Heating setpoint maxes out at night

[ksahlen]

Running esphome-rego1000 on an IVT Premiumline HQ with Rego 3.9.0.

All works very well except for one bug(?), or at least a strange phenomena that may (or may not) be caused by rego1000.

During nighttime the heating set point is changed to the maximim allowed heat carrier temp, in my case 55°C. This regardless of the heat curve or the room sensor settings. As shown in the above graph nor the outdoor or the indoor temperature (note that this temp sensor is the one linked to the esphome rego1000 climate entity) justifies such a significant increase in heat demand from the pump. However, for two nights in a row at exactly 21:00 the heating set point is maxed out.

I have experienced this previously as well with the same setup.

Again, the only reason I suspect esphome-rego1000 for now is that I didn't experience this before I installed it.

Any ideas?

[jagheterfredrik]

Interesting behavior 🤔

The esphome-rego1000 doesn’t even have a concept of time, i.e. nothing is being controlled by time. The culprit must be home assistant, the pump or the thermometer.

Are you sure the pump didn’t do this before? Could it be that you didn’t notice?

I’d try to remove the indoor temp emulator (by removing sensor_id from climate) just to eliminate one parameter.

What’s your indoor setpoint and influence?

Make sure you don’t have any indoor temperature programs set in the pump.

[ksahlen]

Thank you for your quick response.

Noted that it doesn't know about time so agree that this is caused or initiated by something else.

I cannot swear for certain this didn't happen before but I cannot recall I have noticed anything before and I checked the logs/data from my now retired Husdata sensors and I couldn't see any similar behavior.

Will remove the indoor temp emulator to test.

Indoor setpoint is currently 10°C (summer house so I'm not there as much this time of year) and room sensor influence is set to 4.5.

Don't think I have any indoor temp programs but need to check when I visit the house next time.

[jagheterfredrik]

Something seems to be happening at 21:00 as you see weird behavior three days ago as well albeit not as extreme

[ksahlen]

So first night without an indoor sensor connected to the thermostat showed no strange behavior at all. Heating setpoint moved only very slightly in accordance with the outdoor temp. Will continue to monitor this a few nights before reintroducing the indoor sensor.

[jagheterfredrik]

That kind of fits with a program being active. If you only have remote access I assume you can read it using a number sensor for ROOM_PROGRAM_MODE

[ksahlen]

Added a number entity with ROOM_PROGRAM_MODE. It came out at 0 (zero) on reboot.

Don't know if 0 represents no program or active program so I will need to test this a bit.

[jagheterfredrik]

Ok, I just tested on my pump.

Apparently the program is set using ROOM_TIMEPROGRAM (Bosch engineers naming things 💯); it's 1 for program 1, 2 for 2, and zero for no program. The times are set using ROOM_PROGRAM_1_*: encoded as 16 bits. The day is split into 48 chunks of 30 minutes each, start time is shifted left one byte. E.g. 00:00 - 00:30 is encoded as 1, 01:00 - 12:00 is encoded as 536 (2*256 + 24).

[ksahlen]

Nice! Thank you for taking your time!

Changed "ROOM_PROGRAM_MODE" to "ROOM_TIMEPROGRAM" and the number sensor came out as "1".

Changed it to 0 and added back the indoor sensor to the thermostat.

Will monitor tonight.

[ksahlen]

So far so good with the ROOM_TIMEPROGRAM set to 0.

Will keep on monitoring a few more days but it definitely seems the issue was the pump and not the code in this repository.

Best,

[jagheterfredrik]

That validates my humble belief that the code is flawless. Great to hear, reopen if the problem persists.