Jee Labs Talk

This forum has moved to forum.jeelabs.net - this read-only archive is for reference only!

1. jcw
   

   Does anyone know a cheap way to detect whether a wire is carrying an AC current?

   I'd like to detect whether current is flowing through a power wire, without galvanic connection. Yes, there are current transformers, and current sense coils, but these are all relatively expensive because they either support large currents or have good linearity for accurate measurement. I just want to know whether some current over say 0.1A is passing through the wire, and it won't normally be more than a few amps.

   What sort of signal would one get from wrapping a few windings by hand around a wire, and then rectifying it? I can't quite figure out whether that would be mere millivolts or kilovolt spikes :)

   Background: if I can place lots of really low-cost on/off sensors around the house, then I could measure the actual change in current draw in the central meter. Things like a fridge, appliances, lights ... (actually, for lights one could even use a light sensor).

   Posted 12/10/2009 17:21:31

2. joop
   

   jc

   I think this is not so easy ,see http://www.faqs.org/patents/app/20090021237
   for some info.

   Joop

   Posted 12/10/2009 17:36:34

3. joop
   

   jc

   Maybe it is easy,
   I just found this http://openenergymonitor.org/emon/node/28

   and software for arduino

   joop

   Posted 12/10/2009 17:44:24

4. kintel
   

   We use a solution which I think is pretty much like Joop posted at the Metalab for detecting when our laser cutter is actually cutting, for accounting purposes. Admittedly pretty high current, but it works.

   -Marius

   Posted 12/10/2009 21:02:38

5. jcw
   

   Yes, I know about current transformers... and hall effect current sensors such as the ACS712.

   But, but, but - if the AVR is sensitive/high-impedance enough to act as touch sensor:
   http://sensi.org/~svo/capsensor/index.en.html

   Isn't there a way to detect current flowing through a wire with virtually nothing? Like a few turns of one wire twisted around the other? Or even wrapped around a nail parallel to the wire, picking up its magnetic field?

   I guess I should tinker a bit. I was hoping there would be some ready-made solution... am really looking for a near-zero priced solution, because then it could be used in dozens of places around the house.

   As I said: all I need is to detect a bit of AC current flow, not measure it.

   Posted 12/10/2009 22:10:09

6. kintel
   

   FYI:
   This is the hall effect sensor-based solution we use at the Metalab for laser cutter billing:
   http://svn.clifford.at/handicraft/2009/currentsens/

   Posted 23/10/2009 16:49:43

7. andras
   

   If you would wrap a few turns of wire around the mains wire, I think nothing would happen. The magnetic field lines around the mains wire look like "circles" around the wire, so they would then be tangential to the wrap-wire. If you want to induce a voltage, the magnetic field must be perpendicular to the wrapped coil turns.

   The wire-wrapped nail parallel to the mains wire will also not induce a voltage. Although the magnetic field is now perpendicular to the wires, it is still not perpendicular to the coil turns (that is, the area of the wire-loops).

   So then you'd get back to the current transformer principle. In a current transformer a ring is put around the mains wire. The ring follows the path of the magnetic field. Then wrap wires are turned around this ring. This way the magnetic field is nicely "fluxing" through the wire turns.

   I suppose you could make such a transformer yourself, but I think quite a few turns are needed in order to get a measurable result. The flux density in the direct vicinity of a wire with 1A running through it is in the order of E-4 Tesla. I'm not sure if the following calculation is correct, but if you'd put a ring around it with a 1cm diameter and wrap a 1000 turns around it, then I'm calculating a voltage in the order of 10 microvolts.

   I could be missing something here though; there's one thing that I don't get about current transformers. They're described as if they are like current sources on the output terminals with a fixed current ratio, but I would think that just like an ordinary transformer a voltage is induced and the output current depends on the load on the output terminals.

   Posted 08/12/2009 21:58:33