0
\$\begingroup\$

I would like to detect if a current flows in a 3-phase industrial environment where currents can go up to 120A. Anyway, for me is only important information current is flowing/current is not flowing. The threshold should be somewhere about 0.5A - 1A. I excluded shunt and current transformer solutions from consideration because I cant put shunt anuwhere (but I also think it is not quite applicable for such high current environments) and the later seems to be more expensive than solutions with Hall-effect sensor which I would like to consider here. In this question I would like to consult you about the principles in order to choose the one that fits the best my application.

Is there a Hall-effect sensor which is able to provide a digital output based on the absolute value of the magnetic field? I checked various types of Hall-effect sensors, bipolar, unipolar and omnipolar. The closest to what I want is unipolar but none of unipolar sensors can actually treat south and north poles in the same way. By this I mean, that threshold should be for +- x T. Using such an element I believe I would minimize my BOM.

Since all three phases are in vicinity of each other, how sensitive might be using Hall sensors in such environments in terms of flux cancelations?

Would it be possible to implement this without a ferrite ring for concentrating magnetic flux? My calculations give in the worst case, 3mm away from the conductor about 0.5G (2/3e-4 T).

Is there any other interesting way of doing this that I missed completely?

\$\endgroup\$
  • \$\begingroup\$ "Is there any other interesting way of doing this that I missed completely?" Measure the voltage drop along some length of a conductor. In other words, since the existing cable has a non-zero resistance, you can use it as a shunt of sorts. Depending on the length and resistance of the wire, the voltage drop that you can measure might be tiny, so this may not be practical. \$\endgroup\$ – Dampmaskin Dec 9 '19 at 14:43
  • \$\begingroup\$ @Dampmaskin I can't, everything must be packed more or less in a wall plug, so I don't have access to more than 1cm of the conductor. \$\endgroup\$ – Nexy_sm Dec 9 '19 at 14:46
  • \$\begingroup\$ You said this I checked various types of Hall-effect sensors and unfortunatelly I can't seem to find one. Please list the sensors that you checked and provide datasheet links. \$\endgroup\$ – Andy aka Dec 9 '19 at 14:48
  • \$\begingroup\$ @Andyaka I edited my question a bit. I can't seem to find a sensor which can ignore the sign of the field and to detect just the magnitude. At the moment I like the most DRV5013 from TI. ti.com/lit/ds/symlink/drv5013.pdf \$\endgroup\$ – Nexy_sm Dec 9 '19 at 15:00
  • \$\begingroup\$ Current transformer should work fine with minimal noise... Area only the field area inside the loop adds signal. \$\endgroup\$ – MadHatter Dec 9 '19 at 18:29
1
\$\begingroup\$

120A is at least 6mm2 of cable.

A current transformer is then the typical way to measure current. However, that is very big. Especially for three. And may be expensive at the precision you require.
You can get a bit smaller with a hall sensor current transformer. Such as LEM LZSR.

enter image description here

When even that won't fit then you can go looking for separate magneto resistive current sensors. At cost of a significantly more complicated design, validation and calibration.
Sensitec and Allegro Microsystems are famous for these.


Using a traditional discrete hall sensor will probably not work reliably at the low ranges you are looking for. If you can find them with proportional output at all.

\$\endgroup\$
  • \$\begingroup\$ It looks like most of the current sensing things are not designed for the application I described, they are either optimized for precise current measurements or for detecting presence of magnets. The problem with what you suggested is that I don't have space where to fit that on my pcb. However, a relay which has self holding capability while the current goes throuh it might help me in the case that it is equiped with an auxilliary contact where one can read its state. Unfortunatelly I don't know the name of such a relay in English. In german it is called nullspannungsschalter. \$\endgroup\$ – Nexy_sm Dec 9 '19 at 16:09
  • \$\begingroup\$ @Nexy_sm You run 120A trough your PCB but don't have space for a LEM module or magnetoresistive sensor? \$\endgroup\$ – Jeroen3 Dec 9 '19 at 18:11
  • \$\begingroup\$ Yes, because conductors for 120A are just passing through the PCB (they are perpendicular to the PCB plane), and the space around them is quite tight. \$\endgroup\$ – Nexy_sm Dec 10 '19 at 8:36
  • \$\begingroup\$ @Nexy_sm Another option is to get custom ferrite core with hall sensor, such as split core AC/DC clamps. But that is beyond my knowledge. \$\endgroup\$ – Jeroen3 Dec 10 '19 at 9:12
  • \$\begingroup\$ Would it be maybe possible to get somehow state of the SCR by measuring its gate voltage. Something in that direction could help detect that no current is flowing without magnetic field detection. \$\endgroup\$ – Nexy_sm Dec 10 '19 at 13:28

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.