Aiming for maximum isolation
Posted 2nd November 2014 at 03:42 AM by abraxalito
Getting isolation from the mains supply is an on-going challenge for me. Here's my latest stab at getting as close to a battery as I can currently manage - a ferrite line-output trafo soon to be pressed into service as a low-capacitance isolation transformer.
The back story is that I'd heard that split bobbin trafos were the way to go for the best isolation from the mains. Having just acquired some for chipamp-drive unit interfacing I was curious to check out their credentials. I measured around 20pF interwinding capacitance which is about the lowest of any mains trafo I've checked. Of that it seems that the majority of the capacitance is directly between the windings, a smaller amount is due to coupling via the steel core (which of course is conductive). So then I figured that even if the windings could be moved infinitely far apart the steel core would be the limiting factor in reducing the capacitance.
Air core trafos are a possibility but then they're hopeless at mains frequency, there would need to be some inverter on the mains side to step up the frequency to something useful. But operating at 50Hz, a ferrite core might be possible as that's a good insulator.
Ferrite cores are never used for mains trafos because the achievable flux is so low that they're very inefficient. But for only a very low power demand such as a DAC I decided to have a play, and the pic shows the prototype. Its not a direct-from-mains trafo for a couple of reasons. Firstly I'm too lazy to wind on the required number of turns - due to the low flux a ferrite trafo needs many more turns than a steel one and I just don't have the patience to wind on enough for a 220V primary. I estimate over 16,000T would be needed. Second the safety aspects with such thin wire bother me - what's its insulation quality?
Hence this isolation trafo works in series with a normal EI core trafo. The EI steps the mains voltage down to 5V and this trafo is a 1:1. Less efficient of course but if efficiency was a major concern I'd not be using ferrite
I have measured just under 1pF of interwinding capacitance (more than 20X better than the split bobbin) but I suspect the real value is a bit lower than the measurement. So I'll measure again once I have this fully built up.
The back story is that I'd heard that split bobbin trafos were the way to go for the best isolation from the mains. Having just acquired some for chipamp-drive unit interfacing I was curious to check out their credentials. I measured around 20pF interwinding capacitance which is about the lowest of any mains trafo I've checked. Of that it seems that the majority of the capacitance is directly between the windings, a smaller amount is due to coupling via the steel core (which of course is conductive). So then I figured that even if the windings could be moved infinitely far apart the steel core would be the limiting factor in reducing the capacitance.
Air core trafos are a possibility but then they're hopeless at mains frequency, there would need to be some inverter on the mains side to step up the frequency to something useful. But operating at 50Hz, a ferrite core might be possible as that's a good insulator.
Ferrite cores are never used for mains trafos because the achievable flux is so low that they're very inefficient. But for only a very low power demand such as a DAC I decided to have a play, and the pic shows the prototype. Its not a direct-from-mains trafo for a couple of reasons. Firstly I'm too lazy to wind on the required number of turns - due to the low flux a ferrite trafo needs many more turns than a steel one and I just don't have the patience to wind on enough for a 220V primary. I estimate over 16,000T would be needed. Second the safety aspects with such thin wire bother me - what's its insulation quality?
Hence this isolation trafo works in series with a normal EI core trafo. The EI steps the mains voltage down to 5V and this trafo is a 1:1. Less efficient of course but if efficiency was a major concern I'd not be using ferrite
I have measured just under 1pF of interwinding capacitance (more than 20X better than the split bobbin) but I suspect the real value is a bit lower than the measurement. So I'll measure again once I have this fully built up.Total Comments 0
