Archived Forum

You are almost certainly reading noise picked up by the analogue input. This might be power line induced from adjacent wiring, or digital noise coming from the processor itself. See:

http://openenergymonitor.org/emon/buildingblocks/measurement-implication...
http://openenergymonitor.org/emon/emontx/accuracy
http://openenergymonitor.org/emon/node/1385
http://openenergymonitor.org/emon/node/2311

Trystan did some wonderful animated pages a few months ago demonstrating how the errors vary here http://vis.openenergymonitor.org/dev/adc/noiseless, here http://vis.openenergymonitor.org/dev/adc/noisy and here http://vis.openenergymonitor.org/dev/adc/noise.mp4.

Some users have said that connecting a 2.2 nF capacitor from the input pin to ground reduces the noise.

Which current transformer are you using - a calibration of 29.41 is wrong for the -000 version, and a 18 Ω burden resistor is wrong if you are using one of the "voltage output" versions (e.g. the -030) because those already have a built-in burden resistor.
 

[edited to add more links]

I’ve built a simple Arduino breadboard test rig to measure voltage and current and have noticed power and current readings when there is none.  I checked the analogue value for the CT and it was a constant 510.  I understand this is most likely noise and have read the various articles you site, but it strikes me that if the noise is constant? Then it could be removed in the calibration.

In photography when I take a photo with a long exposure, I first take one with the lens cap on and at same exposure as intended.  I then have a photo basically of noise which I can subtract digitally from my actual photo.  It works better than any gimmicky noise reduction on the camera.

So if you know there is no power then any measurement caused by noise could be removed.  Emonlib could be modified to include a calibration  constant i.e. y=bx+c.  Where b is the normal calibration and c is an offset.

I’m waiting for components to build a variable gain setup and imagine noise will change at each gain level, so it would be great if Emonlb coped with this.

Thks,

I calmly read what you posted,

As the sensor is Model TCS-013-000, will change the capacitor and resistor Burden.

No, you need a burden resistor for the SCT - 013 - 000. The 18 Ω  burden and a calibration of 111.1 will give you a current range of 0 - 100 A.

[Edit]

To elaborate:
SCT - 013 is a family of current transformers all having the same physical size, but with different maximum currents. All except the 100 A version ( SCT-013-000) have a built-in burden resistor and are classed as "Voltage output" types; the output is 1 V rms at the rated current. The 100 A (-000) version is classed as "Current output" and delivers 50 mA rms at the rated current; this version has protection diodes built in to prevent dangerous voltages appearing if the c.t. is unplugged but they will not prevent damage to the ADC input if the burden becomes open-circuit.