Archived Forum

Tell me more:

What voltage is your Arduino running at?
What voltage are you developing across the CT's burden resistor?
What value is the burden resistor?
What ratio current transformer are you using?

It might well be that the change in power level is too small to be measured.

I'm running my arduino at 5V

The CT is a CT sensor SCT-013-030 (Max 30A) so no Burden resistor is needed.

AC/AC - 9V

What is the change in power that should be noticed? If I'm consuming 300W on my house and producing 150W it should be noticed?

Thanks

Hi Rechena,

If you're running the standard sketch for Voltage and Current, then I would expect you to detect a change of 130W without any difficulty.  The benefit of this approach is that you get to measure real power, by summing the instantaneous product of voltage and current over the measurement period.

If, instead, you are just measuring current, and assuming a nominal value of voltage to calculate the power that is flowing, then your values are likely to be much less accurate.  With no actual current flowing through the CT's primary, my test rig recorded a background current of anything up to half an amp.  But when voltage was measured too, the calculated power value dropped to virtually nil (no more than a couple of Watts).  The benefit of measuring voltage as well as current to obtain power has been well aired on this forum in the last month or two.

I have three old-style bulbs on which the power labelling is difficult to read.  Plugging in each one into my test rig in turn, there was never any doubt as to which one was which.  The power that they took was always close to either 40W, 60W, or 100W, with minimal error. 

I'm simply using the standard V&I sketch, calibrated first to give the expected value for AC voltage, and secondly to give the expected values for the current taken by various known loads.  Calculating the expected values for VCAL and ICAL is always a good idea, but this approach can't correct for component values which are slightly in error, or components whose parameters are not known.  For example, the current clamp that I'm using clearly has a large number of turns in its secondary coil, but I do not know how many.  However, if I double the current that the clamp is measuring, then the reading that's taken by the Arduino is also doubled, so my system appears to be linear over the range that I'm interested in, which is basically up to 13A. 

My suggestion would be to fully check your system using a test rig before trying to obtain any real values for your PV.  My rig, since dismantled, can be seen at 

http://www.youtube.com/watch?v=w0OpfWBmYz8&feature=plcp

At that time, I was only measuring currents but the system worked well enough for demo purposes. 

Hy Calypso_ray, I'm using the normal sketch 

And it's really supposed to reduce on the total W? Meaning if realpower = 200W and PV power = 100W if I turn the PV off it should be 300W realPower?

Thanks once again..

I've made some tests on a isolated rig, but with fans and lights, and I could actually see the values chaging, same has my house, for instance. If I turn the lights on the living room the W go up a couple of W's. But it never gets down or up with the solar PV production... 

I'm measuring the solar PV production with a Chinese Kill a Watt and with a attopilot connected to an arduino. Do i need to add another CT to the Solar PV output and then make some math? 

Thanks, and sorry for all the questions...

Where have you installed the CT?

If you are monitoring current on the meter tails between the meter and the distribution board, then yes, you should read the nett imported power. And if you are generating more than you are consuming, the power should be negative indicating that you are exporting.

You could of course add another CT to measure generation, then you could display the 3 quantities that you are interested in: generated, consumed and nett power. Which maths you do depends on where the CTs are installed.

Here, we have:

Meter(street) -> Amp Selector(company only) -> Differential Breaker -> HouseHold Breaker's...

I've placed the CT after the Differential Breaker.. is it ok? 

I'm using a modified version of the standard sketch, but nothing of significance has been changed.  My sketch was attached to a message that I posted on 28th May at  http://openenergymonitor.org/emon/node/176?page=4

For my test rig, I removed the outer insulation from part of an extension cable so that a CT would fit around just one of the cores.  Different loads could then be plugged in so as to provide known currents for the Arduino to measure.  This provided an easy way of checking and calibrating the system.  When it was all moved out into the garage, no further adjustments were needed.

I'e been looking at your code and the core part is just like mine... except on the calibration part.. Still not sure what my problem could be :(

And where does the PV feed in on this list?

Meter(street) -> Amp Selector(company only) -> Differential Breaker -> CT -> HouseHold Breaker's...

-> CT -> HouseHold Breaker's -> PV

"Meaning if realpower = 200W and PV power = 100W if I turn the PV off it should be 300W realPower?"

You are correct, you should see the current/power fall when you generate, and you should see the power increase when you switch off the PV. Calypso_rae is correct, the 30 A CT should read 100 W without difficulty (from my measurements on the A-D converter, I think it should be accurate to within 10% at 75 W, and it is more accurate at higher powers - because more bits of the A-D converter are used).

Does the power you read (no PV) agree with the actual load - to within (say) 10%? What do you read for Apparent Power - it should be only a little more than Real Power?

Yep, that's what's making me sesing that something is wrong.. If I turn the PV off the power dosen't increase...

This is the current output with no sun, since it's night :P

realPower;apparentePower

660.53;934.71

When I turn the PV off the value that should change is supposed to be the real or the apparent? or both?

Thanks

Sorry about the blank page, it seems to be a problem with ckeditor which I cant get to the bottom of appart from that it look like this problem:  http://cksource.com/forums/viewtopic.php?t=25524, I have reinstated fckeditor in the mean time which seems to work. 

 No Problem :) Thanks :)

 In the meantime, I've changed the CT from the Line to the Neutral, and it is now working PERFECTLY :)

Meaning if:

RealPower = 300W 

Solar PV = 50W

With PV down, realPower = 350W :)

Is this supposed to happen?

