How does a battery-powered HMKE work optimally with the ADA485?

Practical example of ADA P1 Meter and ADA485

In the world after net metering, it is increasingly important that the solar power system uses as much power directly in the home as possible and draws as little as possible from the grid – especially in the evening and at night. For this, having "just" an inverter and a battery is no longer enough: smart control is also needed.

In this article, we show how we configured a Deye SUN-xK-SG04LP3-EU inverter
with GreenHESS ADA485 and ADA P1 Meter devices so that:

during the day it first supplies the household consumption,
uses the surplus for battery charging,
and only then feeds back to the grid,
and in the evening the house runs primarily from the battery,
and the grid only supplies a few watts of “drip” consumption.

Goal: grid-saving, self-consumption optimized operation

The system logic in brief:

Daytime – when the sun is shining:
- the solar production covers the current house consumption first,
- if surplus remains, it goes to battery charging,
- and only the remaining energy after that is fed back into the grid.
Afternoon / evening – when sunlight runs out:
- the house consumption is primarily provided by the battery,
- there is no grid draw until the battery charge is above a safe level.
Safety limit – 30% SoC:
- in the test, we set the system so that it will only draw from the grid,
  if the battery falls below 30% state of charge,
- thus not discharging the battery too deeply, but still covering nearly the entire evening and nighttime consumption.

For an average household consumption, this means that often it lasts until morning until the sun shines again the next day – and meanwhile the P1 meter shows only a few watts of practically negligible grid “drip”.

The two key players: ADA P1 Meter and ADA485

ADA P1 Meter – “trusted eye” on the meter

The ADA P1 Meter connects directly to the distribution smart meter's P1 port (RJ12).

It reads data from the meter updated approximately every 10 seconds:
instantaneous import/export, power per phase, etc.
From this, it prepares well-structured, processable data,
capable of bidirectional MQTT communication and REST API connectivity,
thus the meter data are streamed to the HMKE.APP platform and are also accessible for the ADA485.

The essence:
we do not rely on the inverter’s own measurements but on what the meter sees – exactly what appears on the electricity bill.

ADA485 – “hand on the inverter”

The ADA485 connects directly to the Deye inverter’s RS485 (Modbus) port
(e.g. RJ12 to RJ45 cable).

It reads important inverter data (voltages, currents, power, battery SoC, etc.),
By Modbus register writing, it can intervene directly:
- maximum power,
- feed-in limit,
- battery charging/discharging modes,
- other parameters.

Both devices connect to the same local Wi-Fi / LAN network,
and in the background, the HMKE.APP logic coordinates them.

Test configuration: Deye SUN-xK-SG04LP3-EU + ADA485 + ADA P1 Meter

The specific successful test setup:

Inverter: Deye SUN-xK-SG04LP3-EU hybrid inverter
Battery: Deye-compatible Li-ion battery pack (with BMS)
Meter: distribution smart meter with active P1 port
Devices:
- GreenHESS ADA P1 Meter (P1 → data, MQTT/REST),
- GreenHESS ADA485 (Modbus RS485 ↔ inverter).

Connection:

ADA P1 Meter → smart meter P1 port (RJ12).
ADA485 → inverter RS485 port (RJ12 ↔ RJ45).
Both connect to the local Wi-Fi network and are accessible from HMKE.APP.

How does the system work together in practice?

1. Daytime operation – production, charging, minimal feed-in

During the day, the Deye inverter uses the current solar production in the following order:

Household consumption
- always supplies the instantaneous household load first (lighting, fridge, servers, etc.).
Battery charging
- if surplus remains, ADA485 maintains the charging mode for the battery through the Deye inverter Modbus settings,
- the system “parks” the extra energy into the battery.
Grid feed-in
- only the energy remaining beyond those two levels is allowed to be fed into the grid.

Based on the P1 meter data, it can be seen that import is practically zero at this time,
and feed-in only shows when the house and battery are already “full” of energy.

2. Afternoon / evening – battery as primary source

When the sun sets and the solar production decreases and ceases.
At this time, according to the set logic:

the house consumption is first supplied by the battery,
ADA485 ensures via the mode set on the Deye that the system
does not automatically draw from the grid until it absolutely has to.

3. 30% charge limit – protection and compromise

In the test, we configured the system so that:

the battery is not “squeezed dry” below 30% SoC,
at this level, the inverter gradually switches to grid import,
thus preserving battery life,
yet still with a good chance to run through the evening and night hours mostly from battery.

This value can of course be freely adjusted:

more cautious settings: 40–50%,
more aggressive: 20–25% if the owner accepts it.

4. “Drip” – why do a few watts of import still show?

P1 data typically show a few watts of import even when everything is supposedly running from the local system. This is completely normal:

there is always a few watts difference between meter and inverter measurement,
and some appliances' instantaneous start currents can cause short spikes.

The point is that overall the energy drawn from the grid remains practically at the minimum technical “drip” level, while the main consumption is covered by battery and solar panel.

Why is the meter’s P1 port important for this?

Although the inverter also measures, the official truth is always on the distribution meter.
The advantage of the ADA P1 Meter is that:

we see directly the data underlying the bill,
updated approximately every 10 seconds:
- how much we draw,
- how much we feed in,
- how the house load changes throughout the day,
and the HMKE.APP platform generates graphs and statistics from these.

This way it can be objectively verified that the Deye control with ADA485:

really reduces grid consumption,
and increases the self-consumption ratio in the household.

Experiences with the Deye + ADA485 + ADA P1 combination

During practical tests, we found that:

the system operates stably,
the Deye SUN-xK-SG04LP3-EU inverter Modbus control via ADA485 is reliable,
battery charging/discharging follows the set logic well,
and based on P1 meter data:
- daytime feed-in only appears when the house and battery are already “well supplied”,
- at evening and night the majority of grid import can be actually replaced by battery power.

All this is realized such that from the user’s perspective:

the system “just works”,
and the processes are transparently visible through HMKE.APP.

Who is this configuration recommended for?

The ADA P1 Meter + ADA485 + Deye SUN-xK-SG04LP3-EU setup is particularly suitable for those who:

have a hybrid inverter with battery,
want to minimize their evening and night grid imports,
consider it important to optimize with verified meter data,
and want a system open for possible future aggregator or dynamic tariff control.

Summary

The GreenHESS ADA P1 Meter and ADA485 pair not only “measures” and “logs”,
but actually can intervene in the system operation:

the ADA P1 Meter brings in official meter data via the P1 port,
and the ADA485 smartly controls the Deye hybrid inverter via Modbus,
so the solar panel + battery combination works as much as possible for the house,
and only a thin “drip” remains towards the grid.

If you are considering a similar configuration or want to fine-tune your existing Deye system like this, the ADA P1 Meter and ADA485 provide a ready, field-tested solution.

Comments

No comments yet. Be the first!