The phase shift is only in the current. The phase of the voltage for each phase will effectively be the same throughout the house. The CTs measure the phase of the current for whatever they are measuring. Therefore you can take the voltage reference for each device from anywhere, provided it is the right phase.
1 Like
Isn't that the same as PZEM's?
Imagine if you put all of the PZEM's into one box, and then connected together the three Phase A voltage inputs, the three Phase B voltage inputs and the three Phase C voltage inputs on the PZEM's together (all neat & tidy inside the box).
Then, just like the itoawatt, you would only need the 3 phases feeding into the box once (because each of the 3 phases would be shared amongst each of the 3 PZEM's inside the box.)
Your wiring dilemma occurs, I assume, because you are wanting to put each PZEM at a different location, which would then require voltage wiring being provided at each location.
As far as I'm aware, they would be the same, and accurate.
What could introduce inaccuracy is if you were using a iotawatt and used derived voltage reference mode (which I discussed earlier), which of course you cannot do with individual PZEM's anyway.
You would not be able to sample the data often enough from the Shelly, and also node-RED couldn't carry out the calculations quick enough, so no, not really.
Inside the PZEM & Shelly, the phase angle would be monitored almost continuously, many times a millisecond, and the output that you would get from the devices would be either a mean average of the phase angle between readings, or a snapshot at that point in time. Neither tells an accurate story...
1 Like
Ok Colin, thanks again! Makes sense!
Yes you are absolutely right. I was thinking at each pzem device separately this morning, but that is of course not the case. It seems like I was not awake enough this morning...
Ok that sounds very logical.
No objections anymore for pzem-t004...
A series of pzem-t004 v3 devices with 100A open-close CT clamps have been ordered.
[EDIT] For people interested in this solution: it seems that the white connector on the pzem-t004 is a male JST-XH 2.54 connector, so I have also ordered some cables with female connectors (see here).
THANKS FOR ALL THE FEEDBACK AND PATIENCE !!!!
Within reason though Colin - for instance - i have a small sub board down in my backyard that does all of my back yard devices, pool pump, lighting etc etc
This is approx 60m (in cable distance) from the main board and has 4mm cable to it - we still voltage differences of approx 8v when the pool pump is running and my solar is outputting at the main board
I have a number of other circuits in my house that have runs of 50+ metres when you tae into account all the snaking around between floors etc.
Barts house could be similar - but as a rough rule of thumb you are correct.
Craig
1 Like
Could anybody explain which diodes I should order to accomplish this?
When I read some other discussions (here and here), they mention different types of diodes...
P.S. not sure if it is relevant for this question, but I will measure my values using a wt32-eth01.
Yes, you are right.
Sorry, I have not been paying full attention. What is the reason you have gone for the pzem-t004 rather than the pzem 016?
No seems I have not been paying full attention. I see now that you were talking above in the discussion about pzem-016. Had overlooked it because I see most others talk about pzem-t004...
Is the pzem-016 better?
I don't know. It has an RS485 interface so can be connected directly to the pi using a usb/rs485 dongle. No diodes needed. Ii had not come across the t004 before. Perhaps someone who knows about both can comment.
When I bought mine I got a package deal, the pzem, CT and rs485 dongle for £15 (from China).
As far as I can tell they are functionally the same. But I have only used pzem-004 version 3.
These can be wired directly to the ESP no dongles required. The only issue is that the PZEM is 5v and the ESP is 3.3v. There a a few ways to deal with this eg modify the PZEM and run it on 3.3v or run it at 5v and use a level shifter on the RX TX pins of ESP.
What firmware are you planning to use on ESP ?
Tasmota supports 3, but others may support more on the same ESP.
1 Like
I run EspEasy on all my other wt32-eth01 devices, so will do here the same (to keep things at least a 'bit' simple and uniform). From the post from @edstobi above, you can see that he has the same setup already.
Ah thanks a lot for mentioning that!! Hadn't noticed yet...
Then I will definately go for the level shifters.
Do you know whether those diodes and pull resistor are still needed when I use a separate level shifter for each pzem device?
With the pzem-016 you don't need the ESP at all. The rs485/usb dongle it to allow plugging directly into a pi. On the other hand, if you don't need a pi there then using the ESP may be more appropriate.
1 Like
I would think not, but perhaps @edstobi could show how they are connected.
You would only need 1 level shifter at the ESP as the PZEMs are daisy chained 
So with EspEasy you may only need 1 wt32-eth01, thats good to know.
1 Like
Yes it would be nice to hear the real life experience from @edstobi.
Found here somebody that claims it does not work when using multiple pzem-004t devices with a level shifter:
Ok. So I skip the level shifters and go for diodes and resistors...
So now I need to make up some stupid questions about that. Here we go again
That BAT54WS seems to be a surface mount Schottky Barrier Diode in an SOT23 package:
But would prefer to have diodes in a through-hole variant, i.e. something like this:
Could build a neat solution with those, by soldering them inline (with a heat shrink tube to seal them...).
Anybody an idea which diode would do the job?
I have 6 modules connected to one esp.
The ESP is flashed with espeasy.
All 6 modules go to the same RX/TX ports of the espeasy.
The address of the PZEM 004T was assigned via the ESP. The PZEMs are pre-set to address 0. Each PZEM needs a unique address.
Since I didn't have any BAT in my assortment, I used normal 1N4148 high-speed switching diodes.
I didn't use a level switch, in many forums you can read that the ESP can handle up to 5 V and the diode also steals something. I left everything at 5 V because the CT817c on the PZEM is designed for 5 V. I poll every 10 seconds. The two diodes on the PZEM then flash almost simultaneously on all modules.
It has now been running for 40 days without a problem.
I then replaced the ESC D1 mini Pro with a WT32-ETH0, as I don't have WiFi reception in my heating system.
I flashed the WT32 with ESP_Easy_mega_20221105_energy_ESP32_4M316k_ETH.
1 Like
Hi @edstobi,
Thanks for the useful feedback! Need that kind of info to get started...
So the ESP's are powered at 3.3V, which means their IO pins are also 3.3V.
Which means you would damage the ESP when you apply the 5V signal from the pzem-004t.
Fortunately it seems that the ESP IO pins are 5V tolerable.
Since the voltage drop of your 1N4148 diodes is 0.72V, the signal from the pzem-004t will only be 4,3V.
Which means your ESP is able to handle those without starting to burn, and no level shifter is required...
Did I summarize it correctly?
