Lithium Battery Confusion

[posplayr]

Here is the Overkill Solar BMS protection setting for a 200 AmpHr LiFePO4 battery.
https://overkillsolar.com/product/bms-120a-4s-lifepo4/


This BMS doesn't disconnect till 10V and doesn't come back till 12V and 2 seconds.

How does I come back to 12V if it is disconnected at 10V??? Maybe it is just partially disconnected so that it can revive the cells if some power is presented to the terminals of the battery.


The question for jonyjoe101 is why was his BMS disconnecting and what was the voltage? 11.1V?

Both the Renogy DC-DC and Rich Solar are going to try restarts at below 11.1V and the OverKill BMS doesn't disconnect till 10V (as a total fail-safe).

I don't think these are hard and fast rules but there seems to be some consensus among these various (dare I say quality) brands for LifePO4 battery management.

[posplayr]

OK more info from the Overkill manual defining the parameter limits and functional behavior.

This BMS separately manages the charging and discharging currents. So when this BMS disconnects it is not a total disconnect but only the part of the current exacerbating the problem encountered.

[cotrailhead]

If I understand your question, I’d answer this way. The bms has a charge mosfet and a discharge mosfet. When the bms reads 10.0 volts it will disallow discharging. It will still allow charging. When the voltage returns to 12.0 volts, the bms will resume discharging as needed. Lifepo4 would have a very low state of charge at 12.0 volts. Applying a large load, like a microwave for a few minutes, would likely cause the voltage to sag, my term, and hit the low voltage protect at 10.0 volts this disabling discharge again. Once the discharging is disabled the battery would likely settle above 10.0 volts.

[jonyjoe101]

The bms will disconnect either from the battery reaching its max voltage (14.6 volts), or an out of balance condition forcing the bms to disconnect. This happens if your charging at high amps near the top end, but it can also happen if its slow charging near the top end and it gets cloudy and the sun suddenly comes out of the clouds and the amps go up.

All mosfet bms leak voltage, if charging with solar, eventually you might encounter the bms activating. The only way to avoid is to set the bulk setting so low that the battery never charges at high amps so it never reaches any of the bms high voltage disconnect levels. I tried many bms and they all behave the same way, and I also tried several controllers mppt and pwm and they also produce voltage surges. The only bms that actually worked was the chargery bms8t which I use, but this will only work on batteries that dont have a bms already, this bms uses mechanical relays.



If the mosfet bms activates it won't damage the controller or the battery in my situations. It only damages any 12 volt devices that are connected to the battery and are operating at the time.



This is from the makeskyblue features where it explains it prevent the voltage to overshoot through the bms protection board.

[posplayr]

Originally Posted by cotrailhead

If I understand your question, I’d answer this way. The bms has a charge mosfet and a discharge mosfet. When the bms reads 10.0 volts it will disallow discharging. It will still allow charging. When the voltage returns to 12.0 volts, the bms will resume discharging as needed. Lifepo4 would have a very low state of charge at 12.0 volts. Applying a large load, like a microwave for a few minutes, would likely cause the voltage to sag, my term, and hit the low voltage protect at 10.0 volts this disabling discharge again. Once the discharging is disabled the battery would likely settle above 10.0 volts.

Thanks for the explanation and that is what the OverKill Solar graphs above indicate.


EDIT I posted this before seeing jonyjoe101's post above. You are still using the term disconnect but not saying which MOSFET (Charging, discharging, or both)

The question I have then is for jonyjoe101; "Did his BMS not have this feature" because if the BMS activated (his terminology) it should be discharge deactivated if it is a low voltage disconnect.

Then how is a discharge deactivated battery going to present 11.1v and not accept a charge?

Maybe he had an overcharging disconnect with a load pulling the battery down to 11.1V ?

I'm really not familiar with the inner workings of a BMS and I'm not sure he has explained what happened well enough to determine if it was a high voltage charging disconnect or a low voltage discharge disconnect.

