KWh meter - is there such a thing as a non-contact one?

[Merlysys]

Its to monitor energy use of a subpanel(upto 60A at 120V). I could buy a KWh meter and hook in between main power source (120V) and subpanel.

But it would be easier if it just has a current clamp so I could wrap it around a wire. Such a device should have a voltage set for 120V but even if it only displays amount of current flowed (in coulombs or whatever unit) in a given period I could convert that to KWh.

If nobody makes such a thing what suggestions you have?

[Zero999]

The short answer is no.

Why? It's not possible to accurately measure voltage, without a physical connection to the cable. Capacitive coupling could theoretically be used, but it's only accurate if the area of the conductor and distance are exact and the dielectric constant of the insulator is precisely known.

The only way would be to fix the voltage probe to the cable, calibrate it against a meter connected directly to the cable and ensure the voltage probe and cable are never moved afterwards, otherwise it will need to be recalibrated. Drift would still be an issue, since the dielectric constant of the cable will change due to changes in moisture and atmospheric humidity, which will limit long term accuracy.

[rstofer]

Amps will only get you kVA (with an assumed voltage), you need to account for power factor to get to kW and the only way you are going to do that is to measure the phase angle between the voltage and current.

[soldar]

Yes, you need to measure voltage. Why can't you connect to the wires to measure voltage?

[Someone]

Its to monitor energy use of a subpanel(upto 60A at 120V). I could buy a KWh meter and hook in between main power source (120V) and subpanel.

But it would be easier if it just has a current clamp so I could wrap it around a wire. Such a device should have a voltage set for 120V but even if it only displays amount of current flowed (in coulombs or whatever unit) in a given period I could convert that to KWh.

If nobody makes such a thing what suggestions you have?

If you don't mind the inaccuracy from missing the imaginary power they do exist:
https://au.rs-online.com/web/p/digital-power-meters/8724666/
Plenty of similar cheap clamp meters around, unlike the naysayers above sometimes an estimate of the power consumption is fine.

[Gregg]

E-Mon D-Mon is what you are looking for http://www.testequipmentdepot.com/emon-dmon/
They do need a connection to voltage but most are available with split CTs so that the poser doesn't have to be interrupted to connect them.  Most of them are suitable for billing of power usage. 

[DaJMasta]

Yes, you need to measure voltage. Why can't you connect to the wires to measure voltage?

This.  You get your current information (and if you want it accurate get a current probe/transformer/clamp with some bandwidth), then you have a meter to monitor the voltage.  The voltage measurement isn't inline, so it doesn't need a high current capacity, you just need to tap off the conductors and measure with any normal meter rated for mains work.

[floobydust]

You can get clamp-on current transformers with high-voltage isolation rating, to measure current.
To correct for power-factor, you need to sample voltage as well. A voltage monitoring circuit, must be hard-wired and include a small fuse. Never cheap out there.

The Sensus Smart-meter fires in Saskatchewan (recall of 105,000 smart meters), large part of the problem was the fires caused by kWh meter not being able to withstand Overvoltage Category IV transients.

[NiHaoMike]

It could be possible to get the voltage reference with a plug in device, just silly when you could just tap off one of the breakers in the box.

I had the idea to make a whole house wattmeter that modulated a FM carrier on the power line in order to get around the problem of the metal box effectively shielding wireless signals.

[ejeffrey]

Its to monitor energy use of a subpanel(upto 60A at 120V). I could buy a KWh meter and hook in between main power source (120V) and subpanel.

But it would be easier if it just has a current clamp so I could wrap it around a wire. Such a device should have a voltage set for 120V but even if it only displays amount of current flowed (in coulombs or whatever unit) in a given period I could convert that to KWh.

If nobody makes such a thing what suggestions you have?

It's not clear what you are asking.  If you are asking for a power meter that _only_ uses a current clamp, then no that doesn't exist for the reasons described.

If you have only seen kill-a-watt style inline power meters and you are asking for a power meter that does not require interrupting the feed to install, then yes that exists quite easily.  They use a current clamp to measure the current, and a regular voltage probe to measure the voltage. 

Here are a couple that a quick search turned up:

http://www.extech.com/display/?id=14175
https://www.hioki.com/en/products/detail/?product_key=6417
https://www.omega.com/pptst/HHM98P.html

Note, I don't know anything about these particular products, the just appear to be clamp-on power meters with external voltage probes.

In principle you could make a power meter that used capacitive sensing to pick up the phase of the voltage waveform and then assumed it had a specified amplitude.  But I don't really see the point, it isn't hard to measure voltage.

[floobydust]

The utility industry uses capacitive voltage dividers. I have seen them on 240kV power lines to monitor voltage, instead of potential transformers (PT). It's much lower cost.
But at 120VAC, it's easiest to use a resistor divider.

There are single phase DIN rail mount energy meters for under $30 on eBay. Probably burn down your house unless it uses pass-through CT's, fused voltage sampling.

[EEVblog]

Not commercially that I am aware of, apart from the clamp based ones, but I know someone working on a multi vector magnetometer based design for remote non-contact switch board current measurement.
Resolution isn't great and requires calibration, but it works. Can't measure voltage of course.

[soldar]

You really need to measure (connect) voltage. Voltage variation, wave shape, reactive loads will all affect the measurement. You cannot just multiply current by an assumed voltage.

[EEVblog]

You really need to measure (connect) voltage. Voltage variation, wave shape, reactive loads will all affect the measurement. You cannot just multiply current by an assumed voltage.

Well, you can, it's just a matter of how much error you are willing to accept.
Our mains voltage for example doesn't vary by a huge amount, so multiple a fixed value by the RMS current and you are in the ballpark.

