They are unaffected by dvcc adjustments.
The Pylontech is requesting a 53.2V CVL 24/7
(But DVCC is only using 52.4, when it detects a 15S-Pylontech BMS, as 52.4/15 is already 3.5V)
Why is CCL 0.0A?
Normally when the jk bms goes into float mode the CVL is lowered to the float voltage (RFV (requested float voltage in jk-bms).
For now I have set the RCV (requested charge voltage (55.2V)) and the RFV (now 55.2V, prefered 53.6V) in the bms to be equal. So now there is no discharge when the bms goes into float mode. CVL always 55.2V.
RCV 55.2V - 100% - RFV 53.6V works well with DESS off.
CCL = max charge current in jk-bms -/-5%. (I have 4xbms, 4x100A-/-5% = 380A). This value only changes if there is a problem in 1 of the bms. So 1 bms blocking charge → -/-95A etc.
DCL = same logic as CCL
1 Like
Because SoC is 100%. That’s just how the Pylontech handels to say “I’m full, thx”.
Vice versa, when it reaches 5% SoC, it will request 0 DCL.
For a new test I wanted to turn off “AC-coupled PV - feed in excess”. This doesn’t do anything with DESS on.
If I turn off DESS it works just fine.
DESS - AC-PV feed in disabled → still AC-PV feed in (bug?)
DESS off → AC-PV feed in disabled → no AC-PV feed in (just self consumption)
“Maximum feed in” works with DESS and without DESS.
For my test I want just no AC-PV feed in.
I noticed that too. Guess DESS overrules ESS settings is some way.
Otherwise just turn off your PV system completely so the batteries are only charged by the grid. I won’t effect the issue we’re experiencing.
I can do that. (grid set-point -40Watt → 1000Watt (to supply loads)
I tried it. (turn off SolarEdge)
It doesn’t work that way. I have to think about it for a moment. Without PV, the batteries are not simply charged from the grid. This is also the reason why I wanted to turn off the AC-PV feed, so that there was still PV to charge the battery. Or I should set the minimum soc to 100%, but that doesn’t seem like a good idea to me.
btw
I have now set RCV and RFV both to 3.45V. The jk-bms 100% (after 1h “absortion time”)logic still works fine and CVL always at 55.2V. Not really a solution, but certainly temporary.
You should be able to run DESS without PV. DESS doesn’t require a PV system.
Otherwise try ‘keep batteries charged’ option
First I will try (maybe tomorrow) to limit de SolarEdge (limit active power) to 10% (so max 1.7kW).
With the 55.2 voltage you just hit the balancing point. 3,45x16=55.2 Volts.
I guess you also got some issues with your SOC on the JK BMS too!?
I limited the charge current on 50 Amps and the voltage so it slightly ticks the 3.6Volts otherwise the rest of SOC won’t be reachable. Voltage i set is 56,8 as hard limit. The Victron is slightly over that voltage so it all runs good.
That won’t be an issue as it will remain at 55.2V for an hour so enough time to balance. 50A charge current is quite low for my system. I’m currently at 100A charge current.
But al this is not a part of the issue that with DESS on it will discharge at full capacity to go to float voltage.
Could this have anything to do with it?
Apparently AC-PV feedin switch has no effect with DESS on. (switch on or off, PV always feed in)
Maybe DC-PV feed in switch also has no effect either with DESS on?
I came across this, maybe it has something to do with it
I haven’t worked with node-red yet.
I think we (@dick) found what causes the big battery feed-in/discharge (jk-bms float mode CVL 55.2V -1h-100%-53.6V)) in DESS mode.
- I noticed that AC-PV & DC-PV feed-in enable/disable does nothing (PV feed-in always enabled) in DESS. Why? Maybe it would be nice to have the option to disable (AC)-PV feed-in.
- Feed in overvoltage, dbus path: /Hub4/DoNotFeedInOvervoltage, type enum (Mode: both)
DESS → Feed in overvoltage → forced 0 = YES
ESS DC-PV feed in disabled → Feed in overvoltage → 1 = NO
ESS DC-PV feed in enabled → Feed in overvoltage → 0 = YES
So I expect that in ESS mode the big discharge/feed-in behavior (jk float mode) also occurs when DC-PV feed-in is enabled.
For now I have set the float voltage (RFV) and charge voltage (RCV) exactly the same (3.45V*16) in the bms. Resulting CVL always at 55.2V.
1 Like
Yes, DESS only knows a “general FeedIn yes/no” option (“Are you allowed to sell back to grid?”), it is not aware of a difference between AC/DC PV.
Based on “Are you allowed to sell back to grid?” it will basically either enable both or disable both.
That is interesting, DESS is not setting that setting anywhere. Will see, If I can figure out where this happens, and why.
Ah, that settings is a consequence of DESS saying “Feedin is allowed”.
