3 Phase - Identifying the Phases

I want to monitor the ouput from a 3 phase PV array using an emonTx with the 3 phase firmware loaded.

The 3 x live cables from the inverter to the smartgen meter are neither labelled nor colour coded.

I understand that CT1,2 & 3 must be connected in phase sequence. How to do that?
I guess by trial & error …
What input readings will I see when I’ve got it wrong compared to when I’ve got it right?

Also does it matter to which phase the ac/ac adapter is connected?


As the notes in the 3-phase sketch state, the a.c. adapter must be on Phase 1. But which is phase 1 doesn’t matter, only the phase sequence is important. You could:

  1. Use an oscilloscope to determine the order.
  2. Connect a 3-phase motor and see which way it rotates.
  3. Borrow, rent or build a phase rotation indicator/meter.
  4. Check the meter manual. Although unlikely, it may be that it will indicate incorrect phase rotation, in which case the terminals will be labelled in the correct sequence.
  5. Trial and error. If you calibrate phase 1 current and phase calibration on a known resistive load, and the phase 2 & 3 c.t’s on the phase 1 cable for the correct current and phase calibration for exactly half power (and negative), then with the same load on phases 2 & 3, they are right when they give the right real power.
    I see an “indicator” for under £10 on Amazon - but can’t comment on its safety nor efficacy. Also, searching for “3 phase rotation tester circuit diagram”, there are many circuits, the simplest is only two neons, two resistors and a capacitor. But be aware that you’re dealing with up to 440 V between phases, and contact with that could easily prove fatal.

What ring mains circuits are in the vicinity? Are the ring main clearly labelled at the DB?

With the use of an extension lead if needed I would be inclined to try powering the emonTx’s AC adapter from a “known phase” and then identify the PV cable that appears to be producing the most power (due to being in phase with the AC adapter supply). As long as you can reach at least 2 ring circuits on different phases (with a extn lead to another room if needed) you should be able to establish and label the 3x PV cables. From there you are in the same position as any “labelled” 3ph PV installation and can probably assume L1, L2, L3 as identified in the main DB as the correct rotation.

Rather crude by it should work ok, I’m sure @Robert.Wall will correct me if that not the case.

You would be better doing this while the std emonTX firmware is still in place so all the CT’s are referencing the same phase and a direct comparison of the 3 PV circuits can be done rather than swaping the 1st CT of a 3ph emonTx from cable to cable.

I was assuming that there are no phase markings anywhere on the system. If there are, then it’s a lot simpler: measure the voltage between the unknown phase and the three known ones (or two known ones should be enough) - when you measure 400 - 440 V, it is not that one. When you measure a very low (< 20 V, and most likely a lot less than that, but a small voltage), you’ve got the same phase.

My method using the 3-phase sketch maybe wasn’t too clear. Assuming there are no phase markings anywhere, you start off on the downstream side of where the PV joins, probably upstream of the consumer unit, but somewhere where you have access to the three phase line cables separately.

  1. Calibrate the voltage. Any socket is good for this.
  2. Put all 3 c.t’s on the Phase 1 cable - the one with the a.c. adapter on it (by definition it is phase 1 for our purposes) and calibrate the current so that all three read the same correct current. Any substantial load will do for this.
  3. Take all other loads off the system. Put a pure resistive load on phase 1 and set PHASECAL for input 1 to give a real power equal to the apparent power, and a power factor of 1.0
  4. Set PHASECAL for inputs 2 & 3 to give a real power equal to half the apparent power and negative, and a power factor of -0.5

That should have set up and calibrated the emonTx.

With all loads off the system still:
Put a pure resistive load on phase 1 and note the power.
Transfer the c.t’s for inputs 2 & 3 to the other two phases.
Transfer the load to one of the other two phases. If either input 2 or input 3 gives the correct power, you have the c.t’s on the right cables. If not, swap the c.t’s 2 & 3, when one of them should give the right power. If it does, you have identified the correct phase rotation, so label it. If it doesn’t, something has gone badly wrong.

Then you can go back to the inverter output and identify the phases there by measuring the voltages between there and a place where you know which phase is which.

no, I wasn’t sure either, it not clear even if there are any “other electrics” in the vicinity.

I wasn’t sure there would be any exposed contacts and/or Johns willingness/ability to probe 400v+, so with safety in mind I opted for the extn lead method.

