I’m monitoring my washer using a single CT on a dedicated 120v circuit (North american single phase setup). When I load the status page I see the “reversed” icon appear and disappear randomly during operation.
For example, these screenshots are seconds apart:
The input is configured like this:
The graph, in case it helps:
Am I doing something wrong? or is there something about washers that would explain this, and I should not worry about the reversed icon (just focus on watts)?
To take a closer look at it, can you check “allow negative values” for the washer CT and post a graph of the next run? It would be good to zoom in on a about a 5 minute section of moderate Watts. (On a computer just select the part you want with the mouse.
There’s nothing that I’m aware of with washers, although many modern washers are inverter and so could be highly reactive.
Will do! Probably later today. I looked closer at the labelling on the washer, and wouldn’t you know it: “Inverter direct drive motor” 
Sounds like this is to be expected, then, but I’ll update with the graph so you can confirm. Thanks!
Hmm I see no negative values, despite checking the allow negative values. And if I check the usage after the cycle is complete, I see minimal negative values (-800 milliwatts).
I guess as long as the graph data is correct, I shouldn’t worry about the status page. (I forgot to monitor the status page while this cycle ran 
, I’ll see if it happens again next time)
That last graph has negative power albeit in the mW range where electrical noise could be causing it. Nevertheless, it is consistent for more than a few seconds so worth looking into.
It may be a witch hunt, but a hi-res picture of the washer CT in the panel also showing the CT cord and what it runs near.
I’m also not convinced there isn’t something about the service. I’m thinking maybe a poor neutral connection to your panel. Do your lights get dim and bright? It’s harder to tell these days with LEDs but anything strange like that would be of interest. To that end, could you plot your voltage for the last hour and include the statistics tab with min and max.
(would have responded sooner, but the forum blocked me from replying for 24h 
)
Voltage graph from the last hour; Not sure how much variation is normal, but I included total power as well since it seems to line up with the larger dips:
If the connection issue is only on the washer circuit I would have no way of knowing, the washer is the only thing on it.
I can’t easily get a picture of the panel, I had an electrician put in the CTs when they were doing other work. I’m good with low(-ish) voltage (I’ll run PoE cables, for example), but I’m not touching 120/240 if I can help it
It’s possible the proximity of so many leads (especially them being taped together) is causing some signal issues, but overall the numbers being reported are more than good enough for my usage. Given the negative milliwatts happen only for a short time after the washer stops running, I’m not too concerned, I could easily believe signal issues caused by the washer too.
I’m also seeing fluctuations in the 100s of milliwatt range for circuits that are off, both positive and negative, so there’s definitely some noise in the system. It also varies based on what else is operating, especially the dryer, which I guess is to be expected.
I was more surprised by the positive negative flip during washer operation since it was at ~170 watts, so the noise should not have caused that, but now that I can’t seem to get that to happen again (I just stared at the status page for a few minutes with the washer running), I’m fine with chalking it up to ghosts in the machine.
Just wanted to close this out by saying that I’ve been comparing the 15-minute interval data provided by my utility billing with the data collected by IoTaWatt and the results are well within acceptable parameters. The error is <3%, and you can’t even tell that there are two separate graphs in this screenshot 
I’m very happy with the setup overall, thanks for all your work on building and maintaining the platform
