Hi All,
I just finished setting up an Arduino Uno with SCT-013-000 CT sensor following the instructions found on:
http://openenergymonitor.org/emon/node/58
I uploaded the `voltage_and_current' sketch after changing the
emon.voltage calibration number from 234.26 to 115. I'm using the
5V supply for the current sensing electronics with 33 Ohm burden
resistor. I clamped the CT sensor around the powerstrip cable below
my desk. A Kill-o-watt shows ~50 W power (~0.4 A) of draw. However,
instead I'm seeing the following on the serial output:
real power, apparent power, power factor, rms voltage and rms current
------------------------------------------------------------------------------------------------
7.57 37.21 279.49 0.13 0.20
4.55 27.91 279.54 0.10 0.16
-15.34 99.33 279.73 0.36 -0.15
8.77 63.31 279.58 0.23 0.14
2.98 26.76 279.51 0.10 0.11
6.44 30.41 279.59 0.11 0.21
-3.59 63.25 279.54 0.23 -0.06
11.47 55.79 279.43 0.20 0.21
4.11 24.76 279.45 0.09 0.17
3.93 24.65 279.55 0.09 0.16
4.20 26.79 279.65 0.10 0.16
3.28 22.40 279.76 0.08 0.15
FWIW, I'm using a Linksys brand power supply (Mdl. AD9/1C). It
measures 11.34 VAC unloaded. I suspect that in this case the numbers
don't reflect some error in either the current or voltage calibration.
They appear to be all over the place. Do you all know what my
problem might be? Should I only be clamping the hot or neutral conductor
but not both?
Thanks in advance!
-mt
PS. Here's a pic of my HW:
http://www.flickr.com/photos/63654603@N04/7218485894/in/photostream
Re: Bad Data from New Build (Newbie in need!)
Exactly. The CT picks up the magnetic field generated by the current flowing in the cable. If you pass both line and neutral through the core, the currents are equal and opposite and the fields cancel each other out. You should be measuring on only the line, or only the neutral, conductor.
You are using a 33 Ohm burden, so you probably need to calculate your own calibration constants. The theory and how to derive the numbers for both voltage and current is here: http://openenergymonitor.org/emon/buildingblocks/ct-and-ac-power-adaptor...
The random numbers you are seeing are most likely due to interference pick-up. The ADC is not accurate below about 1% of full scale current, and the slightest pickup can give you a few tens of watts error.
(Out of interest, are you using a plugboard as shown here http://openenergymonitor.org/emon/node/58 or the emonTx kit with pcb?)
Re: Bad Data from New Build (Newbie in need!)
Hi Robert,
I'm doing what you call the `plugboard' (British English for `breadboard'?) on the URL you mention.
Both the current and the voltage figures are wrong in my data. I believe a correction factor of ~116 is correct for US `Mains' power, but I'll take a look again at what I calculated when I get back to work and see if I've made a mistake.
Thanks!
-mt
Re: Bad Data from New Build (Newbie in need!)
I've just spent a day analysing the measurements from an emonTx kit pcb, and what I found makes me almost certain that the random results you are seeing are due to pick-up.
I wrote a special sketch to save and then dump consecutive raw values to the serial output. Using a "Bimboard" like that illustrated to hang a potentiometer and a few resistors on the input of an emonTx, the numbers spread over 4 adjacent counts. When I put the emonTx inside an earthed metal box, dispensed with the Bimboard and wired direct; the spread dropped to a maximum of two counts (and if I 'pulled' the bias voltage so that it lay almost exactly in the middle of a step of the analogue to digital converter, almost all the measurements were exactly the same number).
So you may find that screening everything improves things for you (but if you are transmitting the data by radio, I see an obvious problem with that idea...).
For what it's worth, I call the prototype board illustrated "plugboard". I call it "breadboard" when you have pins that you insert into a perforated insulated board, and then solder the components between pins. As they say, two nations divided by a common language ;-)