Values too low / 0?


I now have my emonPI and 3 nodes connected via 443mhz

They’re monitoring a collection of circuits in a DB.

The main incomer is on node 1 which I know has ~8kW on it, however it reads as ~360 (I assume the value is power, i.e watts)

Other two incomers (phases) are reading at ~4000 which suggests 80kW (looking at the ratio alone)

I know the board in total has ~100amps (23kW) so I’d expect 6-10kW per phase.

My nodes 2 and 3 are monitoring circuits from the board, when I put my trusty ~1kW fan heater on, it still reads 0…

When I tested even low (20w) loads where registered…

So what have I missed?

I am using just a single phase for the my AC voltage, nodes 1-3 show it at between 240-245v so about what I’d expect.

Also is there anyway to get the apparent amps from the CT’s before its manipulated by the voltage?

I don’t understand your setup. I need more details.
“Nodes” - do you mean emonTx? Is your system single or three phase, and what exactly are you measuring? Which sketch(es) are you using?


Yes so 1 x pi (not measuring anything yet) 3 x emonTXv3 (nodes) sending to it.

Its 3 phase and I’m measuring

Phase 1-3 on Node 1
3 Circuits from Phase 2 on Node 2
3 Circuits from Phase 3 on Node 3

All AC current with CTs’ all shop bought

Nb. Not sure what you mean by sketches… all ‘as received’ from the shop.

I tested with a emonTX on its own (via uart) and the values were watts so expected the same…?

So you’re measuring a 3-phase supply with a single phase sketch (the default).
That explains a lot. You need to get a programmer and load the 3-phase sketch into emonTx No.1. The reason is explained in the Resources > Building Blocks in the articles explaining 3-phase power, and in the comments at the top of the 3-phase sketch. A “sketch” is the Arduino name (don’t ask me why) for the user application software inside the processor in the emonTx.

Do emonTx’s No.s 2 & 3 each have their own a.c. adapter? It must be on the same phase as the one they’re measuring to get real power. If not, there’s a 120° or 240° phase error between voltage and current (or more properly the real component thereof).

What do you mean by that? If you mean the quadrature (imaginary) component of current - no. The real component is used (but not calculated and not available) to calculate real power. But if you mean rms (scalar) current, then yes, it is calculated and available in the sketch, but not sent by default. You’ll need to modify the sketch (software and programmer required) to send, and to modify the emonHub configuration to receive, the additional values.

Thanks - that makes sense.

Is there a way to measure different phases with the same emon?

I’ll try again:


I understood that, I’m just trying to improve my understanding so I can see how these can be used.

In that case, you’re not explaining yourself properly. There’s nothing in your profile - if you’d written something there, it might have given me some sort of a clue as to the extent of your knowledge and experience in electrical engineering. Remember, all I or anyone else knows about your installation and your circumstances is what’s written in this thread. None of us are mind readers.

Sorry Robert I thought I had!, I’ll try and be more verbose in future.

I’m a little confused about two things.

1). With 3 phase, if all the circuits I’m monitoring are on the same phase (so one emontx has 4 circuits, all on phase 2 and is powered by an AC adaptor on phase 2) the standard sketch should work (I think…?) it has no knowledge of the other phases…

As I understand it the emontx gathers the data, formats it and feeds it to the pi which it collates and graphs etc… OR does the phase the PI is on come into consideration? I’ve tried moving the pi to the same phase but the inputs are reading the same…

Or is the problem somehow related to magnetic interference (which would kind of explain what I’m seeing) from the other phases? (Pure speculative guess?!)

2). How do I upload a sketch, is this sent to the PI or the to the emonTX via serial?

I’m just trying to fill in the blanks on how this all works so I can get it working

Thanks again

Correct. The emonTx is designed for the UK domestic supply, which except for a few rare cases, is single phase.
The fudge the 3-phase sketch uses is to delay the voltage wave by 120° and 240° and pretend it’s the voltage for the 2nd & 3rd phases, then use that to get the real power for those two phases. Of course if the amplitude is wrong or there’s a phase/timing error, then you have an approximation only.

Yes and no. In your case, no. Data coming in is already processed - it comes in as watts. But the analogue front end of the emonPi is a cut-down emonTx, so when you’re using that, then yes, that part of it is single-phase, a.c. voltage and currents must be the same phase. (But the USB 5 V power doesn’t matter - it’s d.c!)

It could be, but unlikely to account for that much error. There are numbers in the test report for the CT.

For the emonTx, it’s Computer > USB > programmer > FTDI Interface on the emonTx.
For the emonPi, it’s more involved and spelled out in the Wiki (I believe, unless it’s moved).


So as long as I power my EmonTX’s using AC from the correct phase and I only monitor that phase with that EmonTX, I don’t need a new sketch as the EmonTX will collect the right data and feed it to the PI numerically regardless of phase.

I only have a problem if I want to monitor 3 different phases from one EmonTX, then I’ll need either one per phase OR the 3 phase sketch on JUST on that EmonTX.

If so, it explains most of the problems I’ve encountered…

Thanks again

As a rider, apparent power (VA) is also available in (but not sent from) the emonTx, and that might be of interest to you in a commercial setting. But I’m afraid we can’t do vars.