BMS common port vs seperate port

[GXMnow]

When everything is working correctly, the BMS is just along for the ride. If the cell balance goes out, it will try to pull it back in. IF the cells are well matched, it does not have much to do. A few milliamps for a few hours will keep god batteries in balance. The Charger will stop charging before the batteries get too high, and the inverter will shut down before they get too low, so the voltage protection in the BMS should never be needed.

BUT>>>

There is always a but.. If the cells do drift out of balance for any reason, and the BMS balancing is not able to keep them in range, you could run into a cell going to high during charge or too low during discharge. Depending on how far the inverter or charger is trying to use the pack, an imbalance could be enough for the rouge cell to go far enough out to hit the protection threshold in the BMS, even if the whole pack is still in the safe voltage range. Without monitoring each cell, the charger and inverter do not have any way of knowing if one cell is out of range. The better the batteries are matched, the less the BMS needs to do, but due to how lithium batteries work, a BMS is always needed to make the system safe. Even perfectly matched cells could age differently and become unbalanced as they age.

The question becomes what happens if the BMS has to disconnect. If it does, it means something is wrong. It should never need to happen. So the main thing is we don't want it to damage anything else. Disconnecting the battery from a solar charge controller is not a good thing. Ideally, it should see the voltage jump up and go into constant voltage mode. This would be a good question to ask and should be in the manuals since more and more people are turning to Lithium batteries. If a charge controller can't withstand the battery being pulled due to a BMS cutoff, then I would say that charge controller is not compatible with lithium batteries.

On the discharge side, most inverters will just shut off when the battery is disconnected. It is certainly no worse than a fuse blowing. And that can happen easily from an overload. If an inverter can't survive it, there is a bigger issue.

Unlike a fuse blowing due to an overload, a BMS shut down could have a few different reasons. If a cell keeps going out of balance, it is telling you to repair the battery pack. If it is a current shut off, it is being overloaded, and you need to adjust things to prevent that. And if it is a temperature issue, you need to correct the cause. If it is working correctly, it never should shut off.

Since my system is based on an inverter/charger unit, I had to use a common port BMS. But it is nice to know it can stop charge and still allow discharge, and stop discharge and still allow charge. If a battery is run to low, we certainly want to be able to charge it again.

If a system has separate charge controllers and inverters, then it makes sense to use the separate port design. but common port will still work. It would be interesting to look at the diagrams and see how they layout the mosfets to see if you could easily convert between a split and common port system.

 

[boatmonkey]

interesting read.

A separate port seems to be preferable if it handle decent charge rates. Can anyone confirm a working 50A charge current ( as Will's OP is already a year old) with a daly bms?

 

[boatmonkey]

After reading the first two pages of this thread again, i have a question: Because the charge rate of the separate (charge) port is too low, it is advised to get the common port bms, and connect the SCC directly to the battery (bypassing the bms).

But how do you protect the battery from over voltage in that case? Does the SCC need to take care of that? It cant, because it does not know the voltage of the individual cells...So how is the battery protected against over voltage?

 

[GXMnow]

No, you should not bypass the BMS without some form of protection. Will did do a trick where he used a lower current BMS, but had it's output lead drive a contactor that connected the inverter to the system, but this would also work for the charge controller. The whole point of a BMS is to protect all of the cells from over voltage, under voltage, over current, and temperature extremes. When using the relay bypass trick, it does eliminate the current protection, but you can (and should) still have a fuse or breaker to limit the maximum current incase something goes wrong. The rest of the protection systems still work because when the BMS goes open, it drops the contactor (relay) and shuts off the load or charger. If you were going to do the relay trick, I would recommend a separate port BMS and 2 relays. One for the charger, and one for the load. This way if it runs low, it can still charge, and if it goes too high, the loads can still pull it down.

When using a common port BMS, the inverter, charger, and any other loads, should all connect on the output lead of the BMS. Mine is a common port that allowed me to set different maximum current for charge and discharge direction. And even though it is on the same physical wire, it can block charge when the voltage is high, or block load when the voltage is low, but still allow the other direction so a low batt can be charged etc.

