3-phase power consumption measurement

[BartButenaers]

Hi @smcgann99,
I will explain my questions a bit more in detail.

@smcgann99

Not sure what you mean about green terminal block ? This device has screw conectors built in.

I had forgotten that @edstobi had already shared above which terminal blocks he had used (see here). What I meant is that the original terminal blocks on the pzem-004t devices look not very firm to me, but I can be mistaken. I mean wiring for 230VAC should be a bit more solid and safe compared to our low voltage signals that we are used to deal with. When I see for example the setup of this discussion, the terminal blocks are simply too small for me compared to the wires attached to it:

Something like doesn't look very stable to me, so that is not ok for my home electricity setup...
And I certainly don't want to have multiple wires into a single screw of the terminal block (like in the above image), but more about that below...

@BartButenaers

I am also not sure how thick the wires should be (both for the current and voltage inputs)

This is a bit related to the quality of the green terminal blocks. Normally all wiring for 230VAC signals is very rigid due to the tick copper being used. So then you need very decent terminal blocks to attach those wires.

But when I see for example this discussion, the wires to the CT clamps are very thin. And the wiring to the pzem-004t voltage pins are bit thicker, but still thin compared to the normal 230VAC wiring:

So not sure how thin the voltage wires are allowed to be?
And also not sure how long the wires to the CT clamps can be made, without loosing quality of the signal? Because it would be nice if I could add all those pzem-t004 devices into a small separate electricity closet for safety.

@BartButenaers

a series of pzem-004t v3 devices

I would like to measure N of such pzem-t004 devices via a single wt32-eth0 device. So I was not sure how to connect those together serial. I found in this youtube video the following diagram:

Not sure if this setup is correct? And does anybody knows how many pzem-t004 devices could be attached together this way, or how I could increase this number of devices?

@BartButenaers

And also not sure how to connect N of these devices to a single phase

I mean how the wiring can be done in a safe and solid way. Suppose I want to measure 3 separate circuits on a single phase, then the voltage pins of each of those pzem-t004 devices needs to be attached to the same phase:

On a drawing this looks simple, but how do I do this kind of wiring in practice? I don't want to attach two wires into a single screw of the terminal blocks, because that is not stable enough for me:

For my 230VAC electricity closet I have nice isolated rails all over the place, like e.g. these ones:

Such isolated rails are very nice to interconnect N devices. But have never seen something similar that I could use to interconnect e.g. pzem-t004 devices in a stable and secure way...

[Paul-Reed]

So if you are using just 1 phase for the voltage measurements, how are you going to shift the phase for the other 2?
It's not just the voltage level it's measuring, but also it's relationship to the current (from the clamps) to determine the phase angle.

[BartButenaers]

@Paul-Reed

So if you are using just 1 phase for the voltage measurements

No no, that is not what I meant.
Of course I will measure circuits on every of the 3 phases, and then all the pzem-t004 devices of a single phase needs to be interconnected again. I had mentioned a single phase, because otherwise my drawing became more complex:

So if some kind of isolated rail (or something alike) would exist, I woud need two of those for every phase.

[Paul-Reed]

Ah Ok, you are measuring 3 circuits, each being on 3 phase.
It's a typo, but there is a short between L1 & L2 which needs removing.

[BartButenaers]

@Paul-Reed

Ah Ok, you are measuring 3 circuits, each being on 3 phase

Yes, something like this:

I have only showed the measurements of devices on phase L3, but of course on the other phases the same happens. But then my drawing becomes unreadable, and I am running out of time for tonight

P.S. I have removed the short circuit you mention. Was indeed just a line drawed to quickly...

[Colin]

Do you really need to measure that many circuits?

[Buckskin]

Bart; the wire to the PZEM devices only needs to be low current, 2A 2 core would be more than adequate. As for joining all the wires - use a 'chocky block' which can be secured in something like this
or even simpler something like this

I don't know what your incoming mains setup is like but mine, for instance, has all connections completely secured in a consumer unit. The only wires visible are the incoming supply. I think you should consult an electrician about the regs in your locality, especially about having three phase supplies in one accessible box. (remember there are over 400v between phases) This is probably the reason a lot of 3 phase measuring devices only use 1 phase for the supply. I must admit, I find the thought of three phases to a domestic dwelling for anything other than a three phase machine a bit disquieting.

[BartButenaers]

@Colin

Do you really need to measure that many circuits?

Hey Colin,
Well I am not going to share the electric drawings of my house in public, but yes too many circuits to start buying Shelly devices.

