I have a PV system which is setup in two branches. One set of PV goes straight to the mains. The other set of PV goes to a protected panel which is also connected to a 10kWh battery. I also have a car charger.
I can get the PV generation data and battery data via APIs. ( I have a spare Raspberry Pi which I could to poll the APIs and push data into the emonBase).
To monitor all of this I’m thinking the emonTx to monitor the consumption / distribution from the grid and EV charging. And then the emonBase to collect everything. I would then set data via the API to the emonBase for PV and battery.
Does that make sense? Or is there a better way to go about it?
There are some details that you’ll probably need to address. First is the 4th current input of the emonTx will probably need to be “de-sensitised” to make it the same as the other three - because obviously the input will be used as two pairs - one pair for the grid connection and the second pair for the car charger - assuming it’s greater than 16 A per leg. That’s a relatively simple case of adding a resistor inside - but depending on the c.t’s you use, you might need to do that for all the channels anyway. Don’t assume you can use our “standard” c.t’s on your cables even if you do run at less than 100 A, you need to measure the overall diameter, and if it’s more than 12 mm, you need to find a c.t. with a larger opening (from the “Use in North America” page in the ‘Learn’ section).
You realise that an emonBase IS a Raspberry Pi, so in fact all you need is the RFM69CW Radio Module to add to your RPi.
Based upon our wiring, I would need two 200A CTs, and two 100A CTs (the leg to the garage is only 50A, but I suspect there’s no harm in going with the 100As to ensure they fit around the wires). Do you have any documentation on the what needs to be done for the “de-sensitising”?
As far as the emonBase, I could use one of my existing Pis, but I prefer to use those for other experimental work. And, because I code in C#, it’s easier for me to run them with .NET Core and deploy code on them written in C#.
You’ll lose a bit of accuracy at the lower end, otherwise they will work well enough.
I’ve written it here a few times, all you need is to add a wire-ended 27 Ω resistor in parallel with R18, the 120 Ω resistor (“1200” printed on it) that’s immediately behind the CT4 input socket of the emonTx (bottom left here):
There are holes ready for it. The resistor need not be any bigger than 0.1 W, though 0.25 W or 0.3 W will probably fit (the holes are 8 mm pitch) and be easier to obtain. That will make CT4 input the same as the other 3. You’ll need to change the calibration in the sketch to make it the same too.
You might need to change the same resistor on the inputs you use for the 200 A c.t’s, I can tell you when you decide which you’re going to use. And the calibration for those must change likewise. You can do that with a programmer and the Arduino IDE (or any serial terminal emulator).
If I read that correctly, it sounds as if you might have CT rating confused with wire window size.
Unless you have a 50 kVA or larger transformer feeding your electrical service, you don’t actually have 200 Amps continuously available.
Excellent point. Thank you for the clarification. I do have two existing CTs on the inbound wire (from an Eyedro meter). Am I correct to assume that any CT can measure current running in either direction in the case of net over-production from my PV? If so I suspect I could just reuse the existing ones.
It’s alternating current, so it’s reversing - in your case - 60 times a second. No transformer will work on direct current .
I suspect you mean power running in either direction. That’s down to the phase relationship between voltage and current, and one of the reasons why you need an a.c. adapter.
It’s worth investigation. It will depend on whether their VA rating is adequate to generate the voltage required by the emonTx’s input circuitry. If you can identify the make and model, then provided the VA rating is available in some form (it’s rarely given directly, more often in the form of a recommended load), we can check.
N.B. NEVER disconnect a c.t. while it’s on a cable and that cable is carrying current.
That’s not good. Our input voltage is 1.1 V max, so the emonTx would be operating over less than a third of its range, so resolution would suffer - in effect, you’d amplify it in software and noise picked up is likely to be the bigger problem, masking very small currents. Plus, you’d have to remove those SMT resistors, or you’d have even less voltage to measure.
I’d recommend a c.t. with a current output: 50 mA at rated current (or even better, the maximum current you are ever likely to need to measure) is ideal and would mean you only need to change the burden for CT4.
In principle, the emonTx can handle any secondary current, but 50 mA or 100 mA are the best choices considering burden resistor values and the connectors, and if you can get a 50 mA secondary current at the maximum you want to measure, then you don’t need to change the burden.
You could even use a 5 A secondary if you had the burden external to the emonTx - but the c.t. would need to be about 6 VA so quite big, and the burden would dissipate up to about 5 W. It’s possible but it brings problems that are best avoided.
So, at the end of the day it seems like picking up two 200A CTs for the mains (the 100A don’t have a large enough opening). And then two 100A CTs for my car charger. In this case, the car will actually pull 40A @ 240V.
Then I can integrate the rest of the data via the API interface.
The Wattcore WC2 is available with 100 A rated max. current, 100 mA current output is available as an option. I’ve no idea of the price, nor delivery. The opening is 1" × 1.4".
I can’t find any listings for the WC2. Here’s one for the WC3. 100:0.1 A. The wire window is 0.75 x 0.96 inches,
plenty of room for AWG 4/0 Service Entrance Wires.
It’s listed on Amazon, but sold by a 3rd party vendor. Price is 39 USD.
.