PZEM-004T - detect electrysyty failure

[LevKos]

The problem

If there is no power to the high-voltage part of the PZEM-004T, it does not output anything, but the platform has processed that if there is nothing, the values do not change. Electrysyty failure cannot be tracked. I suggest making the following changes to pzemac.cpp:

void PZEMAC::on_modbus_data(const std::vector &data) { if (data.size() < 20) { ESP_LOGW(TAG, "Invalid size for PZEM AC!"); // start add this->voltagesensor->publish_state(0); this->currentsensor->publish_state(0); this->powersensor->publish_state(0); this->frequencysensor->publish_state(0); this->power_factorsensor->publish_state(0); //stop add return; } ... Anern Grid Voltage - Hybrid Solar Inverter Energy Monitoring External Voltage - PZEM004T

Which version of ESPHome has the issue?

2023.12.7

What type of installation are you using?

Home Assistant Add-on

Which version of Home Assistant has the issue?

2024.1.3

What platform are you using?

ESP8266

Board

d1_mini

Component causing the issue

PZEM-004T v3

Example YAML snippet

No response

Anything in the logs that might be useful for us?

No response

Additional information

No response

[DAVe3283]

I like that idea, but it should publish NAN (which displays as unknown) and mark the sensor as failed too. I don't like the idea of publishing 0 because we don't know that power is out, just that communication failed (cable disconnected, opto-isolator failed, etc. all would cause the same behavior).

I will try to put up a PR for this when I have time to test with my PZEM-004T to verify it works as expected.

[LevKos]

The “not available” state is logical when turned on, but if the measurement results have already been transmitted, then their absence, in most cases, will indicate a lack of power on the high-voltage part. Can be added with a condition like "if (this->voltagesensor->state > 0 )"

[snoopsbne]

I've done something similar except I use a timeout in the publish part of the code. I have several PZEM-004t's on the same modbus/uart and occasionally run into issues with CRC errors etc, so having it report 0 (or NAN) on the first "failure" is a bad idea (for me at least). I had a deeper look at a better way to do it, but my coding skills are not good enough to deal with modifying Modbus and the PZEM-004ts code. Ideally a timeout should be defined/optional on the Modbus side of things and if no value is reported then return 0 or NAN.

[LevKos]

Thank you. The presence of errors on the line is most likely caused by the circuitry. I implemented a circuit with 10к resistor то 3.3V and 1N4448 diodes on TX lines, 3 in parallel, polling every 10 seconds -checksums do not fall out. I think it is not advisable to programmatically handle physical problems.