The SCC should be set for the whole pack voltage. Set it a couple tenths of a volt below the desired max cell voltage times the number of cells. In my case, I have MNC cells that are safe to 4.2 volts. I have 14 in series for a max voltage of 58.8 volts, but I have my absorption voltage set to 57.6 volts, which works out to 4.117 volts per cell. Add in a little wire voltage drop, the BMS drop, and some calibration error, and my BMS reports right at 4.1 volts per cell when the charger stops. The BMS will not disconnect until 4.2 volts. on a cell. On the low side, I am even more conservative. I have the inverter shutting off at 44 volts, which is still 3.2143 volts per cell. The BMS is set to 3.0 volts per cell, or just 42 volts for the pack.

 

[Dzl]

After reading the first two pages of this thread again, i have a question: Because the charge rate of the separate (charge) port is too low, it is advised to get the common port bms, and connect the SCC directly to the battery (bypassing the bms).

I would argue that this is not 'what is advised' (I know in some of his earlier vidoes, Will suggested bypassing the BMS, but beyond that I haven't come across that advise, and I'm not sure he still stands by that advise for beginners or people buying grey market cells--in fact mostly I've come across exactly the opposite advice).

The reason for this is because your concerns below are valid:

But how do you protect the battery from over voltage in that case? Does the SCC need to take care of that? It cant, because it does not know the voltage of the individual cells...So how is the battery protected against over voltage?

People that suggest it is acceptable to bypass the BMS for charging to save money because they argue risk of overcharging is minimal. Setting aside whether that is true or not, it seems like the wrong solution to a problem that can be solved in safer ways. There are many BMS on the market that can handle higher charge rates, some are FET based many use relays or other measures. Bypassing the BMS pokes holes in your most essential layer of protection. That is my 2c

 

[John Frum]

Here is another strategy.
If your inverter supports it you can use bms+opto coupler to control the inverter without exposting the fets to that amount of current.
Hopefully your inverter has a sane low voltage cutoff so that its very very rare that the bms has to intervene for low voltage conditions.
Let the charger and dc loads go through the fets.
You can additionally use a battery protect for the dc loads.

 

[PeterBC]

OK newbie here, first post so be gentle please..
I'm looking to get a 24 Cell 280AH LiFePO4 bank for 24V setup on my barge. If I run 3P8S I'm told I can run one 8S BMS, but my peak charge current is 2.4kw from a generator (plus 1kw of MPPT Solar) (100 Amps). Peak Discharge is 5kw (220 Amps) to a Victron Quattro 5000. So I gather I need to go common port to get the charging current. Or can / should I run 3 parallel banks of 8S with a BMS on each 8S bank to spread the load more ? If so thats a 33Amp peak charge, and 73Amp peak load.. so can /should I use 3 separate port BMS's 80A load/40Amp charging rated. I see that an 80A separate port is way less than 1/3 the price of a 250A common port so come out ahead anyway ?

Furiously reading and trying to get up to speed with other nuances and bits needed/recommended for my setup...

 

[John Frum]

I vote for 3 discrete batteries each with its own bms and fuse.
Distributes the load/charge stress and provides some fault tolerance.
Search for the document "wiring unlimted" and read the section on paralleling.

 

[Dzl]

I think both options "parallel first" / single pack OR "series first" / multiple packs have there place and have there strengths and weaknesses.

In your situation I would tend to agree with Joey, 3 separate packs makes more sense.

Consider that many people are already finding it difficult to keep 280Ah EVE packs in balance in a 1P configuration (280Ah per logical 'cell') with the BMS's built-in balancing. The EVE cells are not closely matched, and 280Ah is quite a large capacity cell which amplifies the impact of small differences between cells.

In a 3P configuration you will have 840Ah per logical 'cell'. I believe a Daly BMS has a balance current of 35mA. If I'm thinking about this right, at the max theoretical balance current it would take just over a day to burn off just 1Ah (0.1% of your logical 'cell' capacity)

To fully appreciate the magnitude of the difference:
840,000 mAh
000,035 mA

If you do go the route of 3 separate packs (as has been suggested), I would try to use higher throughput (current) BMSes than what you are considering if you want to take full advantage of the 'fault tolerance' that Joey mentioned. If one pack goes offline (for whatever reason), 2/3 of the packs will need to handle 3/3 of the load. If 2 packs go offline, 1/3 of the packs will need to handle 3/3 of the load.

