I have an EmonPi with a CT connected to one cable on my domestic supply between the meter and the distribution board. I also have a reference power source connected to the mains and an optical pulse reader on the meter.
I’m finding that the values from the CT are significantly over-stated. Is there something I need to do to calibrate the reading? There isn’t a whole lot of wiggle room in the placement or alignment of the CT due to the positioning of the cables.
Over the last 24 hours the meter visual reading increased by 13kWh. The meter doesn’t seem to report any decimal places, so the granularity available is only 1kWh.
The optical pulse counter reported 12,596 pulses. This equates to ~12.6kWh and is close enough to the difference in the meter reading given the granularity available.
I have another set of pulse counters on 9 individual mini-meters on each circuit around the house (which account for all the power fed through the utility meter) and it came back with 12.4kWh based on the pulses from each of the meters added together. For my purposes this is close enough to the 12.6 and 13kWh readings from the other two sources.
Meantime the CT reported 19kWh. That is ~50% higher than actual based on the utility meter. Does anyone have any ideas what might be accounting for this difference, and whether there’s anything I can do to correct it? I can live with the pulse readings, but wouldn’t mind having the CT chime in as well so that I could notice if/when something went wrong with the pulse meters.
There are full calibration instructions for the emonTx in the ‘Learn’ section here. Those equally apply to the emonPi, but unfortunately a version of the front-end software that would enable you to do the full calibration is still waiting for release - it requires changes to emonHub.
However, unless you have either the wrong calibration constant for your c.t or your v.t. (aka the a.c. adapter), or both, there’s something seriously amiss somewhere. Although the worst-case values can get alarmingly high, in normal circumstances you should see readings well inside ±10%, and hopefully inside ±5% even before you calibrate. Ideally, you need to check the outputs of c.t. and v.t. to establish whether either or both are wrong, or it’s that input or both inputs of your emonPi (or a combination of all of these).
What I suggest is, in the absence of a “real” faulty component or components, all you can do is adjust the scales = ... value in emonhub.conf for Node 5 (the emonPi itself) for the input in question. The standard value is 1, if you change it to 0.667, the values should come much closer.
Thanks for the response. I had a look in the Learn section starting here - Learn | OpenEnergyMonitor - but couldn’t find anything directly about calibration. There was a link to this page - Learn | OpenEnergyMonitor - but that returns a “page not found”.
I’m going to let the device run for a couple of days/week and will work out the value to use for scales in emonhub.conf based on the observed ratio between the figures, but I wouldn’t mind giving the other calibration method a go.
Thanks for the fix I’m copying the URL (CTRL-C) then pasting it into the message (CTRL-V). In my edit window it shows up as the text of the URL, which is what I want, but when it is displayed after posting the URL is immediately replaced by the descriptive text.