Thanks once again for all the help.

Rechena

That is not supposed to happen!  If you have a single phase supply, you should get exactly the same result whether you measure on the line or the neutral.

I have to consider whether there is a problem with your supply - the conductor marked line really is line and neutral really is neutral?

However, the end result appears to be good - your EmonTx looks to be working correctly.

Meaning if:

RealPower = 300W 

Solar PV = 50W

With PV down, realPower = 350W :)

Is this supposed to happen?

Yes, that looks correct. And if you switch off the appliances when you are generating, you should see -50 W - meaning 50 W is being returned to the grid.

 My thoughts exactly, but acctually on the line was not measuring. Not sure what it is... Could be a grid tie inverter problem?

I don't want to cause unnecessary worry, but I think you need to check all your wiring very carefully.  It seems that there might be a second path for the current somewhere. If I'm right, it could mean that the whole installation is in a dangerous state and needs attention.

As a quick check, have you measured the line and neutral voltages with respect to earth? I don't know which country you are in so I don't know what the standard supply is, but the neutral should be within a few volts of earth and the line voltage should be about 240 V as you measured earlier.

(The worst case could be something like the problem my neighbour had: A plumber fitted a new electric shower and in the process of adding the extra wiring, he exchanged line and neutral for the whole house. Thus all the switches and fuses were in the neutral conductor and if something was switched off or if a fuse blew, the circuit remained live! The only fuse that remained to protect the house was the 80 A fuse belonging to the electricity company, and the only way to make anything and everything not live was to pull that fuse out. Fortunately, I found the error before anyone got hurt and before a fault set the house on fire).

 Ok, I've made some more tests and with the main breaker off the V on the voltmeter is 0V but on the arduino is 17W realPower... wierd isn't it...

I've tried the house plugs, and no current is passing... This is the image of the connection..

The blue wire is the Neutral... everyting looks fine...

" Ok, I've made some more tests and with the main breaker off the V on the voltmeter is 0V but on the arduino is 17W realPower... wierd isn't it..."

That is nothing to worry about. We think it is due to digital noise pickup from the processor combined with where exactly the mid-point voltage sits in relation to the ADC decision levels. See here http://openenergymonitor.org/emon/buildingblocks/measurement-implication...

On the photo, where did you put the CT when it was on the line cable? Because there are two rows of breakers, you did put it on the incoming side of the main breaker (marked "Line in")?  If you put it on the outgoing side ("Line feeding top busbar only"), you only measured part of the current because the line connection to the breakers on the same row as the main breaker is a busbar ("Busbar 1") that is out of sight.  But because all the neutrals are connected together at the top and then one neutral cable comes down to the main breaker, where you have it now you have the full neutral currrent passing through the CT.

Ok, you lost me there,

I had the ct connected to the Line out, after that big breaker there, but, it is supposed to power all the breakers. not?

I had the CT connected on the Black cable over there, near the blue one that I'm now using.

Thanks

"I had the ct connected to the Line out, after that big breaker there, but, it is supposed to power all the breakers. not?"

Not correct. The feed from the street goes in to the top of the main breaker on the brown & blue wires. The switched feed out to the bottom row of breakers is a copper bar not visible in the picture. The black wire from the bottom of the main breaker feeds out to the top row of breakers.The blue neutral from the bottom of the main breaker connects to the row of neutral terminals at the top of the box.

Therefore, if your CT was on the black wire, you were only measuring the current taken by circuits connected to the top row.

Now, your CT is on the blue wire and it is measuring all the current returning whichever row of breakers fed the circuit.

And I think your PV was connected to the bottom row.

Robert, you are a genious :)

I complettly forgot about the bus bar 

And yes the PV was on the bottom breakers. I will try it again, on the brown wire. Thanks once again for the pattience and help :)

RW: A plumber fitted a new electric shower and in the process of adding the extra wiring, he exchanged line and neutral for the whole house. Thus all the switches and fuses were in the neutral conductor and if something was switched off or if a fuse blew, the circuit remained live! The only fuse that remained to protect the house was the 80 A fuse belonging to the electricity company, and the only way to make anything and everything not live was to pull that fuse out. Fortunately, I found the error before anyone got hurt and before a fault set the house on fire).

I recently needed to fix some lights at a house with very old wiring.  It turns out that there are fuses in each of the Live and Neutral lines.  The fuse that had blown was in one of the Neutral lines, so although no lights were coming on, all of that wiring was still live. 

The moral of this tale is always to make sure that any exposed wiring really is non-live before touching it.  If there is any doubt, check it first with a neon-equipped screwdriver, or preferably a multi-meter.

"check it first with a neon-equipped screwdriver, or preferably a multi-meter. "

I always use a neon, and I always make sure that is working by checking that the circuit is live before I switch off / pull fuses, and dead after I've done so. A neon will light up on body capacitance to earth and generally does not require an earth connection, whereas a multimeter will do.

I claim mine was worse than yours: you at least had a relatively small fuse in the line side that would have still protected against a low impedance fault.

Did your "fix" involve a new consumer unit - it ought to have done. I can't remember when double-pole fuses were outlawed, it was certainly a very long time ago - probably before WW2 (and I wasn't born then - I date from the 1940's, but only just!).

The house in question was built around 1930, and the lighting circuits could well be original from that period, so it's well due for a re-wire!