Optimal design for Current probe
flyboy7777, Tue Jan 12 2016, 10:21PMOkay, so I would like to build a decent current transformer for using with my scope... So my question is this: what sort of configuration will provide the least amount of phase shift, a low inductance and higher burden resistance? Would the cascaded 1089:1 versus a single 100 turn winding be better to avoid any phase shift?
I'm assuming that I would probably sacrifice some noise immunity if I have too high a burden resistance?
Thanks!
Re: Optimal design for Current probe
Wolfram, Wed Jan 13 2016, 04:08PM
This is a very interesting question. I've looked into it for a different project, and finding actual information can be a bit tricky.
A good starting point is figuring out what causes the phase shift in the first place. The magnetizing inductance of the CT forms a zero (high pass characteristic) with the burden resistance, which adds a phase shift and sensitivity dropoff at low frequencies, limiting the low end of the frequency range. It does not affect the response at high frequencies however. The leakage inductance of the CT might seem like a likely candidate, but since it appears in series with a current source (the measured signal) it shouldn't affect the frequency/phase response to any meaningful degree. I suspect most of the phase shift in practice will be caused by the CT winding capacitance (+ external load capacitance) forming a pole (low pass function) with the burden resistor. This will introduce phase shifts at the top end of the frequency range.
To minimize these phase shifts, you want as low capacitance and burden resistance as possible if you want to minimize the phase shift at any given frequency. The capacitance can be lowered by using fewer turns, which goes nicely along with the reduced burden resistance. The disadvantage with using fewer turns is that you get higher dissipation in the burden resistor for a given input current and output voltage, and the RMS and peak current handling goes down due to core saturation. Still, you need to reduce the number of turns significantly before this becomes a real problem. I think the optimum ratio in a typical TC case is far lower than 1000:1, I would go for something between 50:1 and 100:1 as a starting point.
Cascading CTs might make some difference in the phase shift you get, but I think the CT ratio and burden resistance are the main factors when it comes to phase shifts and bandwidth.
I found a very nice example in a book, shown in the following link:
. The circuit in question uses 20 secondary turns and a burden resistance of 2 ohms, for a sensitivity of 100 mV/A. The measured frequency response of this circuit is very good, the output signal is less than 1 dB down at 50 MHz.
Wolfram, Wed Jan 13 2016, 04:08PM
This is a very interesting question. I've looked into it for a different project, and finding actual information can be a bit tricky.
A good starting point is figuring out what causes the phase shift in the first place. The magnetizing inductance of the CT forms a zero (high pass characteristic) with the burden resistance, which adds a phase shift and sensitivity dropoff at low frequencies, limiting the low end of the frequency range. It does not affect the response at high frequencies however. The leakage inductance of the CT might seem like a likely candidate, but since it appears in series with a current source (the measured signal) it shouldn't affect the frequency/phase response to any meaningful degree. I suspect most of the phase shift in practice will be caused by the CT winding capacitance (+ external load capacitance) forming a pole (low pass function) with the burden resistor. This will introduce phase shifts at the top end of the frequency range.
To minimize these phase shifts, you want as low capacitance and burden resistance as possible if you want to minimize the phase shift at any given frequency. The capacitance can be lowered by using fewer turns, which goes nicely along with the reduced burden resistance. The disadvantage with using fewer turns is that you get higher dissipation in the burden resistor for a given input current and output voltage, and the RMS and peak current handling goes down due to core saturation. Still, you need to reduce the number of turns significantly before this becomes a real problem. I think the optimum ratio in a typical TC case is far lower than 1000:1, I would go for something between 50:1 and 100:1 as a starting point.
Cascading CTs might make some difference in the phase shift you get, but I think the CT ratio and burden resistance are the main factors when it comes to phase shifts and bandwidth.
I found a very nice example in a book, shown in the following link:
. The circuit in question uses 20 secondary turns and a burden resistance of 2 ohms, for a sensitivity of 100 mV/A. The measured frequency response of this circuit is very good, the output signal is less than 1 dB down at 50 MHz.Re: Optimal design for Current probe
flyboy7777, Wed Jan 13 2016, 06:50PM
Interesting. It's a bit more complicated than I had realized.
flyboy7777, Wed Jan 13 2016, 06:50PM
Interesting. It's a bit more complicated than I had realized.
Print this page