Random extremely large amperage, voltage, and wattage spikes

[dierk017]

Has anyone experience random spikes on amperage or voltage or wattage? I think this started after enabling software spi to get the 3rd add on board to communicate. I know this isn't actual usage as some of the spikes are on a circuit that is turned off at the breaker and the device records -8000 watt spikes when I don't have anything to return power back to the grid. I've had voltage spikes over 600 volts and amperage spikes over 80 amps on a circuit that has nothing in use. This seems to occur on all the boards but not at the same time.

I've tried compiling the build in Arduino vs esp-idf but the esp32 board does not like it being compiled as Arduino and becomes unresponsive. I tried a suggestion online for randomizing the polling time and put in random times that didn't seem to help. I'm using the nodeMCU-32 from circuit setup so I imagine its capable of supporting it. Anyone have any idea's?

[CircuitSetup]

I haven't seen something like this as extreme before, but it sounds like you're picking up interference from something. Is the meter near anything that could potentially interfere with it? Is it only happening on one CT? If so, make sure that CT is fully closed. Also try swapping it with another channel and see if the problem follows the CT or the channel.

However, since you mentioned that you're getting spikes of over 80A, that's very suspicious as the max the meter can report without scaling is 65.53A. I suspect data corruption somewhere, especially since you're seeing erroneous number with voltage too. Is the ESP32 having trouble staying connected to your network? What is the signal strength?

[dierk017]

The meter sits right below the electric panel outside side the panel and has clear line of site to an access point about 10 feet away. . RSSI according to the meter is -42. I can try make a card to monitor it via pings but uptime at the moment is 811,353 seconds. I thought it was having uptime issues so i added the uptime sensor. It seems its random on what sensor throws the weird spike and doesn't seem to to be affect every sensor at the same time. I've also noticed this seems to only occur for one polling cycle which is around 10 seconds for each sensor that I poll but I have them set to poll +- .1 second from 10 seconds with no group of sensors polling at the time. I added filters to keep it from going negative current usage all the time as I was getting -687,000 watts on my dryer on my dryer when not in use. I think the 80 amp part might have been from me looking at my main current clamp with 4x multiplier but I am seeing random spikes of 35-40 amps on multiple circuits that are not in use.

[CircuitSetup]

humm, since you have the multiplier, it's probably not data corruption.

Do you have a spare ESP32 to test with (that has the proper 19 pins per side)?

Try adding harmonic power to one or both of the mains CT's. I'm wondering if it's actually some mid-high frequency interference from your electrical system itself.

Another thing - the latest beta of ESPHome was just pushed, and has some new calibration features that may help 0 things out. Specifically, the offset calibrations. See how things are set up here and here if you want to try this out. Pay attention to this part: Offset calibration can only be performed when all voltage and current inputs are at a 0 value. USB power is recommended. For power offset calibration, only a voltage transformer should be connected.

[dierk017]

Just to update, one of the main clamps recorded a -687,195 W spike this afternoon. Going through amp or voltage spikes I don't see any that match the same time. I can pull some of the filters to see if I get any that match in but i suspect it will be continue to be random.

[dierk017]

I added the harmonic power and will keep an eye on it. Is there anything specific I should look for on the harmonic power? I noticed that my air conditioner circuit was showing the spikes even though the circuit breaker upstream of it was turned off. I don’t have a spare esp32 but have one on order and will try it on different board when it arrives.

…

On Thu, May 15, 2025 at 8:09 AM John ***@***.***> wrote: humm, since you have the multiplier, it's probably not data corruption. Do you have a spare ESP32 to test with (that has the proper 19 pins per side)? Try adding harmonic power <https://next.esphome.io/components/sensor/atm90e32.html#harmonic-power> to one or both of the mains CT's. I'm wondering if it's actually some mid-high frequency interference from your electrical system itself. Another thing - the latest beta of ESPHome was just pushed, and has some new calibration features that may help 0 things out. Specifically, the offset calibrations. See how things are set up here <https://github.com/CircuitSetup/Expandable-6-Channel-ESP32-Energy-Meter/blob/master/Software/ESPHome/6chan_main_calibration_beta.yaml> and here <https://next.esphome.io/components/sensor/atm90e32.html> if you want to try this out. Pay attention to this part: *Offset calibration can only be performed when all voltage and current inputs are at a 0 value. USB power is recommended. For power offset calibration, only a voltage transformer should be connected.* — Reply to this email directly, view it on GitHub <#199 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AXYAUQW65EK44JZJ4JQO3F326SG2FAVCNFSM6AAAAAB5EDUP2OVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGMJVHA3DKNA> . You are receiving this because you authored the thread.Message ID: <CircuitSetup/Expandable-6-Channel-ESP32-Energy-Meter/repo-discussions/199/comments/13158654 @github.com>

[dierk017]

I am not sure what specifically fixed it but replacing the ESP32 board combined with compiling it as arduino and not doing software SPI I've not had any of the random spikes.