I have just installed EmonTx V3 inside a 230V 3 phase power panel. (shown in the below).
The connection and firmware are correct except that when I turned off those 3 Circuit Breaker switches (On -> OFF), the 3 CT's reading still shows value of 460W (230v x 2A) for each.
I understand if there is no AC current inside the clamped CT sensor, there still will be readings of 30W or so but 480W is too high!
Why is it? Is it possible other adjacent "live" wire generated inducted current in the CT as well? Anyone has successfully installed CTs inside panel by hooking with paralled "live" wires?
Note:
CT1 is red phase, CT2-yellow phase, CT3-blue phase. When CB switches are on, the readings increased to 600W which are correct, because those 3 circuits consume about 200W each.
Re: EmonTx V3 Installed in Panel report wrong power when circuit breaker switched off
It is possible that the CT core will pick up a magnetic field from an adjacent conductor, but I'm surprised that it is as much as that. I will attempt to replicate your set-up and report my findings.
[More...]
I set up my CT test rig and tried to get as close to your set-up as I could. The maximum current that I can have is 5 A, so I have a 20-turn coil that in effect gives me 100 A. The coil is approx 100 mm diameter, and with the CT inside the coil but with the secondary winding of the CT lying directly on the bundle of wires forming coil, the highest reading I could get was 241 W (equivalent to approx 1 A). Unplugging the AC adapter to measure current, I read 200 VA - 0.87 A maximum. Positioning was critical, moving the CT away by 10 mm halved the power/current indicated. The most sensitive position was (looking at your picture) when the coil bundle was between breakers 3 & 4.
From your picture, it seems highly unlikely that you have approx 200 A flowing in the cables that the CT is touching, so I think this may only be part of the problem.
Incidentally, with no current in the coil, I read about 20 VA (no ac adapter) and less than 2 W with the adapter.
Re: EmonTx V3 Installed in Panel report wrong power when circuit breaker switched off
Hi Rober Wall, My CT turn is 2000:1 (100A:50mA), not 20 turn coil.
Today I unplugged all the CTs and EmonTx, put them aside for 5 hours and install them back to the panel again.
Now those CTs show correct readings of 530W each (I am having 10 lighting x 52W each) and when I turned off those lights, the readings show 20W each which are correct.
I didn't do anything and this leads me to think maybe it is the CT connection problem with the EmonTx stereo 3.5mm power socket?
Next day when I arrive at office and I turned off those lights again but this time the "off" state power readings show 160W. It is not 20W anymore.
Above oberservations make me wonder if the CT sensor itself has some "memory" effect such that it will remember previous current which has something to do with magnet core or any material made those coils?
Re: EmonTx V3 Installed in Panel report wrong power when circuit breaker switched off
"Hi Rober Wall, My CT turn is 2000:1 (100A:50mA), not 20 turn coil." You do not understand. I was simulating a current of 100 A in a wire touching the body of the CT, and I could get only half of the reading you say you saw. To get the reading you had, I would need approx. 200 A.
Iron-cored CTs can be permanently magnetised, but yours have a ferrite core, and I do not believe it is possible to permanently magnetise that. Even if you did, and although it would affect the calibration, I cannot understand how that would give you the "memory" effect you describe.
Did you turn the lights off using the circuit breaker - when you might have moved a CT connection - or did you use a switch somewhere else? Can you check that the CT plugs are inserted all the way in to the sockets? A new socket can be very stiff and it is easy to not push the plug all the way in.
How and where are you reading the values? If you connect a programmer to the emonTx and use the serial port on the Arduino IDE, what does that tell you? If the powers are correct there, there is nothing wrong with the CT.