@Buckskin

2A 2 core would be more than adequate

Hi @Buckskin,
Is this about the wires? If so, do they need to be able to carry so much current? And can you be a bit more specific about the diameter of the wires if possible?

@Buckskin

use a 'chocky block'

I have used already quite a lot of similar looking blocks in the past, but I don't know the correct English name for the ones I have used. So perhaps it were no 'chocky blocks'. Are those blocks interconnected, or do I need to interconnect them myselve using an isolated rail? But yes that could be a way to go...

While searching for your blocks, I came across another solution:

@Buckskin

mine, for instance, has all connections completely secured in a consumer unit. The only wires visible are the incoming supply.

Oh yes, I don't want this kind of stuff anywhere else than in a closed secure customer box. And even if the box is opened, I want to have all components very solid, isolated and protected.

Moreover devices - like the Shelly 3 EM I had planned to buy - also requires all phases to be wired to it, to have accurate power measurements (as discussed throughout the above discussion). So those devices would require a rather similar installation procedure...

I have already more than 20 years 3 phases in my house, and never had problems with it. All is well wired and protected. And I know a lot of people that have it at home. But don't want to start discussing about that, because then we will get way off topic. And I am not going to switch after all those years to single phase anyway

[Buckskin]

Your solution is even better.

Is this about the wires? If so, do they need to be able to carry so much current? And can you be a bit more specific about the diameter of the wires if possible?

It was about the wires. Cable - 0.5mm² 2 Core Pvc Flex White (3A). As far as current capacity, probably not but you do need the protection of the sheath (covering). Bell & speaker wire for example is only for low voltages.

[smcgann99]

@BartButenaers

And does anybody knows how many pzem-t004 devices could be attached together this way, or how I could increase this number of devices?

I can only speak for Tasmota firmware, which supports up to 3 on 1 ESP, I have only needed to connect 1 but I found this -

Multidrop Wiring

To avoid clashes on response, some diodes need to be added when more than one are used, the circuit also needs to be pulled high and below is a wiring diagram. A fast/low voltage drop Schottky diode is really needed to make the comms work successfully. Remember to cross over RX/TX from the Wemos to the PZEM modules.

@BartButenaers

And also not sure how long the wires to the CT clamps can be made

I have extend mine around 50cm without any issues.

@BartButenaers

So not sure how thin the voltage wires are allowed to be?

As long as the cable is mains rated (which is more to do with the insulation) then the smaller the better.
If you cut the ends to the correct length and tin them with some solder there is no need for any bare wire to be exposed at the green terminals.
You could take 1 larger wire to your separate box then split it there using the connectors you found.

[Colin]

@BartButenaers

do they need to be able to carry so much current?

The wires to the CT are low voltage, you can use anything. I use telephone wire for mine.

[Paul-Reed]

@Colin

The wires to the CT are low voltage

Only if the CT is connected to a shunt.
If the CT does not have an internal shunt, or is not connected to the shunt in the module, then the induced voltages can be quite high, and can give you a surprising jolt (I can speak from experience!!).
That can occur when for example the CT is added to a current carrying cable before plugging it into the device.
But no need for high current wires for the Ct's, almost anything will do, provided it has good quality insulation, and as Colin has said even telephone wire will do, but if you are not hardwiring the CT's, and using a plug/socket at the end of each CT instead, make sure it is insulated, and does not have a metal body which is connected to one of the wires

See the section;
Current Transformer Example No1 in
Current Transformer Basics and the Current Transformer

[craigcurtin]

@BartButenaers - use Wago connectors - quite cheap - and used extensively in the US - i have used them for a few of my things and they produce a good solid connection.

As the other guys have said - you are not moving many amps across the cables so you do not need very thick wire - but the insulation is key - so must be rated to mains.

Not sure how much you value your time (but it would be better spent writing new nodes for all us) - just get an IOTAWATT !!

Craig

[BartButenaers]

@Paul-Reed

That can occur when for example the CT is added to a current carrying cable before plugging it into the device.

Yes indeed. They mention everywhere that it is best to power of the electricity of the main cables, before attaching the clamps to them. Or that the clamps should be wired at least to the pzem-t004 before you attach the clamps to the main cables. Will keep attention to it.

@craigcurtin

use Wago connectors

That is indeed another option. Didn't know they had larger models with more poles. I see now that they have e.g. 8 way connectors:

@craigcurtin

just get an IOTAWATT !!

