Question about Current output

[overeasy]

You are definitely a candidate for integrations. Right now it’s available to ALPHA and BETA users in release 02_07_04. You could change to auto-update BETA if you want to use it now, otherwise it will be a few more weeks before I upgrade MINOR.

You are probably aware that there are a few different ways to connect a Powerwall, and there are other users who do it differently. It appears you have it connected to feed quite a bit of load. Given it has a 25A capacity, I’m curious how your disconnect/transfer works.

Typically, the disconnect activate when the grid goes down so there is no back feed. A transfer is usually to inject power from a generator. I don’t see one in your system.

Are there CTs somewhere that control when the Powerwall charges/feeds-in? Can you show that and explain briefly?

What’s at issue is that Powerwall usually can provide power during a grid outage. When it does, it is producing the voltage. When it isn’t, the grid is producing the voltage. The question is where the VTs are connected in this scenario.

Its a complex setup. IoTaWatt is very flexible and can handle just about anything, but need to get it setup right.

[iotapl]

I’m wondering if what I see is “real” reactive power and some way how the PW works…

@overeasy

Are there CTs somewhere that control when the Powerwall charges/feeds-in? Can you show that and explain briefly?

Yes, there are two sets of “visible” CTs, one for solar and one for mains. One on each phase for a total of 4CTs.

I installed the IoTaWatt CTs alongside these for those two feeds.

For the power wall, I installed an extra pair of CT’s, they measure the current internal to the transfer switch or the power wall…

The transfer switch (see modified diagram below) will detect a loss of voltage and disconnect the utility side from the Solar and the Power wall isolating the grid for self-generation.

I do not have a generator, the Transfer switch (rather an automatic disconnect sorry for the misnomer) will
just open up when the grid looses power.

In my diagram below, I get a 150A breaker feed from the meter to my main distribution panel. On there, I measure with the CTs (as well as the power wall). (Note: all loads I will describe here are 220V loads. On the first two panels, all loads are 220V)

In that same panel, I have 2 loads that are also measured by IoTaWatt: The Range and the Clothes Dryer. I have another feeder to a sub-panel in the garage but I will eventually add another IoTaWatt there to measure the incoming current to that panel and the individual circuits.

From that panel, I have a straight through connection (after the 150A main breaker there are lugs that go yout with no further breaker) into a simple manual disconnect switch, required by code.

From the switch, I go to Tesla’s Transfer (rather called Tesla’s gateway) switch, again, an automatic disconnect. Tesla’s CTs go into that circuit board and I presume they also get voltage measurement there to know when the Grid lost poser and open the switch (rather to know when it comes back).

From Tesla’s Gateway I go to my generation panel. I get 7.4kW on 220V from the solar inverter. And the power wall can deliver/consume 5kW, also connects in that panel. I monitor both loads with CTs. That gives me a total of 10CTs.

From that panel, I get a 220V for IoTaWatt (more of that later) and a feeder to my house sub-panel that are not monitored with IoTaWatt. Again, as soon as I get my two recently ordered IoTaWatts I will add monitoring to the house panel Feeder (total of all circuits) and individual circuits.

So as you can see I am measuring every line in a 220V fashion (both sides of the split phase)

I have 220V to the IoTaWatt because I am using two TRIAD transformers (120 to around 10V) to measure each side of the split phase. I did this just because I can… I did a field calibration with a TrueRMS DMM with 0.08 + 0.03 accuracy spec for 200V@60Hz

So each CT on phase A is referred to VphaseA and same with CTs on phase B with VphaseB

Now that my Power Wall is fully charged, I get this:

Note that the current is the same in both channels, and VA and VAR are the same… with the Power being at 0…

Still trying to find if this is a behavior of the Power Wall or what…

Also, I tried swapping channels and the issue traveled with the Power Wall…

[iotapl]

I just updated to Beta to try integrators.

I have gave up on the current of the power wall as I still need to connect the main CTs and I need them to test a theory (that the PW is helping the grid correct for pf, the interesting thing is that my vector is at 45deg as they have the same va than var)…

But going through this exercise I wonder if there is a way to retrieve the actual value (with sign) of the current… the output seems to give me an absolute number. And I would need this to test my second theory (that current is flowing between phases).

[overeasy]

@iotapl

But going through this exercise I wonder if there is a way to retrieve the actual value (with sign) of the current…

Current in measured in RMS and has no sign.

@iotapl

And I would need this to test my second theory (that current is flowing between phases).

Assuming you are split-phase, virtually all of the current is phase-to-phase. You can measure tho current that is not phase-to-phase by measuring the current on the neutral conductor.

Also, you are using Triad transformers. If you are trying to measure PF and VAR, you would need to know the phase-shift of those transformers.

You would do much better using a single 230V transformer scaled down by half as a reference for both legs.

[iotapl]

You are right… I think the shift should be small, but I will test them. Is there anyway I can write a complex calibration value to correct the phase? I really do not expect it to change that much though…
I have tried doing the same reference for both inputs and I get basically the same results.