[posplayr]

Originally Posted by jonyjoe101

This is from the makeskyblue features where it explains it prevent the voltage to overshoot through the bms protection board.

This is describing load disconnect which I assume doesn't mean a battery disconnect?

[posplayr]

Originally Posted by jonyjoe101

The bms will disconnect either from the battery reaching its max voltage (14.6 volts), or an out of balance condition forcing the bms to disconnect.

The Overkill Solar graphs show clearly that either battery overvoltage or cell overvoltage will result in MOSFET charging disconnect. Low voltage results in discharge MOSFET disconnects.

You would need simultaneous high and low conditions to fully disconnect in both MOSFET directions.

[jonyjoe101]

The only one that will cause voltage surges will be overvoltage disconnect, its the only time I encountered voltage surges. But if the battery is completely dead (and bms has activated to disconnect further discharging) and you try to charge it with solar (something I never encountered) you might also get voltage surges since the battery won't accept amps. When the mosfet bms disconnected the battery it was around 14.5 volts, but once the bms disconnects if I connect a multimeter to the terminals the battery reads 11.7 volts (which is the voltage that the bms is leaking, a good bms would read 0 volts) , I have to reset the bms manually by either unplugging it or trying to put a load on it, then it reads its resting voltage about 13.3 volts.

On the makeskyblue features, they mention load output (shorted an overcurrent protection) but also the battery output which is connected directly to the bms is where it talks about voltage overshoot/surge. The load output on most controllers are design to prevent voltage surges, so if you connected the lithium battery to the "load output" it would be protected from surges, also if you connect all your 12 volt devices to the load output they would be protected. But the majority of users connect there devices directly to the battery.


The best way to describe what is happening is you have 2 types of relays on bms. The one that is used on 95 percent of bms is electronic relays (mosfet) because they are cheaper and no moving parts. But the drawback is they leak voltage. When activated they won't allow current to flow to keep charging the battery, but the battery will leak voltage and the controller will see a battery connected reading about 11.7 volts.
With a mechanical relay (example would be a 12 volt 30 amp automotive relay) when it activates it wont allow current to flow and since it completely disconnect the battery, the controller sees that no battery is connected.


With the mosfet bms, the controller will try to charge the battery since it sees a battery connected, with a mechanical relay the controller does nothing since there is no battery connected.

To me this was the biggest problem when switching my solar system from lead acid to lithium. And I struggle understanding what was happening that was causing the voltage surges and destruction of my fans. Thats why I dont use any mosfet bms with solar. Even the makeskyblue controller I never tested it in anger to see if it would actually work as advertised. Having experience many voltage surges first hand, I avoid them and advise everyone on what to look out for. Imagine having a 200 dollar roof vent fan and running it at 20 volts, it will burnout in seconds maybe milliseconds. It happens quick.

[posplayr]

Originally Posted by jonyjoe101

The only one that will cause voltage surges will be overvoltage disconnect, its the only time I encountered voltage surges.

Thanks; yes that is the only way I could see a BMS disconnect causing a load dump to occur. Thanks for the detailed description.

I'm working on a surge-protected bus bar for installing inside a DIY battery case that only exposes terminals for the solar charge controller, inverter, and DC chargers. This is going to be low quantity so I have to keep costs low with the direct benefit of being less complex and hopefully more reliable.

I have taken some of your ideas and will be ordering parts soon to do some prototype testing to confirm my analysis. Right now I have a design that I think should clamp all voltages to not exceed something like 17V (@ 10 amps of dump)with or without a battery disconnect.

This link simulates a standard "load dump " from an alternator of 2.88A@ 86V for a duration of about 0.3 sec.

Design example for Alternator "Load Dump"https://us.v-cdn.net › uploads › FileUpload

[cotrailhead]

also if you connect all your 12 volt devices to the load output they would be protected. But the majority of users connect there devices directly to the battery.


This is the part of the post that confuses me. The bms should live between the battery and all discharge/charging devices. If the bms shuts down all discharging, but there is a load connected to the battery directly, your battery is still in danger of over discharge.