Heater state always "Off" although battery heater is active

[nixo2]

The MQTT state payload published by GroBro always reports "heater": "Off", even when the battery heater is obviously running.

This can be clearly seen by the battery temperature curve:
In the attached screenshot, the battery temperature shows a steep and sudden increase (orange circle), which cannot be explained by normal charging behavior and strongly indicates active battery heating.

At the same time:

the heater value in the state object remains "Off"

the heater status line (green line at the top, derived from state.heater) stays unchanged

---

Expected behavior

heater should report "On" when the battery heater is active.

---

Actual behavior

heater is always "Off".

---

Environment

GroBro: dev branch

MQTT (Mosquitto, TLS)

[robertzaage]

Thanks for the report! Can you please switch to the latest production release for a quick verification, if we have a regression in our code or something else changed? As far as I can tell nothing was modified on the modbus part or the corresponding register map.

[nixo2]

Thanks for the quick response!

I will switch to v2.1.0 and monitor the behavior over the next couple of days.

At the moment the battery temperatures are relatively high, so the heater is not being activated right now.
Once temperatures drop and the heater should normally engage, I will verify whether the heater state is reported correctly and report back with my findings.

[criticallimit]

is the Heater showing in the Shine App as "Heater on" but not in Home Assistant???

[nixo2]

I did not check the Shine App at the time in question, so unfortunately I cannot confirm what it showed regarding the heater state.

I am also not using Home Assistant. I consume and process the MQTT data directly myself (via Node-RED).

I will try to cross-check the heater status with the Shine App as well and report back.

[Christoph22]

Yes in Shine App Heater is on, but not in HomeAssistant.

[criticallimit]

@robertzaage
Can´t confirm that due to high temperatures here at my home, but is it possible the register for heating status is changed or never be the right one for the NOAH´s?
Maybe someone can read register 17 when heater is on in the ShineApp and check the value provided?

[robertzaage]

Sadly, I don’t own a NOAH - so no validation from my side. It’s possible that there is something off with the mapping. Best thing would be a small message dump taken while the heater is on and one while it’s off.

[nixo2]

I am still trying to verify GroBro’s reporting of the heater value, so I am currently taking a closer look at the data.

What I noticed is that the Noah seems to warm up the cells even under very poor light conditions (like today), despite only marginal charging power being available.

Despite nearly constant voltage and current, the charging power is intermittently reduced at first, which strongly suggests that part of the available energy is being diverted to heating. Later on, the heating appears to run continuously, which is also clearly visible from a strong temperature increase within a short period of time.

Unfortunately, I was again unable to check the Shine App output at that time. GroBro itself did not indicate an active heater (green lines) during this period, neither on the dev branch nor on release v2.1.0.

Starting tomorrow, I will also log the data from ioBroker, which should hopefully reflect the heater status as shown in the app.

[nixo2]

I continued observing the system today and again noticed battery heating under very low light conditions.

However, during this period the heater was not indicated in the Shine App either, and also not by GroBro.
This suggests that this is not a GroBro-specific issue.

It is possible that a heating process is only reported once more charging power is available, while low-power heating under poor light conditions is not exposed via the reported heater status.

For now, I will therefore close this issue.
I will continue to monitor the system, and if the heater status reported by GroBro becomes inconsistent or unclear again, I will follow up with new findings.

[robertzaage]

Thank you for reporting back. I don’t think we have a regression here and I can start to prepare a new (Christmas) release. 😊

[nixo2]

I need to provide another update, as there might indeed be an issue after all.

I suspect that Growatt may have changed something in the register mapping. My Noah units were previously running the last, but not the latest, firmware version, and heater activity may therefore not have been reported reliably in the app anymore.

Earlier today, I allowed an update to the latest firmware. After this update, heater activity is now clearly shown again in the Shine App (upper green line), collected via ioBroker from the Growatt cloud.
The first heating event today was not detected by the app when running the old firmware.

However, GroBro still does not report any heater activity:

• no output on the dev branch
• no output on release v2.1.0

The heater status remains "Off". The only reaction from GroBro is a change in bat_sysstate from charge to idle.

In the charts, this is visible as:

• upper green line (App): heater activity is shown
• lower green lines (GroBro dev / v2.1.0): heater remains unchanged

This suggests that the heater status is available, but GroBro may still be reading an outdated or changed register mapping.
This possibility is further supported by the fact that with the previous firmware version, the heater status was still shown correctly in the app short time ago, and is now only shown again after performing the firmware update.

Thank you for your continued work on GroBro.
Wishing you all a Merry Christmas, relaxing holidays, and a great start into the New Year!