[overeasy]

You may want to read through this

120V/240V split phase voltage reference

I’ve been monitoring my US split-phase service for two years now using a single 120V reference on one “leg” of the service. The results have been excellent, with overall and individual results within 1% of the utility meter and several revenue grade EKM meters on individual circuits. I long ago became comfortable with the notion of the single 120V reference, even though that voltage typically varies between 119V and 121V throughout the day with excursions even further off the mean from time to…

[iotapl]

I did read it… Right now my house has a problem (that I am trying to fix when my new panel goes in) which is that it is heavily out of balance. One leg has about 80% of the load. And part of the reason I am doing both legs is to be able to see how much voltage drop I get on each leg.
So far I have found that it is about 1V. The difference between your setup and mine is that I am measuring with respect to neutral. So technically I am getting info you are not (the drop on neutral) and I am missing info that you were getting (the phase voltage, 220V). I think that none of it really matters but I like data… I might install a 220V transformer on the extra input.

[iotapl]

I’m still playing with settings. My latest iteration on the input settings is:

I am measuring (for now) using the same Vref on all inputs. So I invert each phase.

My home consumption is:

The tesla app shows:

And iottawatt:

It is interesting that the Total Grid is 0, but each phase has some power being imported, and some being exported.

[iotapl]

Regarding the split phase post, I also found that my drier is not very symmetrical:
.
The heater element is 220V and you can see that it has a pf of 1, which makes sense as it is basically purely resisstive.
The motor for the tumbler comes from 110V and it has a pf < 1. in the image above its .90 because it gets averaged with the resistive element. But when the resistive element is off, I get:

[iotapl]

And this is the graph:

So I think my IoTaWatt is configured right… I still need to figure out what the power wall is doing when idle (as there seems to be a real current flow) and how this export to grid works (as one leg is exporting energy while the other is consuming it).

I do have another question, and I will start a new thread…

[iotapl]

This is once the Power Wall is full and I am making 1.7kW with solar and consuming about 1.5kW:

Note that I am not monitoring the main panel from my house (waiting for my 2 new IoTaWatts to arrive). But other monitors tell me what I am consuming…

I think it is important to monitor both legs of main, specially when there is production. Solar production seems to be mainly symmetrical regardless of the load (not surprisingly). What is more surprising to me is that the power wall is doing something… weird… I think it is also important to monitor both legs if one is interested in learning more about PW and the loads on the house… Will try to figure out what it is…

I have ordered the 220V transformer… I will see how much unbalance it has when it gets here… (I is more for technical curiosity than practical usability).

[iotapl]

And now this discrepancy. It’s quite big:

vs

and

It is double the error…

If I click on double for both Mains, I get the same result everywhere…

One thing to note, this IoTaWatt is outside and it is 8degC if that matters.

[iotapl]

Figured it out… wrong CT selected…

[overeasy]

Sorry, I haven’t been paying attention to this thread as I’ve been swamped with orders since Friday. Coming up for air, I see that you have resolved your main question.

I’d like to understand this better and maybe make a demo of how to measure a battery system if it looks generic enough.

Are you using integrations for battery and grid import/export?

How are the primary circuits signed?

• Grid import(+), export(-)?
• Solar feed (+)?
• Powerwall discharge(+), charge(-)?

Are you interested in feeding this into Home Assistant Energy?

[iotapl]

Oh don’t worry… I had ruled out hardware damage by swapping inputs, etc… I knew it was a config issue but had not figured it out… then I remembered that I had bought 200A CT’s but set them to 100A…

I am using integrations for the battery and the grid.
The battery one seems OK, but I am still having some issues with the grid one:

It is always stuck at 0. I was going to ask you how could I delete all data as I do not know if any of the multiple changes in sign, etc has affected the way the integrator works (although I would not think so) (I will eventually want to do so as my logs right now have spurious data from all the experiments)

Grid import is + and export is -
My solar feed is +
My charge is + and discharge is -
I am considering solar generation, but the battery as a load, that can give me energy back. The reasoning is that all loads are + and because the battery is not 100% efficient, it is net as a load.

I will be feeding all this to HA Energy. I am just waiting for the other two IoTaWatt’s to arrive. I will do an electrical panel swap and install both of them on the new panel. I want to monitor about 40 circuits and for now I will be combining some of them… later I might buy another IoTaWatts.

And I want to aggregate all that data and feed it to HA.

I am still deciding if I should just let HA do the aggregation. Or if I should spin an influx db and use that to aggregate the data… Use HA graphna to present the aggregated data but let HA Energy to get the info directly from the IoTaWatts…

I can write up how I have everything connected once I finish with the panel if you want.

[overeasy]

@iotapl

I am considering solar generation, but the battery as a load, that can give me energy back. The reasoning is that all loads are + and because the battery is not 100% efficient, it is net as a load.

I view it as the same thing as the mains. My model of mains with net metering is that it’s a big battery with virtually unlimited capacity and charge/discharge rate. Folks with true net metering charge it up all summer and discharge it in the winter.

Grid + battery+ solar = consumption. Notwithstanding that some of those are negative from time to time.

To clear an integrator

Delete the integrator
Delete the integration file.
Recreate the integrator.

[iotapl]

@overeasy , your thought makes sense… I might change it so that the battery is +… again, because of the battery not being 100% efficient it will always have a net negative (consumption)…

I tried deleted the integrator. But then I went to find the integration file and it was not there.

I restarted the board and now everything is really unstable.

At some point I was able to get to the status page and the grid was still there… and in the graph, the grid name was there but there was an error if you selected it…

It is really slow and loading weirdly:

[iotapl]

I was able to get to the logs…
It seems that the file got deleted but not from the configuration

also note the date/time

[iotapl]

It is more stable now

but also note the grid…

[overeasy]

Grid is currently undefined because you deleted it. You should redefine it now.