question

Do all BMSs talk to the units when you have the maximum 5 parallel battery modules? I really believe this is overkill. Most modern battery designs only have one BMS-equipped unit anyway, afaik.

Maybe u check for another bms, so a secondary bms is connected to a primary bms which conected to victron, like batrium.

@thanar @hardy tuner Yeah, one BMS will be master and the others are will be slave in such parallel BMS configuration. Grouping 5 batteries with one master BMS and powering up one busbar seems like another option.

The installation example I shared was similar to my setup. 20 pieces of 25,6/200Ah Victron LFPs are used. But there is no visible BMS, this frustrated me.

@Melih Manisalıgil ...I am not sure about the battery configuration requiring 5 BMS units. The VE.Bus BMS V2 can handle 5 parallel strings so your system will be fine.

The size of the system eliminates the use of a shunt so the Quattros could be configured as the Battery Monitors. No "DC System" exists so that will be accurate.

One VE.Bus BMS V2 can handle the entire bank. VE.Bus control from VE.Bus BMS V2 to the first Quattro and daisy chain VE.Bus to each other Quattro. Connect VE.BMS V2 to CerboGx. Set up 3 Quattros on 3 Phase configuration. Set up battery bank to terminate to a large Busbar with individual series string fusing. It has serious DC fault current capability so obviously busbars must be well segregated and well designed.

It's not clear to me that the limit is (just) BMS-related. The Lithium Smart battery manual states:

"Connect a maximum of 5 batteries or battery strings in parallel", and "Connect a maximum of four 12.8V batteries or a maximum of two 25.6V batteries in series".

This would limit you in a 12V system to 5 batteries in parallel (5 total); in a 24V system composed of 25.6V batteries to 5 in parallel (5 total), or if composed of 12.8V batteries, to 2s 5p (10 total); in a 48V system composed of 25.6V batteries to 2s 5p. or with 12.8 batteries to 4s 5p (20 total).

There is no mention of building a bigger bank than these limits permit, although the example quoted in Romania would seem to considerably exceed these limits.

Why the 5-parallel limit (independent of system voltage)? The only thing I can think of is a concern about fault currents if a battery has a serious internal failure (which fault current should be limited by the individual fusing required for each string).

As I understand it, the BMS wires from the batteries have 3 conductors, and I believe there is normally continuity on each wire through the battery. The battery apparently signals an alarm condition by creating an open circuit on the appropriate wire (load disconnect, charge disconnect, or pre-alarm). Assuming the switch in the battery is a dry contact, I don't know why there would be a limit of 20 on the number of batteries that can have their BMS cables daisy-chained. Perhaps they use opto-isolators. Certainly the BMS wires must be galvanically isolated form the batteries to permit daisy-chaining of BMS cables from series-connected batteries.

It would be nice if Victron would clarify these various limits!

But assuming you can build a bigger bank (than 4s 5p 12.8V), and that the BMS is limited to 20 batteries, I would think you could use multiple BMS's, create some simple logic circuitry to combine the three electrical outputs of each BMS to make overall signals for the three alarm functions, and feed the overall load and charge disconnect signals into the Quattros using the Two-Signal BMS assistant. AFAIK, there is no way to do this using the VE.Bus outputs of the VE.Bus BMS.

Re the Romania video, you can see the daisy-chained BMS cables on the top row of batteries, but not where they go. They could have extensions to the equipment rack, etc. The Victron BMS's (Small or VE.Bus) are quite small, and could easily be hidden out of view.

Yes, I'm aware of that training video. That's what led me to believe that the BMS may be signaled by a simple open circuit inside the battery between the corresponding wires in the two BMS cables.

I suspect that the (external) "BMS" units Victron has are therefore rather "dumb". The real BMS is inside the battery, as part of the balancing and monitoring electronics. When you look at the VictronConnect app, all the BMS-type information is already available inside each battery: the individual cell voltages, the temperature, the alarm thresholds, etc., and the alarms derived from these.

