Powering 3x emonTx with one power supply and one (common) voltage reference

A few questions here. I have trawled the community posts and wikis but can’t see any answers. Hope someone can help me before I try this for myself.

I have twelve circuits to measure and are using three emonTx v3 devices in an enclosure. All have a WiFI adapter plugged in on the programming header.

I have a 5V DC Power supply unit that I can use for powering the units with DC.

Can this PSU be connected to the block (using connector 1-5V and 3-GND) on each device rather than using a USB cable?

I also want to share the voltage reference so intend to use one 9V AC-AC supply and wire this in parallel to the three devices and plug into AI0 (9V input). Jumper JP2 is removed on all devices.

Aso want to check if the WiFi board can be connected directly to pinouts on circuit board as per EmonTx + WiFi Adapter purchase option, or does it require the Tx->Rx swap as per connection to programming header.

Yes, however there is no protection on those pins whereas there is on the micro usb.

Assuming the obvious, that you have a single phase installation and all emontx’s are therefore monitoring the same phase then, yes, but be very careful to observe the polarity of the AC signal, ie ensure all the emontx AC centre pins are common and likewise for the outer connector.

Yes the emonESP header is designed for the job on post v3.4.3 emontx’s (see emontx3/hardware at master · openenergymonitor/emontx3 · GitHub) the adapter is only need on the original header accessible from outside the case. (I believe! I do not have a v3.4.3 emonTx to confirm)

1 Like

I expect my power supply to be well behaved :wink:

Two of the emonTx devices are monitoring three phase circuits. One is the main incomer to the house. The other is an underfloor heating circuit. We plan to add the kW in a separate application. Will this involve changes to the sketch running on the Tx?

Ok then the answer isn’t so straight forward. Each emontx only has one AC input so it can only really accurately monitor a single phase (any phase, but just the one and the AC adapter must be on that same phase as all that emonTx’s CT’s).

So therefore you could/should have an AC adapter (sensor) on each of the phases you are monitoring and yes you can link more than one emontx to each AC adapter on a per phase arrangement.

Or, you can install 3phase sketches to approximate the voltage signal of the other phases, not so accurate but possibly easier to install. But yes that involves loading new FW sketches.

If you can access a socket on each phase being measured and can group the CT’s into groups of 4 without mixing phases then this is the way to go and involves no FW changes. Otherwise use the 3phase sketches and observe the fact there will be 2 of one phase and 1 each of 2nd and 3rd phase.

You can of course do mix of both schemes should you choose.

1 Like

I have to say that I’m not sure I know exactly what supply you have due to the word “spilt” in the title, that’s a very USofA term.

Maybe the title should read “… one (common) voltage reference”?

(And of course as you (Paul) point out, that’s not ideal on a 3-phase supply when all 3 phases are being measured.)

To clarify, we use “split supply” where a single transformer secondary winding is centre-tapped and connected to an earthed neutral, and two “legs” provide (in the case of N.America) 120 - 0 - 120 V, which can be used at 120 V for low power appliances and lighting, or 240 V for high-power fixed appliances.

To the best of my knowledge, Ireland has a very similar supply to the UK, i.e. TN-C-S
(Earthing system - Wikipedia, Section 2.1.1) operating at a nominal 230 V L–N, 50 Hz.

A 3-phase TN-C-S system is definitely not a “split supply”, but you can have a “split-load” consumer unit - that’s a totally different matter.

Yes Robert. I was talking about a ‘common’ voltage reference. It is an Irish 3-phase supply as you pointed out.

This isn’t going to be easy.

Unfortunately I am going in blind to this house as there is no single line schematic to tell us which phase goes where. Depending on how the local sockets are fed in the area where we plug in our AC-AC adapter we could be picking up any of the phases. This should be OK for the 3-phase measurements if we connect our CTs in the right order and rotation, but it will make measurement of other circuits difficult.

For example if my ‘common’ voltage reference is on L2 and the emonTx is measuring circuits from L1 we would need a sketch that deals with the 240deg rotation (2/3 cycle). We could end up with quite a few combinations all requiring a different adjustment :frowning:

I assume a phase rotation meter will tell us the difference between our reference and the target for each circuit we wish to measure?

I’m not sure that it would. Those are designed to indicate whether you have L1-L2-L3 or L1-L3-L2. I’d not expect to see a meaningful result for (say) L1-L2-L2.

The three-phase sketch’s installation notes tell you how to set it up to get the phase sequence correct. I suggest you download the sketch and the accompanying documentation
Update to 3-Phase PLL sketch - #21 by Robert.Wall
and have a careful read so that you can see better what’s going to be involved.

When you’ve identified the correct sequence at one point, a simple voltmeter will identify which phase any outlet or circuit is on - given a long enough test lead. Just measure the voltage between the unknown phase and the three known ones, when the answer is close to zero, that’s the same phase. When it’s close to 415 V, it isn’t. I’d expect to see a few volts when you’ve got the same phase. And remember to check that it’s live first.

From a point of view of personal safety, I’d expect all the outlets in one room, and certainly all those that are within reach of one another, to be on the same phase. If they’re not and there happen to be two faulty items, you stand to get a 415 V shock rather than a 240 V one. It’s something I’d check, and from what you write, you’ll be documenting it. Over time, I’ve documented everything like that in my house - it’s invaluable when you need to change or add something.

1 Like

Thanks for that link. It makes it a lot clearer how the 3-phase sketch works. I will still have issues measuring power on other single phase circuits where the voltage reference is from a different phase. If we group the circuits by phase as you suggest it will be a lot easier to set up each emonTx.

This is the enclosure ready to go to site. Quite a lot packed in :wink:

That’s true enough - and 10 A d.c. is rather more than you need - each emonTx takes only a few tens of milliamps. (1 A @ 5 V would have been plenty.)

It might be worth mentioning that the 4th input of the emonTx can be uprated by adding a wire-ended resistor in parallel with the existing SMT burden - that would give you additional flexibility. 27 Ω will return it to 100 A max using the SCT-013-000 c.t.

I realise this was over the top, but was also catering for the ESP8266 load of each unit. The WiFi boards are mounted on the internal header.

Good to know. I was wondering what size resistor would ‘desensitise’ this input.

Thanks again for all your help. Will probably have some further updates . . .

Or anything in between 27Ω and ∞.

It’s a 120 Ω SMT on the p.c.b, and the rule of thumb for the resistance is you’re looking for 1.1 V rms across it, given a well-behaved (= decent sine wave) load. (And the calibration coefficient is the current that gives you 1.0 V there.)

But having written that, it’s better to have a lower current c.t. (provided it’s got enough power to generate 1.1 V rms) or a multi-turn primary winding than to increase the burden value.

Indeed. But a very clean install, nonetheless. Nice!

1 Like