This site is now in read-only archive mode. Please move all discussion, and create a new account at the new Victron Community site.
I have a SmartShunt connected to the battery. Downstream of the battery are various loads and a charger.
Since I installed it last week, the Max SOC has declined a little bit on every charge. I have every reason to believe the battery is properly charging to 100%.
Battery: Li Time 12V 200AH.
Charger: Li Time 14.6V 40A charger.
This chart is generated in HomeAssistant, it polls data from Cerbo via MQTT.
SmartShunt 500/50 is connected to Cerbo via ve.direct port.
Here are my SmartShunt settings.
Presumably that half-hour cycling is your battery bms doing what it does. This could involve quite low currents, even below the 0.10A Current Threshold. Try setting that to Zero, so even small currents (of the 'noise' level) are counted.
Unless of course if you wanted to consider the bms doings as noise, then you could consider raising it to get it to ignore the currents both ways. One way may be larger than the other..
Can you do any settings changes on the charger such that it is not trying repeated absorption every 30 minutes, it is quite unusual to keep cycling the charge this frequently. Your battery is being kept almost permanently in the upper voltage region which is not ideal. How about putting the charger on a timer plug so it only comes on once or twice a day for a few hours.
You get the award for solving this.
I had the charger disconnected as I moved to the final install last night, and SOC dropped to 95% before I reconnected the charger. SOC did reset after that charge, but not after subsequent charges.
Zooming in, the charger is doing something slightly different on a longer charge, which is falling within the Shunt algorithm. The ramp-down after the longer charge takes 5 minutes, but only 1 minute after the smaller cycle.
Going forward I will charge nightly using a smart-plug.
I cannot imagine how I would have figured this out without this charting ability. I probably would have blamed the battery...
I'm continually drawing 1.27a from the battery, which is why it's topping up every 30 minutes.
My question is, why is the SOC slowly declining each time it tops up? I am assuming that Shunt isn't properly detecting the full SOC...
100%, 99.9%, 99.8% - the long-term decline in the peaks after charging.
The shunt doesn't 'detect' SOC, it calculates it. Largely from Amps. Then 'doctors' the A as your settings direct it to.
I'm a little surprised that the bms switches like that, maybe something to do with the charge V?
You're talking quite a tiny drift here though, it's only the numerous cycles making it show. Even then, it's not large.
Current Threshold is still relevant. An increase in Peukert and/or Charge Efficiency will also push SOC upward.
Synchronisation comes from voltage, tail current and time.
A graph with a time period of a couple of those cycles, voltage, current will help tune the settings.
This might also be a graphing artefact. Generally the points shown represent a time period and are the average for that period. This would show in a more granular graph.
Interesting thoughts. Here are some more charts:
Current: (a)
Voltage:
and SOC:
And the Shunt settings have a Charged Voltage of 14.1 with a Time-to-go averaging period of 3 minutes. I understand that to mean the Shunt is supposed to reset to 100% SOC if voltage exceeds 14.1v for 3 minutes. The following chart shows it is over 14.1 for more than 4 minutes, from 4:07:31am to 4:11:46am, yet it doesn't reset to 100%. Is my understanding correct?
You have Tail set for 2.0%, that's <4.0A, and that must be satisfied too. It's possibly close, but the charge is shutting down, preventing it from getting there.
So I've tried some settings changes to correct for some of the suggestions.
1. Tail current, I've raised to from 2% to 5%, (10 amps). This should guarantee 3 minutes above Tail current. It did not help.
2. Charged detection time, I've lowered from 3 minutes to 1 minute. This should also guarantee sufficient time above thresholds. It did not help.
3. Current Threshold, I've lowered from .1 amps to .02 amps. It did not help.
4. Charged Voltage. I've lowered from 14.1 to 13.8
Max SOC after a charge is continuing to decline on each charge cycle. Actually, I think I've made it worse.
So...
Is this something I'm going to be able to solve in configuration?
I am thinking my next steps are to: Swap in a lower power charger, and I also will be increasing the DC load on the battery. Both of these should lengthen the recharge cycle.
Any other tips?
Your graphs show the charger cycling.
Interestingly the cycle period is increasing slightly, charger taking longer to kick in.
Also max SOC hits 99% each charge cycle. Where is the figure in the graph coming from?
I think what's happening is that it is/was synching to 100%, but the averaging over the increasing time between charge cycles is lowering the reported values in your first graphs.
The changes you made in the settings will have the opposite effect to what you want. What would confirm my thoughts further is two graphs, one for the SOC for 5 minutes including the charge cycle and accurate time scales. Then an analysis of the charge frequency over time.
I think your charger is cutting out, rather than switching to absorption.
The charger is definitely turning off, and this matches the mfg's documentation, it switches off when current drops below a threshold. They don't tell me what that threshold is, but from observation it's about 3.2 amps.
I'm hopeful swapping in a smaller charger and adding some additional DC load will help keep the charger in absorption mode.
What I'm confused on is why the Smart Shunt isn't resetting the SOC to 100% on each cycle.
The Victron documentation is mostly clear: https://www.victronenergy.com/media/pg/SmartShunt/en/all-features-and-settings.html
"As soon as the battery monitor detects that the voltage of the battery has reached the set “Charged voltage” parameter and the current has dropped below this “Tail current” parameter for a certain amount of time, the battery monitor will set the state of charge to 100%."
The setting "Charged detection time" specifies that time.
So why isn't that happening? It reads to me like the SmartShunt isn't working as documented.
That does make sense to me.
I don't think I can fix this in configuration.
So... Path forward.
1. I will increase the DC load, which will lengthen the charging cycle. (I'm bench testing this now, I'll just move forward with the install)
2. I will swap the 40 amp charger for a 20 amp one from the same manufacturer, again to lengthen the charging cycle.
If all that fails, I'll replace them with a Blue Smart charger which will maintain the absorption stage, which should avoid the negative current.
It'd be helpful if the documentation was more accurate.
The data in the chart isn't averaged. The most recent charts were pulled via Bluetooth straight off of the Shunt. The frequency is whatever the shunt is publishing via Bluetooth.
Is the shunt resetting to 100%? Hard to tell with this level of detail and such a small range on the readings. Something that may be affecting things is if there's decimal to integer conversion going on behind the scenes. Badly done it truncates instead of rounding.
If you haven't done it, it would be a good idea to bring the shunt software up to date, this may require two updates, one for shunt, one for Bluetooth. If that doesn't help, you can ask your dealer for warranty support.
Victron Connect is clearly doing some very aggressive rounding. They drop everything after the decimal, so this wasn't helpful.
I did pull data from Bluetooth using a python library, and and values matched up with the data I was getting via MQTT from the Cerbo GX.
