How to detect presence/absence of mains (220-240VAC) current ?

Hi,

In my search so far, I've come accross the following 3 technologies for sensing presence / absence of AC mains (220-240V) current.

1. Hall Effect sensors
2. Current Shunt sensors
3. Current Transformer sensors

My need is not to measure (as in metering/quantitative measurement), but only presence/absence of AC current, given whatever load (domestic, but say upto 4KVA). So which one might be the recommended ? All other things being equal, least cost, ability to miniaturize, and low power dissipation are other objectives, in that order.

If there are specific makes/models P/Ns, esply available in India, would like recommendations.

Also, my intent is to use 433/315MHz ASK RF comm. between the current sensor and main controller. Could I expect some RF interference issues in wiring up such a sensor, and if so, what could I do to prevent it.

thanks,
jay.

Hello Jay,

I would not go for the current shunt. Having insulation will make it easier for you to meet general safety standards. If price and space permits, I would go for the current transformer. It is not a high tech component, and you don't need high accuracy, so it should be possible to source it.

I do not expect interference issues between a current transformer sensor and your RF module or the other way around. If you make sure to have input protection at the interface of your module, it should be fine.

The input protection may be required because mains wiring can have nasty transients (depending on so-called "installation category") that will capacitively couple from mains wiring to sensor wiring.

Hi WimRFP, thanks for the response.

WimRFP said:

I would not go for the current shunt. Having insulation will make it easier for you to meet general safety standards.

Click to expand...

Insulation on which part are we talking about over here ? I was hoping to achieve this without any mains splicing/dicing... i.e. non-intrusive installation (if possible). Do you think the Hall-effect sensor may be better in that regard ? I was hoping to just put crocodile clips over the insulated mains wiring, and see if something of that sort works !! Maybe that's contradicting physics.

WimRFP said:

If price and space permits, I would go for the current transformer. It is not a high tech component, and you don't need high accuracy, so it should be possible to source it.

Click to expand...

Price and space are 2 things which are a priority. Space being low would allow this thing to fit in the mains gangbox (or whatever those are called, where the mains fuse and primary circuit-breakers are put).

WimRFP said:

I do not expect interference issues between a current transformer sensor and your RF module or the other way around. If you make sure to have input protection at the interface of your module, it should be fine.

The input protection may be required because mains wiring can have nasty transients (depending on so-called "installation category") that will capacitively couple from mains wiring to sensor wiring.

Click to expand...

I guess these are those "spark quenchers", I figured out on another thread ?

Hello Jay,

For installing a current transformer you have to remove a wire to be able to install the device. You are right, the best place is close to the meter/ fuse box as this keeps all mains wiring at a safe place. Other option (more expensive) is to use a split core design (like the ones used in EMC suppression). Advantage is ease of installation, but mechanical design is more complicated (also with regards to electrical safety).

I am not familiar with hall sensors for such applications, but you may need to locate them close to the wiring to get good S/N ratio. For 1A mains current, you may expect 8A/m (rms) at 0.02m distance (that is 0.1 Gauss). Of course adding magnetic material will increase the Gauss/A ratio.

If your circuit can be touched by the general public or people that are not instructed to work on live circuits, safety requirements are more strict. The requirements are laid down in standards (for example IEC 950 / EN60950 for IT equipment). You should think of minimum requirements for distance through material, creepage and clearance distance, flammability, etc.

I would advise you to study at least some electrical safety documents and to figure out the expected level of mains transients. It will save you lot's of redesign time.

Regarding input protection. As the coupling between mains and your circuit is via a small capacitance, protection can mostly be done with diodes towards supply rails, zener diodes and some RC filtering. You don't need (expensive) discharge tubes. If the wiring between current sensor and your circuit is very large (think of tens of meters), energy in transients may be higher.

For the above I assumed that the electricity company installs basic over voltage protection at the mains entrance (I am not familiar with the situation in India).

1. To sense presence of current you can look at the hall effect sensors. But it will be a mechanical challenge to create a air gap and insert a small PCB along with sensor inside.
+ if you have specific current cut off to sense it will be difficult in mass production.

2. Shunts are not advisable due to lack of isolation

3. CTs are better and safe option but will not help you at lower current.

If you give few more details on product and exact feature requirement , can give you better ideas as I have already done some work on this.

To sense the AC current, I use a low cost small transformer having two separate bobbins. I replace its secondary bobbin (low voltage) with a few turns of an insulated thick wire which will be connected later in series with the AC load. The number of turns (which might be 2 and up) determines the sensitivity and the thickness of the wire must be chosen to be heavy enough for the possible maximum AC current. Now what was called the primary bobbin (with N1 turns) becomes secondary and its output current will be approximately proportional to N2/N1 (where N2 is the few turns made as primary, input).

Hope this help.

Kerim

the problem is that all these solutions involve breaking the mains line and inserting the sensing element in series with the load, and this could be dangerous..

Hall effect sensors should be good for the job, similar to the ones used in clamp type multi-meters...

thanks
a

Another way is with a NE2 neon bulb with 180K in series and a LDR. Glue the LDR on the NE2 and seal it light tight. When mains is
available, the NE2 glows and the LDR will have a resistance of about 1K. When the mains fail, the LDR resistance will be 1M or larger.
Easy and safe to implement.

toninph said:

Another way is with a NE2 neon bulb with 180K in series and a LDR. Glue the LDR on the NE2 and seal it light tight. When mains is
available, the NE2 glows and the LDR will have a resistance of about 1K. When the mains fail, the LDR resistance will be 1M or larger.
Easy and safe to implement.

Click to expand...

Cool. Seems like a easy and elegant option. Should be reliable too (weakest link being "light tight").

Is it about voltage or current sense?!

My requirement was current-sensing on mains power -- at least-cost with good reliability and something that requires minimal intrusion with mains electrical wiring.

current transformers may be dangerous, this of course depends on the kind of load on the AC line. The higher the load the more dangerous it gets, HALL effect sensors may be a good choice, or the kind used in digital clamp meters, maybe you could salvage one from a unused clamp meter. You may have to amplify the signal using an op-amp.

thanks @arbj. I used to think that the clamp meters use Hall-effect sensors in those clamps... maybe I am wrong.
I do need have 4-5 prototypes, so salving clamps seems a very expensive option

I remember that the voice signal on phone lines could be tapped by an external small magnetic receptor. So I wonder if placing a coil with an open core near one of the two appliance wires can give enough volt and current to detect the presence of the current of interest.

Jayachar88,
To make the NE2/LDR light tight, shrink sleeve may work best. You may have noticed that this solution is just detecting mains on or off by a changed resistance of the LDR. It is not measuring voltage or current at all. The output (LDR) is totally isolated from mains (if built correctly of course) :smile:

If you can get it in between the live and neutral, Hall effect components should be of use...!
I've looked at this before to detect when I've left the freezer unplugged.... I have a "mains test pen" that lights up near mains (No physical connection) so that became the alarm!!!
NEAL