Who should be in control?

[s/v Jedi]

Originally Posted by SaylorMade

I just found this informational article from a company that has been dealing with this stuff (batteries in general) for 50 years, and has over 1,000 employees.

So I kindda guess that we can put some trust in their statements, but I have admittedly not researched them further than this.

They talk about the need for balancing and also the difference between passive and active balancing.

I think most of the "cheap" batteries does active balancing these days.

https://www.power-sonic.com/blog/lif...ell-balancing/

I don't understand why some find it very complicated to let a BMS balance your cells?

I wrote that I think the BMS isn’t the best option because an active balancer has proven to be 10 times as good in balancing than the very limited balancing that BMS’s do.

The balancer linked below is very popular. It has a little soldered jumper labeled “RUN” that you can replace with an on/off switch or let the BMS control that when it has the output for it.

https://www.amazon.com/Cywhrvzsf-Equ.../dp/B09G74WDN5

[K1MGY]

Originally Posted by SaylorMade

So, in a few threads here a side-topic has arisen, about wether or not the charger should be in control of when to stop charging a LFP bank.

IMO a well crafted BMS acts as a safety gateway. The transition from constant current to constant voltage happens as a consequence of expected cell/pack response along the charge curve. Chargins stops when current flow stops, and this happens when the pack voltage has been reached. "Control" is up to the charge curve. Supervision is up to the BMS. Supplying current, at a specific not-to-exceed voltage, is up to your charging source/s.

In my setup (280 A/Hr/cell from two 4s-2p packs in parallel, each having an independent BMS -- enough to power a small town!) there are three charge sources. An inverter, a solar MPPT, and a DC-DC converter driven by the engine alternator. I have set all three to supply a specific voltage (which the manufacturer calls "bulk") and leave it at that. There are other settings for equalization and float. I set these (presently) to the same specific voltage, nullifying their control functions. Victron (in my circumstance) apparently created their boxes with some charge characteristic management in mind. This may be useful for certain types of cells (AGM cells with carbon, called "Firefly" come to mind as these were originally in BABU), but I have questioned the utility of these settings as pertains LiFePo.

My view: Give the cells a specific not-to-exceed voltage, let the love flow, and let the BMS do its protection thing.

Some fine tuning is still in the works here, as I discover what's best for the cell lifetime, so charge voltage may change eventually.

On balancing? Start with cells within 10 mV or so. If impossible, use an active (capacitive) balancer. There are many good ones out there. I have found this to be of great value as it has eliminated pack shutdown due to single-cell overvolt fauls.

Hope this gives some good thoughts.

[Jef & Marin, Netherlands]

I understand that it is best to not charge an LFP bank to 100%.
You might want to charge it to a SoC (State of Charge) of for instance 50% or 80% or whatever, depending on how the bank's capacity and the next charging (for instance by the sun tomorrow).
When your desired SoC is reached, you want to minimize the current going in or out of your bank.
Both the SoC and the current of the shunt are known by the battery monitor.
This information should be used by a (any) controller to fix the output current of your chargers.
Note that at 50% or 80% SoC you are in the almost flat part of the battery curve. So voltage is a complete useless criterium.



Let me describe my current situation and my desired situation.
In the current situation, after reaching absorption voltage, the solar controllers go in float at 13.31V. (this is by the way around 99% SoC)

Then I switch on the hot water boiler. This will draw 60A. The voltage drops, the solar controllers compensate after a minute or so, and the draw drops to 10A. Even when the solar controllers have enough sun power to supply more than the required 60A. That is because float of the solar controllers is voltage controlled.


My desired situation is that I want to have a battery monitor that outputs current and SoC, for ionstance Modbus. This information is read by a controller (Raspberry Pi). The raspberry pi instructs the solar controllers (2 Morningstars) by Modbus about how much current they should supply.
The Morningstar MPPT60's have a slave mode, controlled by Modbus via Ethernet.
The software on the Raspberry pi (Node Red maybe) measures the momentary consumption, and changes the Morningstar slave current real time to force the battery bank current to zero.
This way you avoid minicycles on the LFP bank. That will cause wear.



Another thing I would like: the same controller should automatically switch on the hot water boiler at a certain SoC.



To realize the above, I lack only 3 things: the battery monitor, and sufficient time and knowledge of the programming of my Raspberry Pi. So I have to learn and invest time. After the current refit....



Of course, also this system is not perfect:
- The autopilot uses rapidly varying current. You can not regulate the slaved solar current to zero fast enough, I think. Same for the induction hob.

