Designing a current sensor for 50mA up to several dozen Amps

[arivel]

Hi everyone .with your help I would like to understand if there is the possibility of creating a current sensor with the method that I illustrate below. it has to measure from 50mA up to a few dozen A, I need maximum precision and linearity.it is similar to the current transformers that already exist but those are made to measure large currents and use a ferromagnetic core.

what I have in mind are many very small toroidal plastic cores (therefore permeability similar to that in air) placed next to each other in order to form a straight line because the wire traversed by the alternating current to be measured passes through the central hole of each one.each toroid is wrapped in one or more turns of turns, this is to be decided later because a compromise must be made between number of turns, inductance, impedance and DC resistance.

the output of the last turn of the first toroid is connected to the first turn of the second toroid and so on until the last one, therefore the turns of all the toroids are connected to each other. this is to exploit a greater quantity of magnetic field (which is at most in contact with the wire where the electric current passes) and to have a single winding.

to understand better, it would be like having many Rogowski coils connected in series but without space between the conductor to be measured and the detection coil.the main question is the following: each toroidal winding provides me with a voltage, since each winding is connected in series can I obtain a voltage resulting from the sum of all the voltages produced by each toroidal coil?.

Thank you

 

[Baluncore]

DC? or AC over what range of frequency range?

If you get the polarity correct, you can stack many Rogowski coils in series to effectively multiply the output. That is because it is the induced voltage that is sensed.
https://en.wikipedia.org/wiki/Rogowski_coil

 

[arivel]

audio signal 20Hz - 20KHz

 

[arivel]

DC? or AC over what range of frequency range?

If you get the polarity correct, you can stack many Rogowski coils in series to effectively multiply the output. That is because it is the induced voltage that is sensed.
https://en.wikipedia.org/wiki/Rogowski_coil

multiply?

 

[Baluncore]

multiply?

One induced voltage per coil, multiplied by the number of Rogowski coils you have.

 

[arivel]

with the same current.
with the same number of turns.
in the absence of space between the sensing winding and the current-flowing conductor.
if I increase the diameter of the conductor, the field intensity should decrease, is this correct?.
keeping in mind the characteristics written above, if instead of a full section conductor I use a thin-walled copper or aluminum tube, what happens to the magnetic field?.

 

[Baluncore]

You are designing something that is becoming more like a directional coupler, than a current transformer.

Have you considered a current sense resistor?
Why not just monitor the voltage drop along one output cable, by feeding the far-end voltage, back along the output cable through a separate wire in the cable.

 

[berkeman]

audio signal 20Hz - 20KHz

with the same current.
with the same number of turns.
in the absence of space between the sensing winding and the current-flowing conductor.
if I increase the diameter of the conductor, the field intensity should decrease, is this correct?.
keeping in mind the characteristics written above, if instead of a full section conductor I use a thin-walled copper or aluminum tube, what happens to the magnetic field?.

Instead of us playing 20 Questions with you, how about you just show your cards and tell us what you are trying to do? This is getting tiresome.

To make the current measurement that you've defined so far (as best as I can tell), you can just use a production Hall Effect current clamp like one from Tek or other vendors. What is wrong with that?

https://www.newark.com/tektronix/tcp202a/probe-current-1-oscilloscope/dp/24W8886

 

[arivel]

I need it to create a feedback network for a bridge power audio amplifier.
sensing resistor is one way to go but I want to explore other alternatives.
in my case the detection resistor is not without problems, even if small (0.1 Ohm), at high volume its noise becomes significant and is amplified.
the current sensor has the task of detecting the current flowing in the speaker then the feedback network has the task of correcting the response of the two bridge amplifiers

 

[berkeman]

I need it to create a feedback network for a bridge power audio amplifier.

The reactance will be an issue that you have to deal with in the design of this feedback network. It will likely take a bit of work and simulation to get the system stable. How much experience do you have in designing feedback systems?

the detection resistor is not without problems, even if small (0.1 Ohm), at high volume its noise becomes significant and is amplified.

What kind of noise? Thermal noise?
$$v_n = \sqrt{4kTRB}$$
https://en.wikipedia.org/wiki/Johnson–Nyquist_noise

 

[arivel]

thermal.
using a transimpedance and operational integrator amplifier you can obtain a flat response

 

[DaveE]

Whatever you can build yourself, these guys will sell you a better version.
https://www.lem.com/en/product-list?measurement=52

There are other manufacturers too. A hall effect based flux balance sensor is the way to go here.

 

[arivel]

https://www.akm.com/content/dam/documents/products/current-sensor/ak310h/ak310h-en-datasheet.pdf

 

[Vanadium 50]

Whatever you can build yourself, these guys will sell you a better version.

"We have a buy or build decision point here and..."
"Buy"
"But we haven't completed the analysis and..."
"Buy".
"But..."
"Buy"

 

[DaveE]

"We have a buy or build decision point here and..."
"Buy"
"But we haven't completed the analysis and..."
"Buy".
"But..."
"Buy"

Yep 100%! NRE is #$@#ing expensive in so many ways. At the laser company our EEs never designed anything we didn't absolutely have to.

Not so true for low cost high volume products though. I had a friend consulting many years ago at a famous mouse manufacturer. They thought they could leave out a pull-up resistor, so they built and tested 20,000 of them that way to see if they could save the cost of a single resistor.

 

[DaveE]

https://www.akm.com/content/dam/documents/products/current-sensor/ak310h/ak310h-en-datasheet.pdf

Also Melexis, if you want an open loop hall sensor. That type may have a problem with the >60dB dynamic range.

Since it's only audio frequencies, you have a lot of good choices. Building your own probably isn't one of them, unless you have strange requirements.

 

[arivel]

with reference to my post 6.
just one thing I'd like to know.
Is there a magnetic field inside the tube or not?

 

[arivel]

Bye and thanks for the answers .
To do a quick test I made 5 small plastic toroids wrapped in 0.20 mm enamelled copper wire, then I connected them together.
the internal diameter of each toroid is 6 mm and the thickness is 2 mm.
I measured the total inductance of the 5 toroids with a special device which measured 6.5 micro H.
in the conductor that occupies all the internal space of the toroids and passes through them all, I passed a 50mA alternating current at a frequency of 1Khz then I connected the output of the toroid chain to the oscilloscope but no signal was displayed.
I don't know if I did something wrong in the circuit or if the voltage produced is too small to be detected by the oscilloscope.
can you give me some help? I calculated 325 nano volts but I would like to be sure of this number, for you it would be child's play.