reinhard-brandstaedter
commented
5 months ago I haven’t seen any need to trigger things on hub change. Usually this is reflected in DTU changes or smartmeter changes.
Closed mavo closed 5 months ago
reinhard-brandstaedter
commented
5 months ago I haven’t seen any need to trigger things on hub change. Usually this is reflected in DTU changes or smartmeter changes.
mavo
commented
5 months ago But if the demand of house does not change (a lot), its not triggering anything. Same for DTU, does not change as its stable fed by the hub/battery.
This can happen with no direct connected panels. And in those cases its actually required. Is it actually hard to incorporate something like this?
reinhard-brandstaedter
commented
5 months ago Not difficult, but usually too many triggers (that are linked) creates unnecessary steering noise. Already today a change happening on the DTU likely triggers a change on the smartmeter as well. A change on the hub usually also (in combination with min charge power) usually also triggers a change on the inverter. And as a backup there is still the regular interval trigger.
mavo
commented
5 months ago Not difficult, but usually too many triggers (that are linked) creates unnecessary steering noise.
Understood. However, for me (and most likely for all having no direct panels), the DTU does not trigger too much, as it only changes when the SMT changes due to changes in demand. Maybe the app can recognize when there is no direct connected panels...?
And as a backup there is still the regular interval trigger.
Which is somewhat slowish.... 🤔
reinhard-brandstaedter
commented
5 months ago The challenge here is that the hub usually reports it's solarpower very frequently, the more change the faster actually. So if we would want the hub to trigger the limit function as well we will need some kind of fixed time window (not a moving avg) to determine a significant change, otherwise it would just be a a short-interval of minimal changes (and we certainly do not want to trigger on those) See this reported data:
mavo
commented
5 months ago we will need some kind of fixed time window (not a moving avg)
Which is tricky, as it might happen that the hub does not report anything if it does not change, such as it was the case for me when the shelly did not update with the fixed interval? As far as I get it, those fixed time window is also used for the usage reported by the SMT, isn't it?
I mean, I get that this is not THE most urgent issue and also it does work and only happens on day time, so you won't feed energy to the grid (at least I did not see this yet and also cannot come with a situation this happens and can be prevented by this trigger). On the other side, it would be great to have this, as in my config the only trigger thats left is the SMT, and if this does not change I need to wait for the regular interval. I was thinking about some sort of script which changes the DTU ever so slightly in intervals in case there was no change in the inverter reported. However, not only feels wrong, but also is as this would interfere with the whole sf-control eventually 😵
reinhard-brandstaedter
commented
5 months ago No, the smartmeter uses a moving average window of 10s, since there are typically up/down movements on the smartmeter this works well for triggering adjustments on change. (because the change is usually not +2W every 2 seconds, but rather something like +60W or more instantly).
I'm running with no direct panels for a couple of weeks now and I haven't seen any issue with this setup. You are not loosing any energy. If it doesn't go to the house it goes to the battery, and since you are running with a min_charge level of 0 (I still don't see the benefit of that) it will be adjusted soon enough. You are just getting a bit of charge power for max 2 mins.
mavo
commented
5 months ago As said, it's not really an issue. I was under the impression the hub would trigger as well. It does not, it's fine.
Regarding the 0w min charge - why not. For me, feeding the house has prio and it does not actually matter when the produced energy is used. Isn't it? For me, why I think it's good to do it like this is that in summer months, I think I might end up with a battery being full earlier in the day. So my goal (and summer might prove me wrong here) is to feed in as little as possible, that's why I prefer home consumption.
reinhard-brandstaedter
commented
5 months ago The latest master does have a trigger on hub solarinput change. Try if this changes the steering cadence for you.
mavo
commented
5 months ago You are an absolute hero! I'll try, but I probably won't have a result tomorrow, as it's time for a full charge cycle :-)
reinhard-brandstaedter
commented
5 months ago HUB triggers limit function now correctly
Hi @reinhard-brandstaedter,
I am not sure if this was like it before, I only started to notice today (I changed to latest master yesterday in the morning ours). For some reason (maybe even wanted) I do not see any "hub triggers limit function" calls, which I would expect when the hub input changes. The result for me is, that when the sun is out, household demand is matched, all good. But then a dark cloud happens to appear, as the hub is not triggering any update, my battery is then used to keep the inverter at its desired output rate (or the other way around, battery is fed instead of home).
Here are my startup logs:
And here are some logs from the app running:
From this log you can nicely see, that the radiation did increase, but nothing was triggered. At some point, some trigger happen and fix the "offset", but its not ideal.
Is this how it should be? If yes, please just close this ticket with a comment so that it does not clutter the other tickets.