27-11-2014, 02:39 PM
Mark,
I did a bunch of work of very high linearity, high frequency current transformer designs for my DRSSTC projects (which run up to about 500kHz).
Getting an "ideal rectifier" that works with low input voltages is tricky, as is matching impedances...
The circuit below was very close to my final design - I like Rod Elliot's pages, but, as he says, they are the basic stuff. I came across a neglected TI app note SBOA068, "Precision Absolute Value Circuits" that eventually put me on the right path (for my app, anyway!) - it's an excellent read and I can't understand why it's not more widely known about... for your app, you only need IC1a/b - the rest is to do with ultra-high-speed over-current detection.
I'll look up the actual CT I used with the 10R burden resistor/load (R7) - it was tricky finding the right one.
Should you wish, I'm happy to loan you a Pearson CT to check your readings - these are normally enormously expensive, having a flat response into the high MHz regions - they are based on Rogowski coils - I have a few different types and they are essential for accurate work.
HINT: I used Linear Technology devices, not just because they are very good, single-rail & rail-to-rail, but also that I use LTspice to simulate stuff and accurate models are obviously included for all the LT stuff in LTspice IV - I've probably got the ASC files for this somewhere if you want to have a play with it.
I did a bunch of work of very high linearity, high frequency current transformer designs for my DRSSTC projects (which run up to about 500kHz).
Getting an "ideal rectifier" that works with low input voltages is tricky, as is matching impedances...
The circuit below was very close to my final design - I like Rod Elliot's pages, but, as he says, they are the basic stuff. I came across a neglected TI app note SBOA068, "Precision Absolute Value Circuits" that eventually put me on the right path (for my app, anyway!) - it's an excellent read and I can't understand why it's not more widely known about... for your app, you only need IC1a/b - the rest is to do with ultra-high-speed over-current detection.
I'll look up the actual CT I used with the 10R burden resistor/load (R7) - it was tricky finding the right one.
Should you wish, I'm happy to loan you a Pearson CT to check your readings - these are normally enormously expensive, having a flat response into the high MHz regions - they are based on Rogowski coils - I have a few different types and they are essential for accurate work.
HINT: I used Linear Technology devices, not just because they are very good, single-rail & rail-to-rail, but also that I use LTspice to simulate stuff and accurate models are obviously included for all the LT stuff in LTspice IV - I've probably got the ASC files for this somewhere if you want to have a play with it.
sıʌǝɹq ɐʇıʌ `ɐƃuol sɹɐ
ʞɔıu
ʞɔıu
