Current clamp sensor on the end of a long wire.

[Simon]

Is it ok to have a current clamp on the end of a long wire back to the measurement circuit?

[Gyro]

It depends what sort of current clamp sensor... and how long you're talking about. Not many clues there.

If it's a traditional current transformer then it ought to be ok. I think it would be important to put the sense resistor close to the transformer, both to limit the voltage burden and to minimise the risk of it becoming accidentally disconnected (= voltage breakdown!). Obviously you would want to use a twisted pair to minmise noise pickup.

[Simon]

Well I've recently had a "powervault" installed and notice that there is a long wire from the current clamp to the main unit, I don't know if there is a burden resistor in there. Would screening help? the installer was told the wire had to be 100mm from any mains cable which is stupid as it logically sits in the same conduit that carries the mains power to/from the unit

[james_s]

A product that I worked on had a fairly long cable between a current transformer and the unit that read it. We had some concern during the design process that there would be problems with interference but testing revealed this to be a non-issue even using a single ended input. Burden resistor in my case was placed within the control unit and even coiling up the 30' wire with a mains extension cord carrying around 1kW to various test loads had no measurable effect.

[Simon]

OK sounds like it might be ok then. I'm told by the installer (who is no fan due to issues on every install including mine) that no one has had problems.

[Gyro]

No, it doesn't sound like a problem in that case. Screening would help with high frequency common mode pickup but twisted pair (which I assume it is) ought to provide good cancellation at 50Hz.

[Simon]

I have no idea if it is twisted, that is my concern

[james_s]

Well, does it exhibit problems? If it works then I wouldn't worry about it. I try to have some faith in the fact that most engineers who design commercial products have at least some knowledge of what they're doing.

If in doubt, fire up some electrically noisy stuff like universal motors, radio transmitter or a Tesla coil near by and see if it malfunctions.

[Simon]

Well, does it exhibit problems? If it works then I wouldn't worry about it. I try to have some faith in the fact that most engineers who design commercial products have at least some knowledge of what they're doing.

hm, well no it's not working very well actually and i have no faith in them at all. Apparently this works fine on other installations.

[Simon]

They are using this one: https://www.ebay.co.uk/itm/100A-SCT-013-000-Non-invasive-AC-current-sensor-Split-Core-Current-Transformer-/271409507165 available on ebay for under 5 quid

[SeanB]

Well, it is a 50mA shunt, so probably best run with a shielded cable, easiest would be some Cat Whatever shielded network cable, as you really are only worried about the shield, or you can just use regular unshielded network cable with the unused 3 pairs shorted at the one end only. As it is also a floating system with no direct mains connection there is no issue with isolation of the cable, though that also depends on the instrument, so check if the current input is isolated from mains using a transformer of some sort ( probably not, likely the power meter simply references neutral and uses a capacitive dropper as power supply) and if not then treat the cable as needing to be isolated to mains specs. Most likely is it is mains referenced ( because it needs a voltage reference for power) so just use unscreened twisted pair, and the 2kV isolation of the inner and outer insulation, along with it being double wall as well, should be enough. Plug should have a 3A fuse in it as well in any case in the UK, and be connected to a RCD as well.

[Simon]

whoa, whoa, back up, this is not my shit. They will be fixing it if it's an issue.

see strange thing is that this sensor is all over the place for using with an arduino...... AND the main controller they have used is a ATmega2560....... I'm really starting to suspect this pile of junk is a copy of an arduino mega2560 with any extra circuitry they need.

[Simon]

Thinking about it how does one determine the direction of current with a current clamp transformer ? I can see how it could achieved with a burden resistor measuring the voltage of the live relative to neutral so that you know the part of the cycle your in and then the polarity of the measurement across the burden resistor but on a totally floating current transformer how do you know which way power is going?

[fcb]

Thinking about it how does one determine the direction of current with a current clamp transformer ? I can see how it could achieved with a burden resistor measuring the voltage of the live relative to neutral so that you know the part of the cycle your in and then the polarity of the measurement across the burden resistor but on a totally floating current transformer how do you know which way power is going?

You don't.  As you said, you have to compare it against a reference.

[Simon]

I suppose you have to still look at the current polarity of the voltage with reference to neutral and then you can tell from the polarity of the output from the current transformer at the same time if the current is going one way or the other.

What I am seeing is power going the wrong way, but up to 150W but then the resolution on the input is 1024, the sensor is 100A range, that is 141.4A peak, 141.4/512=0.276A resolution (and I'm not going to go into the accuracy of the ADC argument) 0.276*240=66.3W. I understand that current sensors are not brilliant at near 0 amps ? that is exactly where this thing has to work.......

[Simon]

So say I needed a very large measurement range, would it make sense to switch different burden resistors in? so in the case of this sesnor 33R for the 0-100A but say 330R for 0-10A and even 3.3K for 0-1A ?

[fcb]

So say I needed a very large measurement range, would it make sense to switch different burden resistors in? so in the case of this sesnor 33R for the 0-100A but say 330R for 0-10A and even 3.3K for 0-1A ?

Theoretically that would work, in-practice you want to make sure that you don't o/c the circuit. Also, you might find that cable positioning in the centre of the coil becomes important. But with only a 100:1 range at mains frequencies I think i'd amp up the 0-100A, something like cascading a pair of x10 opamps and then tapping off the at the three positions.

[Simon]

Yes amplifiers would be another way, some sort of zener or TVS would be essential in any case, the transformer in question does have a TVS built in.