VK6ZFG wrote:The discussion has been on how good is one balun design vs another. In other words, the quest for the ultimate in balun design.
What has not been touched on in all has been what is an acceptable level of isolation. Here I would imagine having at least 15dB of isolation should suffice in most instances. However I stand to be corrected on this as this statement as it is not based on any research on my part.
It's a good question but there isn't any universal answer, because problems with common-mode current can vary so much. How good a choke is needed will depend on how much of a problem there is... which is extremely difficult to quantify.
What Igor measures as "isolation" is specific to his particular type of test equipment, which measures the isolation or loss when the choke is inserted between two 50-ohm test ports. However, it's always better to translate those results into an equivalent value of
impedance, because impedance tells you more about how the choke will actually perform.
Experience tells us that the old "500-ohm" wet finger estimate is often pretty useless. 500 ohms isn't enough impedance to suppress common-mode currents reliably in every application, and may also lead to the ferrite cores overheating with high power. That is why most writers today are suggesting at least 5000 ohms and preferably higher, to make sure that the choke will be able to tackle even the worst common-mode problems. But chokes like that tend to use large ferrite cores, and also special types of ferrite which can be expensive. We don't want to overspend on big ferrite cores for problems that don't actually need it.
So what do we do? Do we start with cheaper but less effective chokes and work our way up? Or do we always go for the 'big ferrite' to nail the problem first time, even though we might be overspending? It probably depends. If it's entirely my own problem then I might experiment a bit; but if it involves other people, then I usually go straight for the 'big hammer'.
That's why I haven't answered your question, Igor.
73 from Ian GM3SEK
) to use a more complicated approach?
