Aaah Auto-recharge. I know the bastard who implemented that part quite well… unfortunately 
So one possible way to give people what they want is to have the option to disable auto-recharge (default to on). That way, you simply set the MinSoc to 95% (no matter how low they are) and all PV will be used to charge the battery, with self-consumption kicking in at 98%.
We always come up with great ideas and plans, to sort something else out. That is just who we are here in SA/Africa.
If the SOC slowly climbs during the day, the problem is substantially mitigated of drawing big loads for extende periods in hot summer months.
Because no batts have been used, enough panelling, and me increasing the SOC manually just in case, the temps are ok today.
3:40pm, batts are ready for evening, geyserS are at 65deg, dishwasher is done and the min SOC is set at 95%, it is sommer really good.
Scheduled Charging still kick in at 4:30pm, to get it to 100% SOC by evening.
Just a thought.
How about variable max inverter power to match available PV ? and changes Min SOC as the battery charges
A few less cronjobs @TheTerribleTriplet
How does one determine the PV potential?
I know @Paul has done that with success if I recall, but it is tricky at best.
And it is only one Cronjob I need … 
Upping the SOC the last thing to sort.
Haha! I have that same slot! 1630 to 1900. Gets everything nice and balanced and covers the large loads in the kitchen. Don’t want to run ovens off my batteries. Happy to just supplement with whatever PV is available.
If it will help some one 
.
I am using “Limit inverter power” form ESS.
Then I track my battery watts during day time only. If the battery watt goes into a negative I set the “Maximum inverter power” to 85% of the max solar output , in my case that seems to be the point where the batteries will still charge a bit and the rest of the load will be carried by the grid.
If the batteries is charging more than 200w then I increase the “Maximum inverter power” to the current value of the solar output this ensures that you use the maximum available solar available.
At night time (if solar watt is < 100) “Maximum inverter power” is set to a fixed value , in my case 350w. This ensures that the batteries is discharged at a constant rate through the night and you end up with the desired SOC in the morning.
Hope this all makes sense , it does in my mind but again I have been running this now for almost 2 years with no problems , yet …
It is part of the home automation. Running on Node Red.
But I am sure it could also be run though Python. Memory serves me the latest Venus GX has Node red running on it already so it should be fairly easy to implement.
You can have a look at the code but keep in mind I am not a programmer and this might only make sense to my self but if it helps its cool:
settings = global.get(“maxDischargeSettings”)
minEss = settings[0].Data1 // Min Battery Discharge value for ESS : 350
MaxBattDc = settings[1].Data1 // Battery Discharge value before Activation : -350
factor = settings[2].Data1 // Factor by with the ESS vaule is set - this is to ensure that the pannels match the dischrge : 0.09
MaxBattCh = settings[3].Data1 // Maximum Battery Charge - 50 w
if (sol <= (minEss+(minEssfactor))){
newMax = minEss;
}else{
newMax = sol; //-(solfactor)
}
if (gridSw === 1){ // Confirm Grid is On
if(maxstatesw==“ON”){ // Check if Switch is ON
if (batw <= MaxBattDc && maxstate > minEss || maxstate < minEss ) { // Battery is Discharging and ESS is still to high
msg.payload = (newMax0.1).toFixed(0);
return msg;
} else if (batw >= MaxBattCh ){ // Solar is enough to sustain the carry of the house load
msg.payload = (newMax0.1).toFixed(0);
return msg;
}}}
What happened to the project where you used that teeny solar panel to ascertain solar potential?
There is support for such a sensor already. It’s called the IMT Si-RS485TC Series. Uses a normal RS485 cable. At the moment the info is not used for anything (other than logging it), but a sufficiently determined developer can use the info to do something…
Jip, know of them … and their price tag.
Paul made a clever move, was discussed on another forum in detail, and it was quite affordable compared to the Si-RS485TC Series.
But then I stumbled onto this one: https://www.geewiz.co.za/sma-inverters/22831-sma-sunny-sensor-box-11gb-measuring-system.html
It died a quiet death , that was until you resurrected it…
The problem that I had was the shunt to measure the current. The panel was just to small the get a consistent reading. You could get reasonable voltage readings and if there is some clouds it will react but I was looking for a “bell curve” on a clear day to start making comparisons. I am still sure you could get much better results with a 12 panel but so far price tags on those was just a tad high for my liking.