Believe it or not, on at least 2 occasions, when a qualified electrician has been provided for me by the client (as I am not a qualified sparks), when a question has arisen over identifying a phase and I have suggested the “the one that’s not over 400v” route, they have just looked at me as if I had grown horns, so despite other sparkies making sense of this test, I assume it is not something that’s included in the standard qualification training or just theory that gets forgotten.

Don’t talk to me about qualified sparkies. Someone I know moved into a “new” house - actually a very old one heavily modernised, and in the corner of the hall is a staircase that goes up in two flights. Next to the staircase is a chimney breast, so there’s an alcove behind the staircase (not sure if I’m making sense, it’s hard to describe). There’s a consumer unit in there. To get to it, you go into a low cupboard under the stairs, head first on your back, sit up, then stand up in a space about 2 feet square, and the distance between your nose and the consumer unit is about 10 inches. And it’s been signed off, so it meets the rules for “accessible”? No way in my book is it accessible even to reset an MCB, let alone safe to work on.

Then again, if it’s only the PV output that’s being monitored, then it’s probably a safe bet that the power factor is close to 1.0, and modifying a standard sketch to report apparent power rather than real power won’t be too far out, and in that case, phase sequence is immaterial.

Good point!

Guys …

Many thx for all yr ideas and suggestions.

Safety is a concern so I do not want to remove covers and probe 400+ volts.

The ground mount PV array (114 panels) & inverter is 100 metres from the house. Between the array and the house is an outdoor heated swimming pool with an ASHP in situ but not yet plumbed or wired. Extension leads are not really feasible.

There will be 3 x emonTx’s - at the inverter, in the pool plant room and at the grid supply meter (house). The site has wifi coverage.

When I open the cabinet at the ground mount PV array, I’m faced with 3 unidentified cables from the inverter to the smartgen meter.
Here will be an emonTx (with 3 phase firmware) serially connected to an RPi running emoncms plus 3 CT’s and a pulse counter on the smartgen meter.
There is also a 13 amp socket outlet (for the ac/ac adapter) connected to one phase but I know not which. See picture …

From an earlier forum exchange with Robert, I understand that at the inverter/smartgen meter, the PF will be close to 1.
Also I assume the power on each phase will be close to equal when generating – correct?
The inverter itself has an LCD panel showing what is currently being generated in total.
Once up & running, I envision applying a simple correction factor (based on smartgen meter pulses) to the phase powers as shown by the emonTx.

So my trial & error questions – if I clip all 3 CT’s to one live cable – what emonTx powers will be indicated if the selected live is not nominal phase 1 (ac/ac adapter phase)?
What will I see if by chance the selected live is nominal phase 1? Is this a way of identifying nominal phase 1?

If so I’d then be left with 2 x unidentified phase live cables and 2 x CT’s. I could try hooking both CT’s to each of these phase lives in turn – what powers would I see? If one were negative would this be the final clue?

To complicate matters, the installation is at my son’s place an hour’s drive away and I’m spending days trying to set up emon gear in my single phase home with no PV where I’ve made an experiment with an emonTx running 3 phase firmware.

I clipped 2 CT’s to my one live phase (making sure they were in same direction orientation). One CT was in CT1 input and the other in CT2. The house consumption was 500 watts approx but varying as indicated by CT1 PowerL1. Then I boiled a kettle …

CT1 PowerL1 3465 watts
CT2 PowerL2 minus 1880 watts (at the same sampling instant)

I switched the second CT into CT3 input, rebooted the emonTx and boiled another kettle …

CT1 PowerL1 3678 watts
CT3 PowerL3 1606 watts (at the same sampling instant)

Is this data in any way helpful or relevant?

Whoever wired that cabinet didn’t believe in securing the cables, did they? That’s pathetic.

And it’s not a good idea to leave the c.t’s on their cables and without a burden. Although they are supposed to be safe, I’d rather not risk it. If there’s any current likely to flow, pop them off the cables before you unplug them. Short-circuiting a c.t. is always safe, open-circuiting one might not be.

I think I’ve actually covered all the points you raise in post 4.

You can set up all the emonTx’s at home and calibrate them. (Bear in mind that calibration corrects for sensitivity and component variations in both the emonTx and the sensor, so once calibrated, keep the same sensor (voltage & current) with the same input (label them all!). This is steps 1 - 4 in post 4 above.

Then go to the other house and install the emonTx at the inverter.