Yeah I have to admit that those devices look very neat, and will be easy to install. But due to the setup of my house (a renovated farm of my grand parents), I have a number of electricity closets which I would like to monitor. And then all these solutions are just way too expensive, if you take into account all the stuff you need (clamps, ...). Still need to negotiate with my wife about much other iot devices

[craigcurtin]

@BartButenaers

That is indeed another option. Didn't know they had larger models with more poles. I see now that they have e.g. 8 way connectors:

Yep they are pretty nifty and make a good connection - i just use mine and then put them into a sealed electrical box

I reckon by the time you mess around and add all the bits and pieces you need you will end up over the cost of incrementally adding another IOTAWATT for each of the cabinets in the house.

Thats how i started mine - an IOTAWATT on my main meterboard where my solar, batteries and mains all feed into

Then added a 2nd one in one of my distribution closets for the main circuits i was interested in - then when money permitted added a 3rd one and divided the ciruits down further - i have a fourth one of the way now

Craig

[BartButenaers]

For completeness, here is another one that I came across accidentally during my search:

It is a din-rail power meter, and they offer both wifi and zigbee versions.
I like such kind of devices because you can create a very neat/solid/secure setup:

• You just need to click it on a din rail
• They also have open-close CT clamps
• Via the N/L pins on at the top, these devices can very easily be interconnected and powered via the isolated rails I have been talking about above. So no voltage wiring needed.
• It allows you to build a very extensible system

But they are again more expensive (26,7 eur) and they are Tuya, which means they require a cloud connection? No idea whether you can connect with it locally, e.g. via node-red-contrib-tuya-local (which is not maintained anymore).

So I don't think this is the way to go...

[Paul-Reed]

@craigcurtin

I reckon by the time you mess around and add all the bits and pieces you need you will end up over the cost of incrementally adding another IOTAWATT for each of the cabinets in the house

Are you on commission from Bob Lemaire, Craig

[Buckskin]

@BartButenaers

But they are again more expensive (26,7 eur) and they are Tuya, which means they require a cloud connection? No idea whether you can connect with it locally, e.g. via node-red-contrib-tuya-local (which is not maintained anymore).

The nodes and API are still going strong for existing kit, but that could change for new equipment.

[BartButenaers]

Yes I was reading a bit a Tasmoto on Tuya last hour, and the ATMS1603 (from my last screenshot above) is not in the list of devices supported by Tasmota. Moreover seems their latest firmware makes it very difficult to flash alternative firmware. And on top of I read here that Tuya is also switching their chipsets:

Tuya has begun manufacturing some Wi-Fi modules using a Realtek RTL8710BN Wi-Fi SOC instead of an ESP82xx chip.
Tasmota cannot run on Realtek devices and there are no plans on supporting them .

So I am not going down that road...

[BartButenaers]

Hey guys,
I need to have a bit more knowledge, before I start buying my hardware. Want to avoid having an (extra) large box of unused material in my basement, because I should have bought other devices...

I am not quite sure if I understand the phase shift correctly...
The phase shift is depending on the load (resistive, capacitive, inductive) and so it differs from phase to phase. However is it the same on all circuits of a single phase? I assume not, because there are a series of sub-circuits on such a phase each with different kind of loads. So I assume that each sub-circuit has its own phase shift?

Let's explain my question with a nice picture I copied from here:

When you buy pzem-t004 devices, you need to wire those so they can measure both voltage and current. So they can measure very accurately the power on every point in your setup. But that is hell lot of extra wiring. Which is not very attractive in an electricity closet that has already been certified officially...

The devices like Iotawatt work differently: you need to wire the 3 phases only once to it, and then you need to specify for each CT clamp to which phase this clamp belongs. And then they know the phase shift and they can calculate the power. But is that correct? Suppose I install two CT clamps for two different sub-circuits of the "same" phase:

• One clamp on a sub-circuit with a capacitive load
• One clamp on a sub-circuit with an inductive load

I assume the "total" phase shift is measured for each phase by those devices, since you need to add a voltage wire to the start point of each phase entering your house? Something like this I mean:

Am I correct that these readings are less accurate compared to e.g. a pzem-t004, but they are still accurate enough for a normal household?

Reason for my question is this. Suppose I measure the phase shift per fase, by wiring 1 shelly device to each phase. And then I measure for all my circuits only the current, via one of the devices mentioned above:

Would it be possible to calculate the power for each circuit in Node-RED, based on the voltages and phase shifts and current measurement? Or does does phase shifts change to quickly, so you cannot join the related messages good enough in Node-RED to make good calculations perhaps, or some other pratical issues I didn't think about?

Because this would be an affordable solution with not much wiring involved...

Thanks again for supporting my old brain!!!!