- In case of high current consumption (making tea using 2000W) the battery is still depleted. The solar controllers will go out of slave mode to replenish the battery from 79% to 80%.


To summarize my view, the chargers should be controlled by the battery monitor.



2 Morningstar MPPT60's with 960 + 890W of solar. 920 Ah 12V LFP bank. We have seen 60+60A of charging at noon sometimes. Even here in New Zealand in winter, we see more than 40A in sunny spells, between the rain showers

[hjohnson]

Originally Posted by SaylorMade

So, in a few threads here a side-topic has arisen, about wether or not the charger should be in control of when to stop charging a LFP bank.

I think it's worth having a seperate thread about this, since it seems quite important to me, and I would like to hear the different arguments.

For the debate let us assume that the LFP bank has a BMS which will let the user set the different parameters.
We will also assume that the BMS is setup with appropriate values for OVP as per LifePo4 Cells specs, and also an OVP on pack voltage.

My own point of view is that the BMS should control when to stop charging, since only the BMS knows the voltage of each individual cells, and it is my understanding that a charger can only feel pack-voltage.

The arguments I have heard for NOT letting the BMS be in charge ("first-line-of-defence) is that BMS are not to be trusted, but this seem counter intuitive to me, since we will rely on the BMS anyway to keep the cells balanced anyway, so that pack-voltage can be any safe indicator of when max voltage has been reached.


My own setup:
12KWh 48v bank for propulsion controller by a JK BMS with a charger from the supplier.

1.2 Kwh 12v house bank with JBD BMS which is charged mainly from a "stupid" charger just putting out 14.6 volts until the BMS disconnects.

Both of these are relying on disconnect from the BMS when the first cell reaches the set value for max voltage.

I realize that the sudden cut-off from a BMS may have implications for an alternator so that should be adressed seperately.

I'm absolutely from the camp of let the BMS be the one in control. The BMS knows the cell temperatures, it knows the cell voltages, it knows what the allowable c-rate is for the battery, and it knows what the charge history is. As such, it can set the charge (and discharge) limits for the battery, and adjust them on the fly.

The goal of the system should be that there shall never, ever be a disconnect. If you have a disconnect, something has gone seriously wrong and it was an emergency. In 19 months of having the system on my boat, we've never come close to that, and that's with an alternator, solar, and shorepower charging the battery.

And why wouldn't you trust the BMS? yeah, there are a lot of cheap knockoff JBD type FET BMSs... I wouldn't trust those either, but if you get a good one, such as REC or TAO or Electrodactus (or Victron if you buy their batteries), they're extremely reliable. Plus, even if they do fail, you set in decent failsafe/limp values into your chargine sources.

My system is as follows:

• 230Ah EVE cells in a 2p4s configuration, giving 460Ah capacity
• REC ABMS managing the battery and providing balancing
• Victron Cerbo GX as the brains of the operation
• Wakespeed WS-500 managing the alternator
• Victron MultiPlus Compact 2kva 120v inverter/charger
• 2x Victron MPPTs for my solar
• 18A Victron Orion-Tr for maintaining my starter battery
• Various other sensors and toys linked into the system

All of these systems play nice with each other and communicate (other than the Orion, but I have a BMV-712 on the starter battery). The thing I like most about it is that I can log in and check on my boat from anywhere in the world. Given my day job, it gives huge peace of mind.

[CaptainRivet]

Originally Posted by SaylorMade

So, in a few threads here a side-topic has arisen, about wether or not the charger should be in control of when to stop charging a LFP bank.

I think it's worth having a seperate thread about this, since it seems quite important to me, and I would like to hear the different arguments.

For the debate let us assume that the LFP bank has a BMS which will let the user set the different parameters.
We will also assume that the BMS is setup with appropriate values for OVP as per LifePo4 Cells specs, and also an OVP on pack voltage.

My own point of view is that the BMS should control when to stop charging, since only the BMS knows the voltage of each individual cells, and it is my understanding that a charger can only feel pack-voltage.

The arguments I have heard for NOT letting the BMS be in charge ("first-line-of-defence) is that BMS are not to be trusted, but this seem counter intuitive to me, since we will rely on the BMS anyway to keep the cells balanced anyway, so that pack-voltage can be any safe indicator of when max voltage has been reached.


My own setup:
12KWh 48v bank for propulsion controller by a JK BMS with a charger from the supplier.

1.2 Kwh 12v house bank with JBD BMS which is charged mainly from a "stupid" charger just putting out 14.6 volts until the BMS disconnects.

Both of these are relying on disconnect from the BMS when the first cell reaches the set value for max voltage.