My guess is that the only thing communicated by the battery through its BMS cables is those three alarm statuses, signaled by a simple open circuit in an otherwise continuous wire. (That is, there is no sophisticated digital communication mechanism, like CANBus being employed.) The external BMS then has no knowledge of individual cell statuses, or how many cells there are. Nor does it know which battery in a daisy-chained group is causing the alarm. It can't distinguish e.g., whether a high cell voltage or a high temperature or even a low temperature is causing a "charge disconnect" alarm. To get this information, you need to use the VictronConnect app to get the detailed data via Bluetooth from the batteries That is where the sophisticated communication occurs with the battery. Again, this is just my theory based on what information I've been able to find. I'd be interested to hear any further thoughts on this subject.

BTW, your argument about cell count fails to explain the maximum limit of 5 12.8V batteries in parallel - that is only 20 cells. The 80 cell limit you proposed would permit 20 12.8V batteries in parallel.

I should also amplify my thought about the paralleling limit. When you have paralleled batteries, it is not just a question of how much current the other batteries in a paralleled group can deliver into a failed battery. There is also the question of the available energy. A fault may not cause enough current to flow into a failing battery to blow the fuse for its string. The more batteries there are in parallel, the greater the possible energy delivered into the fault at currents less the the level to rupture the fuse. Again, more speculation!

I did a little more research, and found the following photo: https://shop.pkys.com/assets/images/Inside.png

If you zoom in on the top left of the circuit board, you can see what I believe are three opto-isolator chips, two next to each other in the extreme top left, and a third just to the left of where the upper BMS cable is soldered to the board.

So I believe each set of BMS wires (brown, black and blue) connects across the output side of one of these opto-isolators. The opto-isolators will be normally on, turning off (open-circuit output) for a fault condition. Thus when you daisy chain many batteries, the on-state voltage drops of the opto-isolator outputs will add up in series, and eventually the total voltage drop will exceed whatever voltage the BMS applies across the series string to detect whether all opto-isolators in the series string are turned on.

So I further believe there are two quite separate limits here:

(1) the maximum number of opto-isolators in series (i.e., the maximum number of daisy-chained batteries) that the BMS can support;

(2) electrical safety limits (or whatever) that set the maximum number of batteries allowed to be in series and parallel.

The latter limits are very clearly stated in the Lithium-Smart manual, and the maximum number of batteries allowed in a bank is for a 4s 5p 12.8V configuration, with 20 batteries total.

Since this configuration is supported by the Victron BMS's, we know the limit for item (1) above must be at least 20 batteries. Possibly it is more, but Victron would not appear to have a reason to support more than 20 as that is the maximum bank size they permit.

But I think before you worry about the BMS setup, you first need to talk to Victron about whether you can build a single bank of the size you described, as it involves significantly more paralleled strings than Victron permit according to the Lithium-Smart manual. Obviously you could build it, but would they honour the battery warranty?

Hi, there are some updates on this project.

Firstly, the MG Energy LFP280 batteries will be used. MG also limits the batteries up to 5 parallel only but you can connect to the same Master LV up to 96 batteries which is more than necessary.

• 30x MG Energy LFP 280
• 1x MG Energy Master LV 1000A
• 3x Quattro 48/15000
• Cerbo GX and GX Touch 50

I'm not familiar in detail with MG equipment, but it's certainly a fine choice from what I know.

One suggestion: consider dividing the bank into two halves, each with its own Master LV. I know another owner with MG batteries / Victron inverters. He created such a setup, the point being that a major battery issue (e.g., a cell problem) or trouble with the controller that causes a Master LV to disconnect does not shutdown the whole bank.

I don't believe he has had an issue where a BMS disconnected, but I have heard of others having such issues and have experienced it myself, with a different brand of battery having a fully internal BMS and an external disconnect switch.

I think a possible cause of such issues is charging a battery with poor cell balance (this was apparently the situation I experienced). As the battery voltage approaches the absorption level, one cell may have excessive voltage, leading to the BMS calling for a "stop charge". However, possibly the cell voltage overshoots a little and hits the BMS threshold for opening the disconnect switch.

Another cause (which does not involve any equipment failure) could be misconfiguration.