Firstly, again let me emphasise how much I appreciate the replies and help I’m getting. Thanks!
Brian: I guess it’s not a problem. It’s just not quite what I was expecting.
Looking at “Learn” on CT sensors, this page is basically saying because of the way that it, and the Sampling works on the Arduino, it’s accurate to about 30W.
https://learn.openenergymonitor.org/electricity-monitoring/ct-sensors/extending-ct-cable
This page mentions that an extension for CT cable might pick up noise. I don’t have an extended CT cable, but will see if I can re-route away from other cables.
It also mentions earthing the case of the emonpi. I may try that and see if it makes a difference. I checked with my cheap voltmeter and resistance meter, and couldn’t find any measurable voltage or resistance problems as discussed, however I did feel a small tingle at one stage as I touched something in one hand and an earth in another - the display corrupted and I had to cycle the power. I could not reproduce it, and as I said, I couldn’t see a measurable voltage between anything and earth.
Something else: If the power is to the emonpi, moving my finger across the case feels normal. If the power is on, moving my finger across the case, my finger seems to vibrate. I’ve noticed this effect before on a bedside-light that has metal case with a capacitive-touch-sensitive on/off switch. Is that normal?
https://learn.openenergymonitor.org/electricity-monitoring/ct-sensors/yhdc-sct006-ct-sensor-report
I’ve tried to read about the different sensors. I’m afraid the electronics characteristics went over my head a bit, but given the price of the alternative CT sensor, I may try it anyway despite the difference “being the equivalent of having an 11 bit ADC instead of a 10 bit” - well it’s a bit different. My solar will be around 10A max I think. 10A on the SCT006 seems to give 12.5mA approx, wheras on the SCT-013-000 it seems to give 5mA. This ties in to the comment that there’s about a 2.5 times advantage for the SCT006 over the other one, so I guess if I try it I’d have to divide the sensed figure by 2.5.