Hi,
I am running a setup where I am measuring fridge energy consumption, with 2 setups in parallel.
One measurement is with emonTX V3 and Raspberry PI.
The other is with emonTx Shield V2 and NanodeRF SMT.
I performed calibration with a resistor bank as explained at http://openenergymonitor.org/emon/buildingblocks/calibration .
The values (Vcal, Ical, phase_shift) are as follows:
- (241.1, 88.6, 2.7) for emonTX V3
- (239, 58.740, 1.6) for emonTx Shield V2
As you can see the phase_shift value for emonTX V3 is far beyond the recommended range (0.0 – 2.0).
This causes a problem when changing the load to the fridge.
The V_RMS, I_RMS and Apparent Power values match, however the Power Factor and Real Power values are not consistent.
I believe that it has to do with the emonTX V3 phase_shift value being outside the recommended range, as the same EmonLib functions are used for calculations of both setups.
What confuses me is that having performed calibration for the emonTX V3 at a different power socket of the same room previously, all the values were within normal range (241.1, 88.6, 1.5).
Please help me understand what is causing this strange emonTX V3 performance
Thanks
Re: EmonTX PhaseShift Value Outside Range
Phasecal 'corrects' for the combined effect of 3 things, phase shift in the two transformers (voltage & current) and the time delay between readings of voltage & current. If the processors are the same, that more or less rules out a difference in the time delay. You can easily see whether the VTs are to blame by swapping them. Then we come back to the CT. My first concern is that you might have calibrated and be using the two systems at a very small current - in the order of a few percent of the operating range.
The two have different value burden resistors, so that will automatically mean that even the same CT in the two units will have a different phase shift. The other fly in the ointment is the CT is, if I'm right, working at a point where the phase shift is increasing very rapidly, which will amplify the effect that the different burden resistors generates.
That probably explains, but does not help, your problem. I'd need to know the actual measurements that you are seeing. It might be that using more than one primary turn for the CT might help (and of course changing the calibration to suit).
Re: EmonTX PhaseShift Value Outside Range
Thanks for your advice Robert !
The resistor bank I am using has 3 X 47 Ω in series, giving a current of ~1.56 A RMS.
I first switched the VT, then the CT.
The CT was effectively to blame, as it gave a strange phase_shift calibration constant on the emonTx Shield V2 as well.
Replaced it and now the setup works fine.
Re: EmonTX PhaseShift Value Outside Range
I assume you've seen the spread of phase shift values for different CTs (same model) and different currents and burden resistors - graphs are in the report in Building Blocks.
Do the CT cores meet properly without any air gap or rocking? It needs almost nothing to introduce a huge phase shift - too much to bother measuring!
Re: EmonTX PhaseShift Value Outside Range
Dear Robert,
Thank for the additional advice and sorry for my late reply.
When you refer to the CT cores, you mean the 2 pairs of magnets that are in contact when closing the CT.
If that so, indeed one of the magnets is loose and adding paper to better position it indeed improves the readings.
Re: EmonTX PhaseShift Value Outside Range
We are referring to the same thing, but they are not 'magnets'. The 'core' is a ferro-magnetic material, sometimes soft iron, sometimes electrical steel, but in this case two moulded ferrite 'U' shape parts that fit together when the assembly is clipped around a cable. But they are not magnetised, and indeed must never become magnetised. They must fit together perfectly so that the magnetic flux created by the primary current gets linked to the secondary winding, which is wound around the core in the larger plastic part, and which you can see when the CT is opened up. The slightest air gap ruins the performance, as you've found. If you look carefully, the 'top' half of the plastic moulding has two plastic 'springs' that are supposed to press the cores together. I presume these are not doing their job.
The YHDC part is modestly priced, this is the first time I have heard of a manufacturing fault, so considering the number that must be in use, I think you were just unlucky.