[time-nuts] NIST isolation amplifiers
csteinmetz at yandex.com
Wed Nov 26 15:54:21 EST 2014
>Almost. 1:1:2 (turns ratio) transformers used in each stage and 1:1
>transformer on input. This allows a lower power supply voltage to be used.
I spent a little time (emphasis on "little") fiddling with the
simulation, and I did not immediately find any solution with 1:1:2
and 1:1 transformers that I liked as well as the design with 1:1:1
and 1:2 transformers. For those who are winding their own
transformers (which I recommend, partly for the reason given below),
the simplicity of three equal windings may, by itself, outweigh any
potential advantage of using a lower power supply voltage. Each
constructor should evaluate this for him- or herself.
>One thing to watch with minicircuits transformers is core saturation due to
>dc flowing in the windings.
Good point, I too have found that some MCL transformers have skimpy
cores. I have no idea whether the MCL T622 (1:1:1) or T613 (1:1:2)
would work in this circuit. They are both specified for 30mA, and I
had the 3904s biased at 20mA -- but I'm not sure what MCL means by
30mA. In this circuit, 20mA flows in the same direction in two of
the windings. If the 30mA applies only if one winding has DC flowing
(or more generally, if the allowable net DC is equal to 30mA through
just one winding), then the core would not be adequate.
More information about the time-nuts