how do clamp meters work?

Hello John,

Use Google to search "current transformer". Then you'll know ;-)

A power cord carries current back and forth so if you run the whole cord through a clamp meter it should not read anything. The field cancels because of opposite phases of current flow between line and neutral. If it does read something make sure that the equipment at the other end of this power cord is taken out of service and repaired. Because that would mean some of the current returns outside the power cord which is dangerous.

BTW, this is also how GFCI switches work. They rely on the fact that all current phases cancel because all wires that can carry current go through the core. PE doesn't so if a leaky device lets enough milliamps slip onto PE the balance is upset and the GFCI is supposed to trip.

You won't see much if you enclose both wires in your clamp... You have to isolate one single current carrying conductor to be enclosed by the clamp. Then you have to design the thing.

Most of the magnetic field will be between the conductors, but there will probably be a detectable field close to one of the wires. An air core inductor with many turns of fine wire can be used as a magnetic field sensor, and you can probably detect some voltage if you connect it to a sensitive voltmeter or scope, and position the coil properly. There will also be some component from other EM radiation, which may also be directional.

Paul

Why not use an electrostatic sensor instead?

As in the $7 Radio Shack units?

You can easily build your own from a CMOS hex inverter.

I've got a good schematic somewhere on my website, but I can't seem to find it.

That'll tell you if the circuit is energized, not carrying some level of current.

Oh crap... I see. I've never actually used a clamp meter before so I didn't realize that it worked like that. That makes sense.

Thanks!

-Me

So I should try placing that next to the cord instead of wrapping it around the cord?

-Me

I noticed that if I wrapped the wire around the powercord and then connected only one probe from the voltmeter, I would get a reading when the powercord was pluged in, but it didn't seem to matter how much current was being used. That must be the same type of principle.

-Me

Yes. Your 'winding' acts as a capacitor and couples the AC voltage from the cord into your voltmeter. Most DVMs have very high impedances and will pick up ambient E fields and indicate a volt or two.

But if you wrap a sense wire around one of the conductors, you still won't get a usable signal; there's no magnetic coupling in that direction.

A current transformer, even a clamp-on, has a ferrous core that both the current-carrying wire and the sensing wire pass through. The power wire makes one pass through the core, "one turn", and the sense wire usually makes many.

You could make a crude CT by winding a bunch of turns of fine wire through a steel washer and also running the current-carrying conductor through the center. It's not a very good core at 60 Hz, but you will get a signal that depends on current.

You can also place an air-core coil near a current-carrying conductor in the proper orientation and induce a small signal into it, proportional to current.

Google "current transformer" and "current transformer burden resistor"

John

To sense the current in a line cord, you have to SPLIT the cord so you can sense only ONE wire.

Or else go to some Hall Effect scheme where the sensitivity to ONE wire is MUCH GREATER than that of the other.

While we are talking about clamp on ammeters, lets not forget the DC clamp on ammeters from HP with a hall sensor in them. Way cool.

Hello John,

Golly, I've got to try that. That would have to be the cheapest CT there is. Should have kept the rusty washers from out old basketball hoop ;-)

Maybe there is a pre-calculated setup but if not that would probably require a serious tour through Maxwell's equations. Or just calibrate it if nobody is looking...

You could stack 1" washers 1/2" high. But I'd buy a cheap audio transformer that I could still run a wire through.

Hello Homer,

I used to do that with audio transformers. But they are neither cheap nor widely available anymore because new designs don't use any. Sometimes luck has it that there is a common mode choke with enough clearance to squeeze through.

Google "Rogowski coil". The idea here is to make an air-core toroidal coil that encompasses almost all of the flux around a wire. The output is the time derivative of the current. The good ones, the ones that do capture most of the flux, can be beautiful.

John

Hello John,

This is a nice write-up about them:

My very first lesson at my alma mater happened to be at their Rogowski Institute.

Found it.

Schematic of the Radio Shack e field sensor is in

Or you could just tap off the CT in a GFCI receptacle with a second regular receptacle providing connection to whatever appliance whose current draw you want to measure:

---------- --------- | | | | | | Hot-----0|Line Load|0---------0| o | | | | | | GFCI | | x | | | | | 0|Line Load|0 +----0| | | | | | | | o | ---------- | --------- | Neutral-------------------+

Mount both inside the same junction box and connect it up with a heavy duty line cord.

Since the GFCI is not connected to the neutral, it can't trip, and will always induce a current in the CT when the connected device draws current. You need to open the GFCI receptacle and solder on a couple of wires to bring the CT secondary to the outside.

Ed