If I were in your shoes I would try to at least design for the 2/3 condition. 66% of the battery bank, carrying 100% of the load. 3 x 150A separate port BMS would allow 2 of 3 BMS to carry the full rated continuous current of your inverter and 3 of 3 to carry almost the full surge rating.

 

[MisterSandals]

In a 3P configuration you will have 840Ah per logical 'cell'. I believe a Daly BMS has a balance current of 35mA.

I am completely onboard with all your recommendations. Balancing such a high AH “cell” would be very difficult.

Just curious to see the results and experiences of someone who tried a paralleled first pack. I do not recall such a discussion nor could i find one.

 

[Dzl]

I do not recall such a discussion nor could i find one.

In terms of parallel 280Ah EVE cells specifically or just generally?

 

[MisterSandals]

In terms of parallel 280Ah EVE cells specifically or just generally?

The 280Ah Eve’s.
I do not recall any parallel first results/problems/successes.

Not trying to make a point, just an observation for those making a case for both battery configurations.

 

[juanmijm]

there is no problem or contraindication in connecting two battery banks in parallel ?, each pack with its own bms

 

[MisterSandals]

contraindication

Wow i learned a new word today. I was certain that was a typo or made up word.

 

[Berseker]

Wow i learned a new word today. I was certain that was a typo or made up word.

Common in the medical/pharmaceutical world

i am eyeing a 200amp discharge and 50amps charge Ant bms. actually 2 units for a 16s2p 280 lifepo4 bank. am a newbie, so trying to catch up on the bms bit. i understand the discharge end should be underated by 50%, but what of the charge side?. should i assume 50amps is ok, if yes, so i can charge my bank at 90 - 100 amps?

 

[John Frum]

Common in the medical/pharmaceutical world

i am eyeing a 200amp discharge and 50amps charge Ant bms. actually 2 units for a 16s2p 280 lifepo4 bank. am a newbie, so trying to catch up on the bms bit. i understand the discharge end should be underated by 50%, but what of the charge side?. should i assume 50amps is ok, if yes, so i can charge my bank at 90 - 100 amps?

Did you just ask if you can charge at 90-100 amps through a bms rated for 50 amps?

 

[John Frum]

I read that again after a second cup of coffee and it seems you want two discrete batteries.
Presumably each will draw approximately 50% of the the charge current.
That would work if an Ant BMS can actual perform reliably over time at 100% of its rating.
Thats probably overly optimisitc.
Field effect transistors=FETS are prone to failure from heat stress.
They also have a reputation for failing closed.
That means that you won't know there is a problem until the BMS just doesn't disconnect when you need it to.

 

[Berseker]

Did you just ask if you can charge at 90-100 amps through a bms rated for 50 amps?

Lolz
Emmm.
I intend to run 16s2p, hence 2 bms rated for 50amps charging, hence if i send 90amps charging current, it will be borne by the 2 bms, giving me 45amps each on each bms.

Is my thinking in order, kind sire??
If nay, do tell max charging amps you think is ok

 

[ArthurEld]

Most people try to use a BMS that will handle twice as much amps. So you should get two 100A BMSs.
There are endless stories of people destroying their BMS or batteries because they used a BMS that was supposed to handle the amps they use.
It's much safer to double it.
Also, you might need to use one battery while you work on the other.

 

[Etienne]

Hi,

Just read the whole thread, interesting topic. Sorry if my question doesn't directly concern solar application.
I would like to build a lifepo4 12V pack to eventually replace a lead acid battery in a car...

I 'm looking to buy a DALY BMS and need to choose between common or separate.
With traditionnal lead acid, charge and discharge are done from same port so I would say the common version would be more appropriated here...
(it seems that we cannot use a separate version and still charge and discharge from same physical port?)
My concern with common port is about overcharge. It is specified from specs charge voltage is 14.6V, which I think should be close to what a car provide generally. Do you think there would still be risk of overcharge and battery disconnection?

Also I'm wondering which amperage should be taken.
From the specs, it is specified 100A version has 300A sparkle current (not sure to know exactly what it means) .
As lots of current is only going to be required at start, and for a few seconds, I wonder if 100A version would be enough.
Do you know if BMS would handle any momentary extra current or if it will go to overcurrent protection and forbid the car from starting??

Thanks