Is It split-phase or 120V/208V “two-phase”?

I recently installed my IoTaWatt, and I’m having a few measurement discrepancies that I’m unsure of. I’m in Canada, in a home with 2x 100A 120V split-phase power service, ECS24200 CTs on the main lines and 4x ECS1050s on other circuits.

One of the circuits I’m attempting to monitor is a 2x 120V 30A breaker for a clothes dryer. I read a related thread on this (Question about monitoring clothes dryer) and it seemed like the simplest approach was to pass one conductor in the opposite direction through a single CT (ECS1050). I didn’t care about measuring the individual power on each leg, but thought this approach would get me the total power usage.

When the dryer is running, things look about right… there’s draw on the both Mains, and a reasonable looking power usage for the Dryer.

When the heating element isn’t active is when things start to look strange, as the Dryer circuit appears “reversed”. My interpretation of this is that the direction I clamped the red & black wires / orientation of the CT is inverse of what would be needed for the 120V leg running the dryer motor to be “positive”. I could either flip the CT, click the reverse button, or ignore the reverse arrow; but regardless my measurements seem plausible…

But then I also noticed that when the dryer motor-only is running and the rest of the house’s power draw is low (relative to above, the “Garage” circuit was shut off), the Main_B suddenly also shows a reversed draw. My 200A CTs on the mains are installed in opposite directions from each other, and generally don’t show this icon:

So I’m not sure what that means. My best guess is that the current on Main_B in this screenshot is so low that measurement errors are creeping in? Any other possibilities?

I also captured this weird situation earlier in the day – here (Main_A + Main_B) is less than the power of all the other circuits, even though there is definitely miscellaneous draw throughout the house as well. If I take (Main_A + Main_B) and subtract all other circuits, I get -55 W.

Both of these two situations feel like the same thing – small scale measurement error. Is that the right analysis, or is it possible that I have I messed up the installation in a way to cause these discrepancies?

As I review these screenshots, I noticed that the second Status view (running time: 5h 8m 12s) also has a discrepancy where (Main_A + Main_B) is less than all the sub-circuits; in that case by -175W. :man_shrugging:

Is this an apartment or condo with 120V/208V?

Nope, it’s a house. No solar or other power generation, either. Here’s a shot of the installed CTs on the mains; although the shadow makes it unclear, they are on the two separate cables. :slight_smile:

Those numbers sure look like a 208V service. Can I see a picture of the whole panel? Do you by any chance have a voltmeter to measure the voltage between the two mains?

Here’s a full shot of the panel.

I’m not sure I know how to measure the voltage between the two mains; but I do have a multimeter… but probably not the confidence to do this safely. Would the reference voltage from the IoTaWatt show something other than ~120V if this was the case?

Yikes! That is one overloaded 100A service. But I guess it works.

I guess you don’t have a voltmeter. I’m pretty sure this is a 208V service, which is not split-phase. 208V is two legs of three phase. I’d like to try something simple to prove it out. This may take a couple of tries because I don’t know which main the VT is plugged into or the phase relationship of the second leg, but we can narrow it down.

At the bottom of the Inputs configuration display, please check the “Enable derived three-phase” box. The click on your Main_B input and in the configuration for that, select Mains Phase B. Save it all and post the status screen for me to see.

OK. Here’s the updated inputs page, just to ensure I didn’t mess it up. And the updated Status page.

By the way, I appreciate your time to help figure this out; thank you.

Dont mention it.

Looks like the VT is on Main_A, but Main_B is off. Can you change it to phase C and report please?

Changed Main_B to phase C:




That’s a little better.

Looking at the picture of your whole panel, I don’t see the branch circuit CTs. Where are they?

Can you turn on the dryer for a minute and give me a shot of status with that running? (The dryer will be wrong, but I want to see what happens with the mains)

OK, I’ve been trying to get the voltmeter reading in the meantime – between the two mains reads 217V. I guess I need to read up on three-phase power; I had just blindly assumed this was split-phase because it sounded like what it looked like.

Here’s a photo of the panel with the branch circuit CTs; my earlier photo was before I had installed them.