I realize that the sudden cut-off from a BMS may have implications for an alternator so that should be adressed seperately.

It simply depends on your overall system design.
In general i full agree with you that BMS is the only instance in your system that knows the cell voltages and if a good one also the most precise SOC value.
I also design system so the defense lines are in order of precises of info means BMS has best info, 2 best is charge sources. Exception because a seperate instance is a victron BMV as last defence line to avoid single point of failure.
I am also a fan of seperate load and charge bus, if you have a hybrid inverter/charger put it on a seperate hybrid bus too. This also makes a very clean system design.
And no its not a single point of failure if BMS is 1st control/action instance, so its normal operation and bank maintenance. 2nd instance can be your BMS (my preference) or all your charge sources(depending on system design), which is your 1st line of defense. 3rd is then all your charge sources(my preference). Then there are more lines of defense.
As last resort i use a BMV712 which
a) you can see difference in SOC between BMS and BMV, trigger an alarm
B) triggers OVP or UVP the BMV relais which switches relay or SSR of load and/or charge bus
C) triggers charge delta of block 1+2 and 3+4, warning or disconnect again.
So you have an additional different instance in case BMS is dead or broken. So no single point of failure.
If the BMS has CAN and all sources connected have CAN too the best is if management goes over CAN.

And if the BMS is in control of all charge sources, there is no sudden disconnect as BMS can disconnect field current of alternator before cutting of all other sources. Or simply put an ArgoFEt splitting diode and a reserve lead battery as dump load.

[CaptainRivet]

Originally Posted by hjohnson

I'm absolutely from the camp of let the BMS be the one in control. The BMS knows the cell temperatures, it knows the cell voltages, it knows what the allowable c-rate is for the battery, and it knows what the charge history is. As such, it can set the charge (and discharge) limits for the battery, and adjust them on the fly.

The goal of the system should be that there shall never, ever be a disconnect. If you have a disconnect, something has gone seriously wrong and it was an emergency. In 19 months of having the system on my boat, we've never come close to that, and that's with an alternator, solar, and shorepower charging the battery.

And why wouldn't you trust the BMS? yeah, there are a lot of cheap knockoff JBD type FET BMSs... I wouldn't trust those either, but if you get a good one, such as REC or TAO or Electrodactus (or Victron if you buy their batteries), they're extremely reliable. Plus, even if they do fail, you set in decent failsafe/limp values into your chargine sources.

My system is as follows:

• 230Ah EVE cells in a 2p4s configuration, giving 460Ah capacity
• REC ABMS managing the battery and providing balancing
• Victron Cerbo GX as the brains of the operation
• Wakespeed WS-500 managing the alternator
• Victron MultiPlus Compact 2kva 120v inverter/charger
• 2x Victron MPPTs for my solar
• 18A Victron Orion-Tr for maintaining my starter battery
• Various other sensors and toys linked into the system

All of these systems play nice with each other and communicate (other than the Orion, but I have a BMV-712 on the starter battery). The thing I like most about it is that I can log in and check on my boat from anywhere in the world. Given my day job, it gives huge peace of mind.

Good system design. The only thing i would do differently is getting a cheap battery monitor for rhe lead and use the expensive BMV712 as last resort if REC BMS fails, its too expensive and 90% of its features not used for a starter. Also the SOC count is quite precise so you can calibrate your BMS correct and have an additional defense line. Had it the same then you and changed it thanks to Jedi explained using the BMV like this. Makes a lot sense to me

[CaptainRivet]

Originally Posted by s/v Jedi

I wrote that I think the BMS isn’t the best option because an active balancer has proven to be 10 times as good in balancing than the very limited balancing that BMS’s do.

The balancer linked below is very popular. It has a little soldered jumper labeled “RUN” that you can replace with an on/off switch or let the BMS control that when it has the output for it.

https://www.amazon.com/Cywhrvzsf-Equ.../dp/B09G74WDN5

An active balancer can also cover up a problem in a bank. If you have a low C-rate usage (dis-/charge) below 0.3C on the EVE/Lishe...or 1C for Winston an active balancer should not be needed if the bank was properly top balanced. If you then get big deltas something is wrong or you have mismatched cells or you couldn't fully charge it for a longer time so balancing didn't happened.

Its actually the best solution if the bms is switching of the charge sources, then the BMS controls the active balancer getting all back to balance if they are not when switched off because eg a cell peaked.
I have electrodacus with passive balancing and one EXTiO controls additionally an 5A Heltec balancer. The reason is i have electric galley and normally i stay below 0.25c and then the passive balancer of BMS can handle that in daily operation but there are some exception i am at 0.6C and my Lishen cells gets a bit unbalanced due to high draw and i need an active balancer as my bank is 1088AH and BMS is to weak with its 300mA balancing current to get it back to full balance quickly.
Still BMS is in charge of full normal operation.

