Two EmonPi Monitoring the Same Circuit - Different readings from one CT Sensor


We have two EMONPI’s running on the same network and monitoring the same circuit (Solar and Grid Feeds).

Both seem to be insync with the Solar feed (+/- 1% difference) but the main grid feed seems way off (+/- 60 - 80%+ difference)

I have tried swapping CT sensors - no difference.

The reference voltage is slightly different between the 2 (one is 236, the other is 232) so I wouldn’t have thought that is causing the difference in readings.

Any ideas? Could one be a faulty unit?

Many thanks!

That sounds likely. When you write “Feed” - are you referring to the electricity supply wiring or a Feed inside emonCMS?

What are the actual values, and is the difference apparently random, is it proportionately less the larger the absolute value being measured, or is it more-or less constant irrespective of the value measured?

What does the Inputs page of emonCMS report? Are the numbers you see reported by emonHub (Setup → Emonhub → View log) similar?
The way to decode the values is:

2022-05-04 15:39:59,932 DEBUG    RFM2Pi     20651 NEW FRAME : OK 5 47 0 41 0 88 0 84 94 0 0 0 0 0 0 0 0 0 0 0 0 156 108 17 0 0 0 0 0 71 83 17 0 42 40 18 0 (-0)
2022-05-04 15:39:59,936 DEBUG    RFM2Pi     20651 Timestamp : 1651675199.931935
2022-05-04 15:39:59,937 DEBUG    RFM2Pi     20651 From Node : 5
2022-05-04 15:39:59,938 DEBUG    RFM2Pi     20651    Values : [47, 41, 88, 241.48000000000002, 0, 0, 0, 0, 0, 0, 1141916, 0, 1135431, 1189930]

This is a snip from my log now. In the last line above,
47 is power1
41 is power2
88 is power1pluspower2
241.48000000000002 is Vrms (ignore beyond the 2nd decimal!)

By far the best test is to put all four c.t’s on the same cable, and then all four should read the same. If you can report the actual numbers, that would be helpful.

Does the c.t. plug going into the socket feel right?

That’s under 2%, so it explains almost nothing! And it affects both powers, not just one of them. You’re looking for something much bigger than that.

Many thanks for the reply Robert - I am going to put all 4 clips on one wire and confirm the readings that way.

Ok, so it looks like one of the CT’s is at fault. It’s occasionally reporting just 3 to 7 watts for about 5 - 10 seconds before reverting to the same as the other readings. I have swapped that CT to other inputs and the behaviour is the same. Occasional and intermittent.

Going to order up a replacement and see if that cures it.

Thanks again for your help

The fault follows the c.t. and does not stay with the same input? So it’s definitely the c.t. and not the input? That wasn’t clear from your initial post:

That’s much more useful than

Anything below 10 W could be noise and pickup - if you’re have a multimeter, check the resistance from the plug tip to sleeve. It should be around 100 Ω. And watch while wiggling
the cable. If it suddenly goes open-circuit, that’s the problem. If the fault appears to be at the c.t. end, which is the more likely, it’s possible to open up the c.t. - the core and bobbin is only held in by the two plastic clips you can see when you open it wide - but it needs 4 hands: one to hold the body, two to hold the clips back and one to pull the core out. Then check for a broken wire, and shorten if necessary. Be careful around the wire of the winding - it’s very fine and easily broken - and if it’s the inside end that breaks, there’s nothing you can do.

One other thing…

Ensure the faces of the ferrite material in the CT are clean and the surfaces mate with NO gaps.

A small air gap affects the phase error significantly, not so much the amplitude. It doesn’t seem to me as if that’s the problem.