Will discuss with the team, the fine difference of “Feedin in Excess-DC-Solar” and “allowing to discharge overvoltage [due to CVL-drop]” is the cutting edge here.
When DC-Feedin is allowed, the MPs need to apply a Terminal-Voltage slightly bellow the MPPTs to discharge that power. So, if the CVL is lowered (lowering the MPPTs ChargeVoltage) the battery will inevitable be ABOVE the MPPT and ABOVE the Multiplus Terminal-Voltage, causing it to discharge.
The absence of DCPV doesn’t change this, instead of V(MP) < V(MPPT) = CVL < V(Bat), it’ll just be V(MP) < CVL < V(Bat)
I am not sure about this..If I turn DESS off and DC-PV feed-in is also off (=Feed in overvoltage → 1 = NO) then I think nothing happens when CVL 55.2V-1h-100%-53.6V. At least my system has been running for about 2 weeks without any problems.
I am still new to the world of ESS & DESS.I find it confusing that in DESS mode I can turn on or off different settings (for example PV (dc or ac) without it doing anything. Maybe it is good to know which settings&limits are forced/overruled by DESS. (preferably “Grayed out” in the menus.)
For now DESS is running in trade mode. I have set the float voltage and the charge voltage equal in the bms. (CVL always remains the same). This seems to work reasonably well.
Yes. Then The MP wont to try to get rid of the “overvoltage situation” arrising from a dropping CVL. Because it is disabled to do so.
From a technical perspective this is the way to go for LiFePos. Mind the JK-BMS can be used with any cell-type, so obviously has a lot of configurable options that may only be required in certain situations.
Lead-Acid needs to be charged OVER their nominal voltage. So, to prevent damage and account for voltage drop, floating is required. LiFePos are charged AT their maximum Voltage. No need to reduce that, current-flow will just physically come down to 0, when fully charged.
I believe I’ve a similar problem as described/discussed here. Would be great to understand why this is happening and how to resolve it.
When my batteries are fully charged, CVL=55.2V, for some reason the configure 1h float mode doesn’t kick in. After a couple of minutes when the 55.2v is reached the systems start flipping into discharging, charging, idle, etc. etc. There is enough solar power to keep the system in absorption mode for the configure 1h. And after this I would expect the voltage to drop to the 53.6V (RFV). But this is not happening.
My system: dunning DESS, external (BMS) controlled (DVCC enabled), JKBMS V15 latest firmware, two batteries (2x16kW), Multi RS Solar and 2x MPPT.
The next screenshots are taking within one minute!
@dognose Hello, would you also be so kind and check what happens on the victron when combined with JK-BMS in DC Coupled systems. It happens only when selling to grid is enabled. ESS continous switching between charge/discharge when battery full and negative grid power
It is well documented in few threads as well. Battery keeps charging and discharging with approx. 1kW of power when at max requested voltage. It seems that victron just ignores maximum charging voltage requested from a battery and increase DC bus voltage higher. SVS shared voltage does not help. Setting maximum voltage (CVL) in victron does not help.
My personal favorites for cause are:
- CCL from JK BMS is still constant even at max charged SOC. But when I tried to limit CCL in Victron it did not help either.
- JK BMS does not turn off charging mosfets at 100% SOC and with higher DC bus voltage it just naturally absorbs power → Victron then sees higher voltage than requested → discharges at maximum pace to compensate → charges → all over again …
This does not happen on pylon batteries, but I do not know if is caused by mosfet deactivation, CCL limit or anything else.
Also it looks like whole Victron battery management is not prepared for this type of BMS behavior - no CCL changes, float stage, mosfets still on… as we can see in this thread
I’m not all to familiar with the JK-BMS, but quickly googling for “JK BMS jumping charging discharging” brings up several threads about this. In case of the JK BMS it seems to have to do with bad voltage / current calibration.
So you should do some research on how the calibration should be performed in a correct way and give that a try, in case you have never done that.
Well, when the battery reaches CVL, the voltage difference should be 0, and there shouldn’t be any current flowing anymore. Still seeing a charge therefore indicates that the voltage displayed / reported by the BMS is not accurate.
For many BMS (passive) the behaviour of charge/discharge around 100% is completly normal, and related to battery balancing, because everytime a single cell is overshooting it’s intendent maximum voltage, the BMS has to discharge all cells in order to prevent overcharging for that single cell.
The JK-BMS uses active balancing, so it does transfer energy between cells while balancing. Yet, doing so is not “lossless”, so that transfer would lose some energy, resulting in a overall reported discharge as well.
What values are you guys using for the Start Balancing Voltage and Voltage Difference Threshold in the JK-BMS settings? How is the actual balancing state reported by the battery?
Eventually your cells are a bit unbalanced, and a single cell is reaching it’s critical voltage, before any balancing kicks in, therefore the JK-BMS has to switch to discharge, to bring that cell down bellow a safe voltage?