[Zero999]

You really need to measure (connect) voltage. Voltage variation, wave shape, reactive loads will all affect the measurement. You cannot just multiply current by an assumed voltage.

Well, you can, it's just a matter of how much error you are willing to accept.
Our mains voltage for example doesn't vary by a huge amount, so multiple a fixed value by the RMS current and you are in the ballpark.

But that will only give you apparent power, not true power. If you could obtain the phase of the sine wave by capacitive coupling, then you stand more of a chance.

[The Soulman]

You really need to measure (connect) voltage. Voltage variation, wave shape, reactive loads will all affect the measurement. You cannot just multiply current by an assumed voltage.

Well, you can, it's just a matter of how much error you are willing to accept.
Our mains voltage for example doesn't vary by a huge amount, so multiple a fixed value by the RMS current and you are in the ballpark.

But that will only give you apparent power, not true power. If you could obtain the phase of the sine wave by capacitive coupling, then you stand more of a chance.

With capacitive coupling you may stand a chance but the phase shift introduced by the coupling may only be calibrated out if it is the coupling is 100% repeatable.
Which it's not. 

[Zero999]

You really need to measure (connect) voltage. Voltage variation, wave shape, reactive loads will all affect the measurement. You cannot just multiply current by an assumed voltage.

Well, you can, it's just a matter of how much error you are willing to accept.
Our mains voltage for example doesn't vary by a huge amount, so multiple a fixed value by the RMS current and you are in the ballpark.

But that will only give you apparent power, not true power. If you could obtain the phase of the sine wave by capacitive coupling, then you stand more of a chance.

With capacitive coupling you may stand a chance but the phase shift introduced by the coupling may only be calibrated out if it is the coupling is 100% repeatable.
Which it's not.

Providing the resistive element dominates, then the error will be minimal. For example, if the meter's input impedance is 100M and the capacitance to the mains conductor is 100pF (31M8, at 50Hz) then the phase shit will be minimal.

Shielding and filtering to remove high frequency interference will be more of a challenge, than phase shift.

[The Soulman]

You really need to measure (connect) voltage. Voltage variation, wave shape, reactive loads will all affect the measurement. You cannot just multiply current by an assumed voltage.

Well, you can, it's just a matter of how much error you are willing to accept.
Our mains voltage for example doesn't vary by a huge amount, so multiple a fixed value by the RMS current and you are in the ballpark.

But that will only give you apparent power, not true power. If you could obtain the phase of the sine wave by capacitive coupling, then you stand more of a chance.

With capacitive coupling you may stand a chance but the phase shift introduced by the coupling may only be calibrated out if it is the coupling is 100% repeatable.
Which it's not.

Providing the resistive element dominates, then the error will be minimal. For example, if the meter's input impedance is 100M and the capacitance to the mains conductor is 100pF (31M8, at 50Hz) then the phase shit will be minimal.

Shielding and filtering to remove high frequency interference will be more of a challenge, than phase shift.

You really need to measure (connect) voltage. Voltage variation, wave shape, reactive loads will all affect the measurement. You cannot just multiply current by an assumed voltage.

Well, you can, it's just a matter of how much error you are willing to accept.
Our mains voltage for example doesn't vary by a huge amount, so multiple a fixed value by the RMS current and you are in the ballpark.

But that will only give you apparent power, not true power. If you could obtain the phase of the sine wave by capacitive coupling, then you stand more of a chance.

With capacitive coupling you may stand a chance but the phase shift introduced by the coupling may only be calibrated out if it is the coupling is 100% repeatable.
Which it's not.

Providing the resistive element dominates, then the error will be minimal. For example, if the meter's input impedance is 100M and the capacitance to the mains conductor is 100pF (31M8, at 50Hz) then the phase shit will be minimal.

Shielding and filtering to remove high frequency interference will be more of a challenge, than phase shift.

 

ot: congrats on your 11999 post! 

I think 100pf is to ambitious, according to my "back on the envelope" calculations with a probe length limit of 50mm (for practical reasons) and dual insulation (required by law) and 2,5mm2 cable the coupled capacitance is max. 10pf.
And then there are repeat-ability issues because of physical probe placement, different wire types and insulation material and thicknesses.
In a controlled (lab) environment something like this maybe doable with usable accuracy (say below 5% error) but
the probe must be calibrated in place and not be touched after.

Sounds like a nice topic for a video to me, including some experiments etc. 

[Zero999]

I think 100pf is to ambitious, according to my "back on the envelope" calculations with a probe length limit of 50mm (for practical reasons) and dual insulation (required by law) and 2,5mm2 cable the coupled capacitance is max. 10pf.
And then there are repeat-ability issues because of physical probe placement, different wire types and insulation material and thicknesses.
In a controlled (lab) environment something like this maybe doable with usable accuracy (say below 5% error) but
the probe must be calibrated in place and not be touched after.

Sounds like a nice topic for a video to me, including some experiments etc. 

How did you come up with 10pF? Perhaps 100pF is a bit high, but I was thinking about a foil or length of wire wrapped around the conductor.

The input impedance of the meter could also be much higher than 100M: 1G is doable, but the higher the impedance, the more noise pick-up will become an issue.

I agree, it would be an interesting experiment to determine if this is workable.

[soldar]

Many years ago I built probes, similar to multimeter probes, but with short tips made with sewing pins or needles. This will penetrate plastic insulation and leave it intact when pulled out. I find this very useful for finding where a cable is broken, identifying one cable out of several, checking for voltage, etc. This would work in this case. If I wanted to make it a bit more sophisticated I could include a fuse.