Thanks - I'd actually come across that and have already had a play with it. It worked reasonably well, and I liked things like the sanity check for the next possible increment in case of a bad ocr reading, and it was very flexible in that it did dial meters too (I think from memory), the only issue is with the focal length of the camera lens. My cupboard has a door that goes up to the front of the meter, and so the image was a little blurred. However I was surprised that I could unscrew the lens for a pi camera and got a usable image for the last 3 digits of my meter (too close to read them all). I'm currently having fun playing with opencv and thinking of other applications around my house.
It does require reference voltages, which are obtained via voltage transformers.
As I recall, iotawatt has 2 modes, one which requires 3 voltage transformers, 1 for each phase which are wired directly into your supply (just like PZEM). The transformers reduce the AC voltage, making it easier (and safer) for iotawatt to analyse.
The second mode uses derived reference by just using 1 transformer on one phase, and then shifting it's phase by 120 and 240 degrees to assimilate the other 2 phases.
This is less accurate.
So the upshot is that it does also require connection to your live wires.
Sorry @craigcurtin but I don't share your enthusiasm for iotawatt 
1 Like
Unless it is battery operated then you would need a power connection anyway, to power it. For single phase it is not usually a problem as it is just a matter of connecting the Power In wires into the sensor circuitry directly, I suspect this is automatic in the case of IotaWatt. For three phase it is more of an issue, but the IotaWatt's ability to use just one phase for volts (which it needs anyway to power it) and estimate the others may well be sufficiently accurate.
1 Like
Further to the above, it seems it is not necessary to have the electronics close to the current clamps. I was surprised to find that (with the PZEM anyway) it works perfectly well with 8 metres of telephone cable between the PZEM and the clamp. My meter is in a box on the outside of the house, and this allowed me to have the electronics inside the house rather then in the meter box.
In case anyone is interested this is what it looks like in my 3D printed box.
2 Likes
Just for info, I installed a CT for a inverter to feed power back to a clients house from a battery pack, The cable length was 100 M+ cat 6 , with no issues.
1 Like
Cat6 and Telephone cable is not designed for AC, you might as well have installed in the meter box, one is as risky as the other.
A few meters of AC cable is a few bob, not wort the risk to save this.
What risk? There are no volts on the current clamp connection. It clips round the outside of the cable.
The voltage is minimal much less the 48v used for POE. The alternative is a 1 pair twisted screened cable. If the client installed inverter in meter box then there would be a need for 100m of 15kw rated 50v DC cable. As Colin asked, what risk?
As long as it is not a Mains AC feed.
You are also working with induced current, if the cable is too long additional induction could be introduced, then you should look at screened cable.
As I said, it works ok so for me it is not a problem. In some situations screened cable could be useful, certainly.
1 Like
Lol. I had thought that you could put a hat on a zero or zero 2 W or use a different device and eg do POE - in these times of high prices I had guessed that accurate overall consumption measurement was the best overall outcome as the big users of electricity can normally be guessed by wattage and time we know devices are on for - I only have experience of a 1 phase supply but as I say the values I got were indicative but not absolutely accurate.
I follow threads such as these to find out what others are doing and learn - that's the fun
Bart,
OK a couple of things
The IOTAWATT needs you to couple a power supply transformer to it - there are a number of recommended ones - this provides the Voltage reference - one needed for a single phase - if you have 3 phase (and the voltages of the phases and quite close) then you can use a Derived voltage from the the first unit for the other phases - in my case we have very similar levels on our phases and i experience less than 1% inaccuracy - or if you want to sacrifice more reading channels then you can have 3 voltage references - one on each phase.
The unit is then powered through the USB power and you need to provide a quality mains to USB charger of at least 2 amps
Be careful with the Emporia vue - it is a good unit but out of the box it needs the cloud to work and only updates every 6 seconds (it may be longer now - you would need to check)
There is a project on ESPHOME to reflash the emporia with ESPhome and then this feeds directly to Home Assistant - (as does the IOTAWATT)
If you want a quick home energy dashboard wiht pretty graphics i would highly recommend setting up Home Assistant with an IOTAWATT - if you have a spare Virtual Machine or RPi you can literally have it up and running in an hour.
You can then do whatever else (fancy wise) in NR by grabbing the data out of HA or directly from the IOTAWATT as time permits.
Craig
1 Like
Can i ask why Paul ??
I have 3 of them here and they have been rock solid and give me all the data i need in a nice neat package - to have 40 total circuits nearly any other way would be a nightmare of units stacked around the place etc
I started off with a single unit to monitor my Solar, Mains, consumption and Battery system (all 3 phase) and then purchased additional units to give me circuit level monitoring
Craig
Not wishing to move this off topic... lets just say previous technical disagreements with it's developer whilst in the openenergymonitor.org forum some years ago, when Iotawatt was first being developed.
