Looking to get started with home monitoring and solar in the US

I’m thinking about getting started with the enomPi. I love the fact that it’s open and that the parts can be swapped out. I had a TED 1000 that’s finally giving out. I am pretty upset that TED is a closed system. I’m more upset that they discontinued supporting it years ago and forced users to upgrade or get left behind.

I’ve been familiarizing myself with everything that’s needed to get started, but being in the US, I’m a little fuzzy with what’s needed for the system (since it’s configured for the UK). I’ve read over the ‘North America’ page, but I still have a few questions. I want to monitor the amount of electricity I’m using from the power company, as well as the electricity generated by my solar panels. Although the power company won’t pay me for any extra electricity, I can ‘bank’ the extra power for less sunny days, i.e. winter. I would like to know the amount of surplus or deficit I have.

I think I need the following

  • assembled emonPi PCB
  • LCD case
  • US AC adapter (?)

I have a couple Rpi3s so I’ll be using that (where do I download the image?). I’m wondering how many CT’s I need. I know the standard ones in the kit work. I measured my incoming feeder wire and it’s a 0 gauge cable (8.75mm). It should easily fit the 13mm gap of the clamp.

Currently, I have a dumb monitoring system, which I’m hoping to replace. As shown in the picture, I have 2 clamps connected to the feeder wires; top center. I also have 2 clamps connected to the solar lines to the fuse box (not sure if this is correct).

Fundamentally, I’m not sure if this is how I should setup the enomPi. I know the reference voltage (via the AC adapter) will help calculate the correct energy draw from the two feeder lines, but I’m not quite sure it’ll calculate the correct energy generation for the solar panels. I guess I need help determining if this will work for me. Thanks

Breakers #1&2 are for a 240V Clothes Dryer
Breakers 12&13 are for the Range/Oven (240V)
Breaker 14 is an odd one. It ties to the neutral, I think it’s for the GFCI. It’s for the jazuzzi.
Breaker 4 and the block under it are the incoming solar lines.

I need to clean the box up some… actually, I probably need to expand this one.

Edit - changed Fuse to Breaker(s). BT - Moderator

GitHub. It’s listed on the Shop page for the emonPi.

That depends on how your PV is connected. I’ve got to refer you back to the “Use in USA” page. If the PV has no neutral connection - 1; otherwise, 2 for the PV. And almost certainly 2 for the incomers.
You’ll need to connect them as explained so as to add the currents in the two legs. Where there’s no neutral connection, the current in the two legs is guaranteed to be the same, and multiplying by two in the maths is cheaper than a c.t.

Why shouldn’t it? the voltage is the same across the system, so you only need to measure it once. When you know the voltage, the current and their relative phase, the power can be calculated.

You have only two current inputs on the emonPi, so you won’t be able to measure any of the outgoing circuits unless you add an emonTx or two.

I don’t think I’d disagree with that. :grin:

Yep. a sub-panel is definitely a good idea!

Thanks for the help. So since my solar setup has 4 wires, line (black), load (red), neutral (white) and ground (bare/green), I guess that means I need two CT’s. In total, for a US system with PV, you need 4 CTs then, right?

Also, this means that I need an emonTX in addition to the emonPi? Or can I just run the emonTX by itself?

Yes and no!

Yes, you need two c.t’s for your PV. (Unless you can disregard the current in the neutral - I don’t know whether that would be valid, so 2 c.t’s is the safest answer.)
So yes, you need 4 c.t’s in total.

But no, unless you want to know what is happening on each leg, you don’t need an emonTx. You can combine the two c.t’s outputs into one input of your emonPi, giving you the combined current for that feed. The easiest is to parallel their outputs. Remember, a c.t. is a current source, so you can parallel the outputs as shown on the N.America page - Fig 5b. That way you’ll get one set of numbers for the PV overall, and one set of numbers for the grid connection.

When you’ve decided what you’re having, if you wish, you can run the numbers past us and we’ll do a quick check. With the standard burden resistor inside the emonPi, the standard 100 A c.t’s will give you 50 A max per leg with the same current flowing in each leg (and none in the neutral, i.e. 100 A total), or 100 A max in one leg with none in the other.
If 50 A per leg is not enough, it’s relatively easy to add a second burden resistor in parallel, so you could have up to 100 A in each leg - limited now only by the c.t. rating.