I do like the SMA sensor - must have a proper look at it as I see you can also measure wind speed with it…
Indeed! If we can disable auto-recharge, I can keep it on “Keep Batteries Charged” during sunlight hours (warm parts of the day I’d prefer to not use the batteries otherwise they heat up more) because for some reason or another, when I use the “scheduled charge” and my batteries becomes fully charged, the MPPTs doesn’t ramp up to power my loads. This works perfectly on “Keep Batteries Charged”.
I know the workaround is to limit the inverter charger to 0, which I will do… as soon as I get the time to turn my home computer on again… turns out the effect a baby has on available time is to absorb it all, doesn’t matter how much you had before.
This is one of the things I’ve heard before that makes zero sense to me. Both KeepCharged and scheduled charging use “overvoltage feedin” to make sure the loads are covered, or that excess is fed into the grid (for those who do that).
Don’t forget the motor in an electric weighs far less than an engine and all its accoutrements, so its a wash.
I’ve seen this many many times where the MPPT only charges the batteries, not powering the loads, with Keep Charged, with Scheduled Charging and when I up the SOC with a Cronjob like I used to. MPPT goes into a kinda “hibernation” and just charges the batts, and that is with full potential available from the array.
To “kick start” the MPPT:
Drop Min SOC from 80% back down to 75% for example.
Or take it out of Keep Charged then back on.
Or disabled Scheduled Charging, then back on.
Can I replate on-demand … off-course not. Who wants it easy!
Something has changed and I cannot put my head on a block on what that is as I have swapped from lead-acid to Lithium in the same period I suspect improvements were done on the Venus.
I run the “Latest Release Candidate”.
Because I’m told each time, “kyk na jou battery spanning”, I’m first going to focus on replacing the Daly with a new BMS for the Daly has to run on much lower than ideal Bulk and Float voltages to not cause a DC Ripple. Yeah, I get all the blerrie goed that no-one else does.
In the mean time I’ve started to use this command again:
dbus -y com.victronenergy.system /Debug/BatteryOperationalLimits/SolarVoltageOffset SetValue 0.3
It seems to have made the issue better on my Daly BMS and settings. But I still have to keep an eye out that the MPPT does not decide to just charge the batts, and not power the household loads.
Case sample - and I DO admit that I may be missing a salient point here!
Here is the system on Optimized without Battery Life:
Then I switch over to Keep Charged:
So in effect I cannot use Keep Charged Option.
So I need to up the SOC manually to not strain batts during peak solar production, if it is on a SOC of 76% then I can set Min SOC at 75%, 81% set SOC at 80% to skip the batts from being used during the hottest point in the day, the batts tend to heat up quite a bit if used during that time, in summer.
Winter makes no difference.
I keep on mentioning upping the SOC.
This is what I see when I up the SOC, in THIS case, from 60 to 65%
Min SOC at 60% - all fine and dandy:
SO I up the Min SOC to 65% - Eskom steps in - I must wait till it is 61%:
There are many reasons this kind of thing happens, and the things that affect it are battery chemistry, battery capacity, cabling, etc etc.
The one thing that always gets people, and despite me saying this over and over and over it just does not sink in, is that SOC and voltage don’t correlate, definitely not on lead acid, and at very high MinSoc settings, also not on lithium.
So as an example, you get people who do the whole load-shedding fever thing, usually people with smaller banks. They set the MinSoc to 95%. Battery discharges to 95% and at this point discharge is disabled (the system considers the battery discharged). No energy will be taken from the DC bus until the battery is charged back to 98%. But… a very full battery (even a lithium one) will easily rise in voltage easily. So quite soon the solar charger has it up to absorption voltage and starts to limit, but because we’re still below 98% SoC none of the extra energy is used for loads. Also, because the battery is accepting really little charge current (even lithium sometimes), it’s charging slower to 98% too, leading to just an overall wastage of energy.
Also something people maybe forget. The MinSoc setting is still used during scheduled charging. If you set the ESS MinSoc to 90%, and your SOC is below that, then the Multi will stop taking power from the DC bus. Even if you enable scheduled charging.