I am in the same situation and activated “Limit managed battery charge voltage”, default value being 55v. This was yesterday and battery SOC was already at 100%
Today, SOC reached 98% and not moving from there. Battery is 2 x Dyness Powerbrick DL5C, observed voltage before charging finished is about 55.4v. The CVL is 65.5v.
Is it required to increase the charge voltage at, say, 55.5v ?
Is there an automated system so all points are reached ? :
You shouldn’t need to limit charged voltage if the system is correctly set up and behaves correctly. What charged voltages did Dyness tell you tonset up with the model you have?
Out of curiosity, just disabled “Limit managed battery charging voltage” and MPPT spiked to max. output and energy (1500w) being fed to battery, voltage reaching 57v despite the CVL 56.5v
Being a remote location, quickly re-enabled the Limit charging voltage and battery started to dump 2Kw of power for several seconds.
Was really ackward and tense moments, also no idea what fits to “normal”.
I will straight up say i think Dyness have some work to do.
For feed into happen, the system raises the DC bus voltage 0.4 above (usually the programmed) battery voltage.
Remote sites certainly test the heart strength.
The fact is the system is programmed lower and battery is requesting higher.
Maybe try 0.4 above absorption to see if that helps feed in.
Is there a reason why you want to feed in from dc coupled not ac couple pv?
It is more efficient to use ac coupling and use dc coupling for the battery specifically.
Does the Dyness Powerbrick DL5C shift the CVL ever/at all?
The charger setpoint is certainly raised to enable them to generate but the export will draw the bus voltage back down again towards the CVL so the battery shouldn’t overcharge.
Yes, that has to happend so the current flow will take the intended path.
But, being BMS controlled, the Multiplus charger parameters should be ignored, right ?
The AC coupled always power consumers and feed-in the excess but sometimes, with less irradiation, I thought to enable DC feed-in to reach the maximum allowed.
Is it a good practice to use the MPPT only for battery and, if required, for “helping” the Multiplus ?
In this case it is set in the system as 56.5v by the battery - above the ve bus programmed at 55v and 54.5v. so there is that. If the ve bus now sets 0.4 above the CVL (so 56.5+0.4) there is the issue. The dyness needs to request less than max or what will lead to over voltage conditions.
Basically Radunre needs to work out the happy spot with the cvl in dvcc. It may be higher than 55.4v to allow balancing and sync to 100% and then feedback. {But low enough to prevent overcharge.}
Shall I understad that after disabling “Limit charge voltage”, followed by the voltage spike of 57v, waiting some more should lead to system stabilising to or lower than CVL 56.5v ?
Yes, but that isn’t the net effect. Feeding in from DC won’t result in the battery ending up charging to 0.4V above CVL else exporting would cease again.
The offset allows for a “tug of war” between the inverter and the charger - the chargers try push the voltage up and the inverters drag it down.
The net effect would be a battery maintaining voltage around CVL.
If you turn on exporting on a system that was not exporting, you can see this effect as the process starts, some charging happens, then it settles after 30 seconds or so.
Obviously if there is some voltage loss in the system this can affect the process.
iirc this is one of the test criteria for certification in the Victron world.
You would need to dig into the BMS data, some of which you can access via VRM.
If the CCL is dropping to zero at lower voltages it is more like due to cell imbalances, you would need to look at the advanced widgets for min/max cell voltage to see if this is happening.
If the battery is requesting a higher CVL and charging stops it is usually related to this.
The GX will also report if any packs are offline in the BMS details tab.
Depends. Typically in lithium you get very little capacity in return for much high voltages, so some tend to run a bit lower for longevity.
A few batteries use the preferred Victron method of charge control - voltage - and will vary the CVL at top of charge.
This is a much better system of control.
If you look at the battery page you can see the trends, remembering some are 15 cell batteries.
Unfortunately, Dyness do not have cell data in VRM and DL5C needs a Wi-Fi dongle (which I do not have) to communicate with cloud for extra information.
Installed some PowerBrick 14.33kW with built-in Wi-Fi and all cell data available in cloud. The same behavior for charging is observed, CCL drops to zero before reaching CVL, usually at about 54.96v-55.1v, on all installations.
You won’t be able to check balances when the pack has stopped charging and is at 53.33V, you only see the difference at the top charge voltages. You would want to see the readings just before it sets CCL to 0.
How are they being charged, I see they recommend no more than 50A per pack?
To me that sounds like either a balancing issue or just a bad BMS algorithm.
55.8V is a fairly common voltage to target.
Since yesterday afternoon, DVCC voltage limit has been disabled.
Up to now, system looks good. DC MPPT is being throttled to match feed-in restriction of 3kW. Fronius PV is sometimes also thottled because of line voltage over 253v, almost end of the line + other PV systems in the neighborhood.