(N.B. We never count the protective earth/ground wire when counting the wires in a supply. So even though there are 4 wires there, it’s a 3-wire system. But we do when counting the cores in a cable - just to confuse you!)

One other condition comes to mind: a 240 Volt load that draws more than 50 amps. i.e. one that uses only the hot legs. It’s not likely he has a single load that big, but if he had two or more loads running at the same time e.g. clothes dryer and stove/oven, he could exceed 50 Amps with little or no current in the neutral.

100A in one leg with none in the other would be pretty tough to do. He’d almost have to do it deliberately. The even numbered breakers are on one leg, the odd numbered breakers on the other leg to help prevent that specific scenario. He’d have to know (or be very lucky at guessing) which of his receptacles and lights are on the same leg, then connect 9 or 10 (that use nearly the max 120 Volt rating of 1800 W) loads to those receptacles to get 100 Amps on that leg. In short, not a likely scenario at 120 Volts.

I use three CTs, one on each Service Entrance Wire (incomer) and one on the PV back-feed.

I put this paragraph in to make it clear that a burden change might be necessary, because the total current seen by the two 100 A c.t’s will be 200 A, which needs an 11 Ω burden to give us the maximum 1.1 V rms across it.

His Service Entrance Wires (incomers) are AWG 1/0 which have a max rating of 150 A.

So he will be limited by the c.t. rating to 100 A per leg, the two c.t’s together seeing 200 A (or sightly more precisely, and what might be confusing you: the two secondaries in parallel will be giving 50 mA each,100 mA in total, necessitating a 11 Ω burden). To go above that means moving to a pair of 150 A or more c.t’s.

If his PV system is 240 V, as the majority of them are, why not use one CT on each incomer and one on the PV feed?

My PV system is 240 V, yet uses the neutral. I use one CT on the emonTX and cross check it with a commercial power transducer that also uses one CT. The readings match quite well.

I thought that was a distinct possibility, and it’s what I originally had in mind. But as Marco doesn’t seem sure, I proposed two c.t’s as the safe (certain) option, but recognised that most likely it would be overkill).

My feeling is the controls will be 120 V, but only a few tens of VA at the most. Or maybe it’s for measurement only?

Controls and displays, if any.

My system uses sends its measurment data to the aggregator box via the AC line, and since the aggregator itself is a 120 V device…

Wow, thanks for the responses…

Looks like the CTs are terminated with a 1/8" stereo plug. To combine two CTs, do I just get a stereo plug splitter? I currently have a couple spare CT’s lying around (ie in the picture shown in the first post), is it possible I re-purpose those?

With regards to my PC system, I believe it’s 240V. It attaches (post in the first picture) to the bottom two fuses, on the left side. That should mean it’s going to the two legs… but the sticker on the inverter box says either 248Vmp or 340Voc. Not sure what that means.

So how would I only use 1 CT for the PV system? Voltage is constant at, say 120V. I have a neutral and ground wires, as well as 2 feeders that go to the two fuses. Do I just measure one of the feeders and multiply by 2? I guess it would be easy to check if the math is correct. The inverter has an LCD that displays instantaneous power… should be pretty obvious if the math was wrong.

Provided it is not the stereo to two mono variety, yes. Or cheaper, solder the two into a single plug. We use tip and sleeve, there’s no connection to the ring.

Yes, provided you know their rating. You need the current ratio and you need to know that their VA rating (almost never specified directly) is enough to give you 50 mA (or 25 mA if you are paralleling two) at about 1.1 V rms across the burden resistor.

It actually says: you get max power out at 248 V, the maximum voltage anywhere on the system is 340 V.

Exactly. If you have two c.t’s, you’re measuring the current twice and half the voltage; if you have one c.t., you either measure all the voltage (240 V), or half the voltage (120 V) and double it!

Correct. It’s back-feeding a double-pole breaker, which means 240 Volts.

As Robert mentioned above, 248Vmp is the voltage at which you get max energy harvest from your PV modules.

340Voc is the maximum PV-module-string Open-Circuit voltage, i.e. when they’re not connected to the inverter.