believe iron can be made to work well and is considerably cheaper.
Iron wire I found is used in gardening, to create attractive shapes to grow and drape flowers through. It comes insulated in green PVC but I'm not going to trust that insulation alone as its there to protect it from rust rather than people from shocks. I'm going to order some from Taobao, the price is 3.5rmb for 40m of 1mm diameter. That's really peanuts. I reckon I'll need 2m or so for each choke. Given that its so cheap I might run two chokes in series for even better noise absorption. I'll also get hold of some plastic boxes to put the chokes in.
My original suggestion of 0.5mm wire for these chokes turns out to be misguided. To get the best noise absorption, a larger diameter is better as that results in a wider band of frequencies where the losses are climbing at 3dB/octave. Hence my choice of 1mm diameter, which is turns out is probably the smallest I can get it in easily.
In part III I'll talk about the L3 choke.
Total Comments 0