I'm going to leave it at that
But good to hear that it's working OK for you.
1 Like
Sorry don't have time to catch up the whole thread right now but I bought a Shelly 3EM for our cottage this summer. It turned out to have some unexpected complications and I wasn't able to convince my electrician friend to install it on the main building's electrical panel. Main issue was that it turned out to be rated only for indoor usage at +0C - +40C temperatures (I hadn't noticed the tech spec has been silently replaced within a year!).
I'm thinking of going for Carlo Gavazzi EM340 which is pretty decently priced and is an established brand which even my electrician friend knew. It's not based on current transformers but is a traditional direct measurement type of device. At first I thought a current transformer based solution is handy as "non-intrusive" but in practice you'll also need to measure the AC voltage on all phases by connecting each phase as inputs. To do this properly, they should be fused (or at least this is my current understanding).
The EM340 communicates via RS-485 Modbus. At least Tasmota seems to list support: Smart Meter Interface - Tasmota
2 Likes
As discussed earlier, the CT's can be some way from the Shelley, so possbly it can be indoors. Also I thought someone said that it had a mechanism allowing you only to connect one of the phases, which again can be done via a normal fused plug indoors.
Will try to explain what I learned from the above discussion, and from the iotawatt documentation. Not saying that iotawatt is the way to go, but at least the documentation contains lots of usefull information...
-
When you just want to measure trends in your home (i.e. which circuits use most of the current or cause peak currents or ...) then it is enough to buy a non-intrusive device with CT (current transformer) clamps. So no need to start messing with your power wires.
-
When you want to measure the power (to know which circuits are going to cost you most in your energy bill) then measuring induction current with CT clamps is not enough. Because - depending on the load of your circuits - there might be a phase shift between the voltage and current sine curves (picture copied from here):
Which means you cannot simply calculate power = voltage * current
-
So in this case you need to measure both voltage and current. The Shelly 3em for example measures current via non-intrusive CT power clamps, but requires the power wiring to be adapted to make sure the power wires are also attached to the shelly device:
-
I only had a quick look at Iotawatt, but it looks to me that they solve this by measuring the voltage also in an non-intrusive way. They do that by using (more expensive) VT (= voltage transformer). I did not have time yet to see whether this is also possible via clamps...
-
So you can use the direct reference option in Iotawatt, which means you measure the voltage for every of the 3 phases. And then you need to specify in Iotawatt for every CT clamp to which of the 3 (voltage) phases this circuit belongs (see doc):
-
But iotawatt can also use a derived reference, which means you only measure the first phase and you apply a hardcoded constant shifting of that single reference by 120° or 240°:
Which means the derived references will be an approximation of the real phase shift on those two phases. But in most cases they claim this will be accurate enough...
Imho me it looks like Iotawatt has solved this quite decently...
I would appreciate a lot of somebody could digg a bit deeper into the non-intrusive voltage transformer measurement options. Would be cool if we had a decent/affordable non-intrusive solution to measure not only current but also voltage. Perhaps this is possible. Don't know. Unfortunately I don't have time to figure all of this out on my own ...
1 Like
It's simply a transformer, that would have a primary voltage (input) of 250VAC and a secondary voltage (output) of 12VAC (or similar low voltage).
The primary circuit is wired intrusively into your mains supply (1 phase), and the secondary is wired into the iotatwat, which being only 12VAC is safer than measuring 250V.
Because there is a proportional relationship between the primary & secondary transformer windings, the device can calculate the mains voltage (after calibration).
So, sorry, it's still intrusive.
1 Like
Ages ago I have bought the emonTx from openenergymonitor:
It is still running fine. Never had problems with this hardware.
Maybe in the mean time there is better hardware available on that site - that I don't know.
In my setup it is monitoring:
- 3 phases with 3 CT clamps
- solar power production with 1 CT clamp
- voltage
Regarding the voltage monitoring - I only need to plugin the adaptor of the emonTx in a AC power socket. So it is not only used to power my emonTx but also to measure the voltage of one of the phases. The voltage for the other phases can be estimated by adding a phase as you commented earlier.
The data is sent via 433MHz radio. To recieve this data I have also bought a RFM69Pi hat for my raspberry pi:
So the data is processed on my raspberry pi and sent to node-red via MQTT.
Regarding the software setup there is also plenty of documentation available on the open energy monitor site.
Here below a diagram of the components:
3 Likes
In the newer EmonTX's like @janvda 's, the voltage transformer is built into the emontx, so it can serve a dual purpose; measuring the voltage & providing it's power.
I use an older emonTX version which requires a 5VDC power supply, and a separate 9VAC power supply to measure the voltage. Both are plugged into a double mains socket.
1 Like