The only snag I can see here is, Phase 1 at the inverter might not be the same as the phase you’re calling Phase 1 at the house, and Phase 1 at the pool might be any of the three. Without isolating the whole system and putting shorts to neutral or earth and then testing with a meter, which even I’d be very careful doing so don’t try it, I don’t know a definitive way around that. If you’ve been really careful calibrating the voltages, then with the system off-load, you might just be able to infer which are on the same phase because they will record the same voltage, but even so, all that might prove is two phases are at the same voltage! Certainly, if the voltages are different, it’s a fair bet that you are looking at different phases. If you’re drawing any current anywhere, all bets are off.

That will probably be correct, though I understand that some inverters will attempt to ‘clean up’ the mains so if one phase is low, it may inject more current into that in an attempt to restore balance.

It says you haven’t calibrated it yet!
If CT1 was correct, CT2 should have read -1732 W and CT3 should have read -1839 W.
The orientation of the c.t. on the cable is important - all 3 should face the same way.

Robert …

It’s not all that bad …

Today was our best generation day since installation in Dec 2016 – for the first time broke the 200 kWh hurdle - 214 kWh generation today.
The panels & inverter seem to be performing well – the fact that the inverter rating of 27.6 kW is regularly breached (during a 15 mins period) is encouraging as is the extended almost flat top to the generation bell curve on a sunny day even as early as March.

My calcs are that once the pool ASHP is running, the family will reduce its carbon footprint BY 80% - not a bad legacy for my 4 grandsons.

My objectives are:

  1. A My Solar display on the kitchen counter to encourage good behaviour

  2. A monthly Excel report showing how the investment (and family behaviour) is performing in terms of kWh’s, carbon and importantly £’s return on the investment

  3. Smart control of the pool ASHP by managing the PV system to supply electricity to satisfy the household electricity demand and then operating the ASHP when export level exceeds a threshold level.

1 & 2 will only be in terms of totals across the 3 phases so I’m assuming the snag you identified – ‘The only snag I can see here is, Phase 1 at the inverter might not be the same as the phase you’re calling Phase 1 at the house, and Phase 1 at the pool might be any of the three.’ – will not distort the total of the 3 phases in any significant way – am I correct?

With respect, could you provide guidance on my trial & error questions - if I clip all 3 CT’s to one phase live cable – what emonTx powers will be indicated if the selected live is not nominal phase 1 (ac/ac adapter phase)?
What will I see if by chance the selected live is nominal phase 1? Is this a way of identifying nominal phase 1?

Again with respect, perhaps I did not clearly describe my kettle boiling experiments at my single phase home running an emonTx with 3 phase firmware.

Only 2 CT’s were involved, directionally they were correct and I just switched one CT from CT2 input to CT3 input for the 2 experiments.
Whilst the 3,000 watt kettle was a resistive load, the background circa 500 watts was probably not – LED lighting, TV’s on standby, sundry PC’s, tablets, etc on charge.
So I did not expect the CT2/3 power to be exactly 50% of CT1.

However I find it difficult to understand the power switching from negative to positive (experiment 1 to 2) is entirely explained by component tolerances and the need for calibration. With the second CT on the CT3 input, the power was positive not negative.

Pls clarify – what am I missing?

Yes - the total across the 3 phases will be right - you problem is you just (probably) don’t know which phase is which.[quote=“johnbanks, post:10, topic:3928”]
With respect, could you provide guidance on my trial & error questions - if I clip all 3 CT’s to one phase live cable – what emonTx powers will be indicated if the selected live is not nominal phase 1 (ac/ac adapter phase)?
What will I see if by chance the selected live is nominal phase 1? Is this a way of identifying nominal phase 1?

Have a look at the phasor diagram in the page in the Learn section about 3-phase power. What the 3-phase sketch does is use the voltage wave to calculate the power for its own phase, and delay it by two different amounts so that it lines up with the other two phases. When you have a resistive load, voltage phasors (those rotating arrows) line up with their respective current phasors (not drawn). If you get a c.t on the wrong phase, the current aligns with one of the other phasors, not its own. Because the voltages are 120° apart, the projection of the voltage that it actually lines up with onto its own phasor is half the length and on the opposite side of the centre. So instead of the power being V × I, it is -½V × I.
Summary: On the correct phase, you read the correct power. On the wrong phase, you read -½ the correct power.

(And setting it up on a single phase supply, that’s exactly what you adjust everything to give.)

OK, that’s understandable given that you’d got an overall non-unity power factor.[quote=“johnbanks, post:10, topic:3928”]
With the second CT on the CT3 input, the power was positive not negative.

It was probably a long way out - the phasor only needs to swing by 30° to go from negative to positive, and if the timing is wrong and your load is partly reactive, that’s easily achieved.