IotaWatt and CTs for USA Mains

Glyn tells me that IoTaWatt sales to the USA are brisk, but that none of the orders included SCT019 CTs. Some of the folks may already have them, or Magnalab SCT750’s which are also supported directly in the tables, but I wanted to post my experiences and recommend this CT for anyone with a US 200A panel.

The SCT019 is rated for 200A and has a 3/4" opening, so it’s a good fit for US panels. I know there has been some discussion in the past about how a 200A panel doesn’t really go above about 100A per leg and how there are other CTs with lower ratios (turns) that will give better resolution, but I don’t think those are compelling reasons to use another lower rated CT. The panel is 200A. The breaker is 200A. The cable is sized for 200A. So my feeling is that you should use a 200A CT. There is no downside.

I use the SCT019 almost exclusively on a number of panels, and find no problem with accuracy, even on lightly loaded panels. My home panel has them, and my summer electric use averages 22 Kwh/day. That’s an average of less than 4 amps per CT. Yet the mains power matches my individual circuit usage which typically is measured with 1000 turns HWCT-004 and 2000 turns SCT013-000 CTs. Here is a daily plot of my total power vs total power minus everything else that is measured, which is just about everything:

So Total_Power is the sum of the two SCT019’s on my mains. Misc is that sum, minus all of my individual circuits. You can see that most of the time Misc is flat, even though Total_Power shows that things are switching on and off and power use is fluctuating widely. What this says is that the mains are measuring the incremental changes the same as the individual circuits. When the fridge goes on, the main goes up 120 watts, and the fridge input measures 120 watts, so the net to Misc is zero.

The few significant bumps in Misc are legitimate unmeasured circuits. The smaller ones are either the garage door opener or the lights in my shop. The big one just before noon is a microwave in the pantry that is not monitored otherwise.
I don’t dispute the advice to use lower turns CTs in other monitors. I don’t have any experience with them. What I’m saying here is that with IoTaWatt, my opinion and experience are that the SCT019-000 is the right choice for USA mains.

Update: If you look closely at the TotalPower plot, you will see that it appears to have a little noise. That is that it jitters up and down a few watts. That’s the fluky heater in our fishtank. Someday I’m going to swap that out. Here’s what that looks like up close:

1 Like

If anyone is interested here is the STC019-000 CT listing in the shop:

I just ordered one to replace my somewhat flaky EmonTX. I can’t speak for why others didn’t order CTs, but in my case, I already have two SCT-013 (100A:50mA) CTs for my sub panel feed and two WattCore WC1-100-MA100 (100A:100mA) for the mains. I don’t recall why I got the WattCores for the mains, but the 100W rating was related to my somewhat flaky grasp of how a 200A service is rated (200A per leg for a 200A total or 200A per leg?). Based on recent reading, it seems like it’s 200A @ 240, so at a full 200A @ 240 load, it would be 100A per leg - but my understanding of it could be wrong.

The SCT-013’s shouldn’t be an issue, as it appears they are in the list of CTs on the demo and I’ll probably add some SCT-006 CTs later for smaller branch circuit monitoring. I assume I can just select ‘generic’ for the WattCore CTs with 1000 turns?

That’s not going to work for the mains Dave. IoTaWatt uses 24 ohm burden resistors and has an input limit of 3.2V peak to peak, or about 1V RMS. The 1000 turns wattCore will produce 1V RMS at about 42 amps, so anything above that will potentially overvoltage the ADCs. They have protection diodes, so short term probably no damage, but the peaks will be clipped off.

One way to use them on the mains would be to change the internal SMD burden resistors to 10 ohm. If you have the skills and tools to do that, the rest can be done with the configuration app. The resistors are 0805 and I recommend .5% tolerance or better.

Another way would be to add another burden external - say 18 ohms. That would net about 10.3 ohms and would handle 100A. You could just solder a thru-hole resistor across the two leads in the 3.5mm connector.

Each leg can potentially carry 200A. The 100A number is a practical consideration. You should be able to tell what your situation uses by looking at the data you recorded with the TX. 100A on just one leg would be 12,000Watts. I’ve never seen that in my home.

I don’t think I’ve seen more than 5000 watts on a leg as measured by my EmonTX, which at 120v, would be 41.6A, which is REALLY close to the 42A you said would output 1V. Most of my peaks are below 3500W, and the highest in the last week being just under 4700W.

Technically, I probably could replace the SMD resistors, but I really don’t feel like messing with that. Don’t really feel like ‘adjusting’ the design that you’ve found to work well.

So for a replacement, aside from the SCT-019, I assume anything that is A:mA (unburdened) is what I’m looking for (Or for a SCT-019, one labeled 000 AMP)? No offense to the OEM guys, their prices aren’t bad, but shipping ‘across the pond’ is killer - shipping on a SCT019 is almost as much as the CT itself with no tracking, or more than it with tracking.

I feel your pain there.

Take another look at my updated post. You can just add external 18 ohm burdens to bring the net burden down to 10.29 ohms, which will work fine. I wouldn’t consider that an adjustment of the design.

On the other hand, I’ve never tested the wattCore for phase lead, so there is no table entry for it. If you decide to go with SCT019, I’d swap you for those wattcores so I can test them and add to the tables. If you want to do that, PM me with your address. I’ll send them and you can send the wattcores after you get the SCT019’s. I’m in the USA - so USPS.

I had an existing Brultech system that I was replacing which used 200 amp CTs.