Fox does support Freeze_export via feed in first

[chalfontchubby]

Describe the bug
4538 sets support_discharge_freeze: False for fox inverters to match the fox_cloud.

This effectively disabled Feed In First which is basically freeze_export - but for the OP of #4207 and others this is an important mode. The distinction is for the OP that his export limit is lower than pv output so he cant export all his power. Predbat currently models freeze_export as clipping - but in the case of Fox (and I think my own sigenergy) if the export limit is exceeded the battery will charge.

Expected behaviour
Model the battery charging when freeze_export can charge.

Predbat version

[chalfontchubby]

4207 was in hindsight clear about the needs - I misread it and have been going in the wrong direction for some time. Apologies.

[WyndStryke]

Here's a debug showing some examples of freeze-export working normally, but it's not going to include clipping because my export limit is > my PV.

• 2026-08-17T07:00:00+0100 to 8:30 feed-in-first (freeze export) selected by predbat after an axle export session, SoC 51%
• then self-use (demand) to top up the battery
• 2026/8/18 16:00-17:30 manual freeze export, SoC 100% so will just export all solar

predbat_debug.yaml (1).txt

For completeness here's my apps.yaml, a plan, and a log from this afternoon, but they probably won't be of interest

apps.yaml (4).txt
predbat (8).log
predbat_plan (5).html

[chalfontchubby]

Thanks. I can force the export limit lower in that to trigger what I need.

[daviessm]

@chalfontchubby Do you have what you need or would you like a debug.yaml from me too? The weather up here in the north east of England is pretty poor for PV for the next few days but I expect to have one with real-world forecast PV exceeding my export limit at the weekend. Does it need to be from a specific Predbat version?

[chalfontchubby]

I can probably work with what I have and you shared previously. Debug yaml are pretty portable across versions.

Don't worry.

[chalfontchubby]

#4582 is up. I'll run some more tests locally - finally a pretty grey down down south as well.

[daviessm]

Actually the prediction model seems to have just re-run and updated with expected excess for today.

predbat_debug.yaml(1).txt

Anything over about 2.7 forecast PV in a slot will exceed the export limit - export limit is 2.5kWh per 30 min slot (5kW).

[chalfontchubby]

Looking back at the debug yaml from 4207 which is where we came in - that showed....

With the patch in place we model the battery charging better during freeze_export / fif so it becomes more attractive to the planner...

[chalfontchubby]

Your plan for today becomes...

[daviessm]

Looks promising! In the earlier yaml I'll have set force export early on to mitigate the clipping losses from later in the day.

With the second one I'd still expect some clipping later in the day once the battery has hit 100% but that's a separate issue (#4036/#4086).

[WyndStryke]

That looks pretty good to me

[kyfc2x28mk-afk]

I’m also now having this issue and has worked fine before. Using local modbus is control and not foxcloud ‘inverter does not support discharge freeze - disabled’.

I can upload debug and yank if we need?

[WyndStryke]

I’m also now having this issue and has worked fine before. Using local modbus is control and not foxcloud ‘inverter does not support discharge freeze - disabled’.

I can upload debug and yank if we need?

I don't think that the change is in a build yet. I'm staying on v8.48.6 until it is (you can select that version via the config page, near the bottom, but make sure you don't have auto-update turned on).