Here’s the dryer circuit running:

And running tumble-only:

OK, this is looking good. Still a few things to clean up. How much time do you have?

Plenty of time. :slight_smile:

OK, first I think you should check “reversed” for the mains_B input. That will get rid of the arrow.

Your dryer is going to need two CTs to read accurately. If you have another CT, could you put one on each of the two conductors and configure one on Phase A and the other on Phase C?

It does appear to be “two-phase”, but the voltage between them should be 208V, given the consistent 120V that I see on phase A. What does your voltmeter say about the 120V legs?

Reversed Mains_B – that one is easy.

The two legs read 123V and 121V to neutral.

I don’t currently have an extra CT, but I have a few more on order that I could use. Let me make sure I understand roughly correctly; I can’t fire the two legs through the CT in reverse order because they’re not 180º from each other, so I won’t get an accurate reading. To be accurate I’d have to measure both separately… if I measure one and double it, I would be either under-accounting or over-accounting that circuit.

Exactly. In your case, any three-wire (with neutral) high-voltage appliance needs two CTs. If you have 208V appliances without neutral, there is a way to handle them with one CT.

But there’s more. All of your 120V circuits that are on Main_B have to be specified as Phase C. They will read about half otherwise (cos 120°). So you might want to first identify which of the mains is Main_A and which is Main_B.

Your load center appears to have 13 horizontal rows of circuit breakers. Most of them have double breakers on them. The phase of the circuits on each row alternate as you go down. Whatever Main is the lower of your mains circuit breakers corresponds to the same phase as the top breaker on the right side. For the sake of this discussion lets say it’s Main_B and so it’s phase C.

The next row has an empty slot on the left and the right breaker will be the same phase as the upper main, so that will be Phase A.

The next row has a GFCI breaker on the left and half of a 40A pair on the right. Those are Phase C.

Now the next row is Phase A, but it has double breakers on both sides, so all four of those circuits are Phase A.

And so on. So this is a little bit of the split-phase logistics to the panel layout, but three-phase when it comes to voltage/phase reference. This is timely because the latest release just out has improvements that apply to exactly this kind of service. IoTaWatt can deal with everything in there.

I have you using derived voltage/phase reference. You can add another VT to get a direct voltage and phase reference for the other leg. You would need a convenient receptacle that is connected to Main_B. My recommendation would be to get it all going correctly with derived and see if that works for you.

You can combine any number of conductors on the same phase to economize on inputs. Your order went out today. I’m guessing the 100A CTs are for those 40 Amp breakers. Since you will only be able to measure one side, you could use 50A for those. Judging by the black/white conductors on the upper right breakers, I think that’s a two wire 208V circuit, probably an AC compressor? That can go with a single 50A CT. Those with red/black are typically three-wire and require two 50A CTs.

If you want to return the 100s for more 50s, I can do that. If a 100 will fit comfortably around your mains, you could swap the two 200A for five 50A.

OK, almost everything makes sense… except maybe that I’m misinterpreting K->L on the CTs.

I created a spreadsheet to layout the panel and associate each breaker with its phase, traced which of the main circuit breakers were on which input to the IoTaWatt, and confirmed that the VT is on the the same circuit as Main_A which is on Phase A. The rest of the CTs I had provisioned so far landed on Phase A as well; including the garage & dryer, which I’ve pulled just a single leg through for now. I also fixed Main_A & Main_B to have CTs in the same direction.

I think everything looks good, except that every circuit is has a reverse arrow on it.

Do I just have the CTs backwards, now? I thought K was the source, and L was the load, so I’ve applied the same directional thinking on all of them.

When the 100A CTs arrive, I’ll check if they’ll fit comfortably around the mains, and get in touch regarding an appropriate return/exchange. Is there a better contact method for that discussion when it comes? Not sure if the community postings are the right place when that discussion won’t benefit anyone else. :slight_smile:

Ah, I reversed the polarity of the VT to fix that. :+1: Excellent. I think this is all on-track now.


Use PM (private message). Click my emoji and there will be a “message” button to send me a private message.