[CaptainRivet]

Originally Posted by SaylorMade

So, in a few threads here a side-topic has arisen, about wether or not the charger should be in control of when to stop charging a LFP bank.

I think it's worth having a seperate thread about this, since it seems quite important to me, and I would like to hear the different arguments.

For the debate let us assume that the LFP bank has a BMS which will let the user set the different parameters.
We will also assume that the BMS is setup with appropriate values for OVP as per LifePo4 Cells specs, and also an OVP on pack voltage.

My own point of view is that the BMS should control when to stop charging, since only the BMS knows the voltage of each individual cells, and it is my understanding that a charger can only feel pack-voltage.

The arguments I have heard for NOT letting the BMS be in charge ("first-line-of-defence) is that BMS are not to be trusted, but this seem counter intuitive to me, since we will rely on the BMS anyway to keep the cells balanced anyway, so that pack-voltage can be any safe indicator of when max voltage has been reached.


My own setup:
12KWh 48v bank for propulsion controller by a JK BMS with a charger from the supplier.

1.2 Kwh 12v house bank with JBD BMS which is charged mainly from a "stupid" charger just putting out 14.6 volts until the BMS disconnects.

Both of these are relying on disconnect from the BMS when the first cell reaches the set value for max voltage.

I realize that the sudden cut-off from a BMS may have implications for an alternator so that should be adressed seperately.

Jk BMs is good low cost FET based BMS but i wouldn’t trust it as propulsion bank BMS as the motor creates high surges and spikes that easily can fry the FETs and the fry 90% close so you have a problem. Eg if you catch a net and prop locks is a 99.99% chance your JK is toast. Additionally their FETs are not really well matched means your max current is 80% of rating.
Perfect for propulsion banks if staying in budget range is actually an Electrodacus BMS as its not carring any current and uses factory remote of eg MPPTs to control and switch sources on or off.
Also on of main reason i chose it because i run a Lifepo4 hybrid starter and service bank so starter and windlass can create high spikes and surges killi g a FET based BMS.

[CaptainRivet]

Originally Posted by sailorboy1

to me the BMS is protective device that never should ever get "tripped"

Thats true for the very simple 5$ BMS in cheap drop ins. That must then have a BMV as last line of defense.
A superior external BMS that additionally and actually manages and optimizes the bank/battery acts a manager in normal operation and as BMS protection device in defense lines that should never get tripped.

[s/v Jedi]

So, should the BMS also steer the boat? I mean it knows how the battery really is and could steer more accurate with full battery and then save power for low battery

[team karst]

Originally Posted by s/v Jedi

So, should the BMS also steer the boat? I mean it knows how the battery really is and could steer more accurate with full battery and then save power for low battery

Or at least call for abandon ship.

[CaptainRivet]

Originally Posted by s/v Jedi

So, should the BMS also steer the boat? I mean it knows how the battery really is and could steer more accurate with full battery and then save power for low battery

Balancing is part of normal charging and management of the bank, so BMS as only source who have all information should steer that too.
If you active balance permanently or before 3,45V is contraproductive as you create imbalances. So the BMS steers the active balancer.
Thats how its done in all commercial grade BMS i know.

What steers the boat is the autopilot and if the BMS switches the bank off, the autopilot is not working anymore. Was always like that doesn't matter which chemistry you have if bank is toast, compared to a lead the LFP is switched off by BMS.
Every lifepo4 when properly installed is much safer then any lead and much more controlled

[SaylorMade]

Originally Posted by s/v Jedi

So, should the BMS also steer the boat? I mean it knows how the battery really is and could steer more accurate with full battery and then save power for low battery

No more red wine for this gentleman tonight

[nfbr]

Originally Posted by Panope

How do electric cars do it?

They bottom balance and BMS cuts off charge when one cell hits top limit.
More important in cars to get range accurate than get last few % capacity.

Boat house loads tend to opposite.

[s/v Jedi]

Originally Posted by CaptainRivet

Balancing is part of normal charging and management of the bank

No, balancing and charging are not part of management, they are done by resp. balancer and charger.

The BMS may provide info to those devices, even recommend them to stop or start but there should always be an override to those signals so that the owner keeps the authority for final decision making.

Many owners who have the “completely integrated” systems have been stranded, waiting for help to get things working again.