School me up on BMS in LiFePo4 please

[KS_Kampers]

Can you point me to a thread or URL that discusses the operation of a BMS? I'm wondering if the majority of them operate similarly; I would bet there are subtle differences. I'm particularly interested in the operation of the Li Time BMS in their 100AH TM battery. (Yeah, i know... that is very specific, but it is what I have).

I'm seeing some things that strike me as mildly weird, and I want to try and do some self education on how the BMS operates.

Specifically, what happens when the BMS senses the battery reaches full charge. If there is some hysteresis such that the battery must reach a specific level of discharge prior to turning on the internal charging circuitry again.

Before getting too deep into what I have measured, I'd like to get educated on BMS operation. I've done some googling around and most places simply define what the BMS is, without going into depth on what it actually does. I'm more interested in the details of operation.

Tnx.

 

[Brett V]

The BMS doesn't control charging. It's a simple safety device that monitors the individual cell voltages, charging current, discharging current, and overall battery voltage, and temperature. If any one of these parameters goes outside of the established parameters it shuts down until the problem is corrected.

 

[kommando]

Think of it as a fuse that blows on more parameters than current/amps alone and these parameters look at every individual cell.

 

[Tulex]

The BMS doesn't control charging. It's a simple safety device that monitors the individual cell voltages, charging current, discharging current, and overall battery voltage, and temperature. If any one of these parameters goes outside of the established parameters it shuts down until the problem is corrected.

That's really not true. The BMS monitors parameters and allows or disallows charging based on those parameters. I would say that's controlling charging.

 

[Pappion]

I had a WFCO 9855 in my old TT. With DIY 200AH and 280AH DIY LiFePO4 batteries and solar. It all played well together.
Sometimes I didn't bother to put the heavy solar panels outside on the ground. After 48hrs the WFCO 9855 would drop to 13.2v storage voltage. There is a trick to getting back up to higher voltages: 1. disconnect the batteries or turn off the battery switch. 2. cut the converter power. 3. restore converter power. 4. reconnect the batteries. The trick is handy to do the day before traveling.
BTW: My WFCO 9855 never went to bulk charging. It never sensed enough current draw due to cable resistance (gauge and length).

 

[Pappion]

The BMS acts like a big electronic switch. Can shut down charging and discharging.
Cell voltage imbalance can shut down charging.
Low voltage can shut down discharging, maybe both.
Temperature sensor can shut down both.
Overcurrent can shut down discharging, maybe both.
There is a slower software shut down for the rated output, ex 100A.
Faster hardware shut downs at higher currents. Trading off noise immunity with higher current triggers.
Overkill solar 100A 4s BMS:
Hardware Overcurrent Protection: 830A, Delay 320ms
Hardware Short Circuit Protection: 1550A, Delay 100us

 

[Mahendra Gomanie]

UTube offgrid garage guy has tons of nice videos that deals with them all or most

 

[TacomaJoe]

My BMS has programable settings for battery voltage and cell voltage. At the battery level (sum of 4 cells) 14.6V is the cutout and 14.0V is the release. At the cell level 3.650V is the cutout and 3.500V is the release. So if one of my cells hits 3.65V the BMS will cutout until the cell discharges to 3.5V.

 

[LakeHouse]

That's really not true. The BMS monitors parameters and allows or disallows charging based on those parameters. I would say that's controlling charging.

There's a discussion of semantics here that I'm definitely not interested in having.
But the at the very least, the BMS shouldn't control charging (with the exception of BMS's that communicate directly with inverters, but that's not the topic of discussion here). The BMS is more like the guardrails on a freeway: Sure, it keeps you from driving off a cliff if everything else goes horribly wrong, but if you're relying on it to keep you on the road, you are most definitely doing something else wrong.

Specifically, what happens when the BMS senses the battery reaches full charge.

The BMS doesn't exactly sense 'battery fully charged' at all. It senses that an individual cell has reached some predefined voltage. At that point, it disconnects the circuit (using MOSFETS usually, but solenoids in some very large BMS's) that allows current to flow into the battery to charge it.

If there is some hysteresis such that the battery must reach a specific level of discharge prior to turning on the internal charging circuitry again.

The BMS will turn the charging circuit back on when the cell that went over its protection voltage falls below its recovery voltage. Usually this doesn't require any discharge, just some time for the cell to rest, which reduces voltage down the ~3.4V for a fully charged LFP cell.

 

[KS_Kampers]

OK. Here's where my confusion comes from. I have recently purchased three Li Time Batteries. All LiFePo4 batts.
(As a side note, I used to design power supplies and battery chargers for NiCad marine application stuff a long time ago. I am trying to not let my experience get in the way of understanding BMS and LiFePo4 health and happiness).

Currently using a Progressive Dynamics PD9160 set to lithium mode. It applies ~14.4 for at least one hour, then folds back to 13.6V as long as converter is powered up.

Using a Victron smart shunt to observe voltage and current, record AH. Also have a good Fluke meter for sanity checks.

Batt 1 and 2 : 100 AH TM 12V batteries (w/cold protection)
Batt 3 : 50 AH battery.

Battery 1 currently on my bench :
It appears to me that once B1&2 get "fully" charged. the BMS does a wierd-ish partial disconnect. Sort of goes to sleep. It can provide a little power but does not accept any charge.

Once I remove my 13.6V charge voltage, AND (I think) if the batteries are ~fully charged, AND there is no significant load, I note that the terminal voltage starts to drop somewhat quickly. It goes from 13.6V to 13.05V within 90 minutes. Overnight that one battery (B1), with no load dropped to 12.99V

I also note that this battery, before I disconnect, draws no charging current after what I think is fully charged. If I put less than a 2A load on the battery, it stays at this lower "sleep" state, and the voltage starts to drop further to 12.8V or lower.

However, once I pull 2A or more, the battery wakes up and jumps to 13.35V. I pulled about 10AH off of it yesterday with no recharging applied. I am currently pulling 48A from that battery now, to try and measure capacity.

Here's an interesting note... this battery, B1, after having a 10Ah draw yesterday with no recharge, sat on my bench overnight and never went into this weird sleep mode. I measured it this morning after 12 hours of no load, and it measured 13.30V.

Battery 2 "fully" charged :
No load on this battery, and took a voltage reading this morning. It was sitting there at 12.98V. I did not attempt to wake it up as my load is currently on B1 above.

Battery 3 "fully" charged :
This is my 50Ah portable battery. When I initially put it on the PD charger, it pulled a heavy charge current for quite awhile. Perhaps 90 minutes. I don't have data for this, but the converter cooling fan was running and the heatsink was warm. So that's my metric for now.

Here's the weird part... that battery has been sitting for days without load or charge, and the terminal voltage is 13.41V. It doesn't seem to go into "sleep" mode like the 100Ah batts.

Here's why I care :
Other than being stupidly OCD and craving uniformity, those 100Ah batteries are going into my camper for boondocking and charging from solar panels. I want to better understand this sleep mode, as there will be long periods of the day in which I am NOT drawing power from them but I do want them charging if they are at 75% SoC or lower.

I don't want to be using <2A all day from leakage, or a small fan or LED light only to keep the battery in sleep state and draining overnight and then ignore the available charge once the sun comes up.

Angst :
I've been trying to communicate with Li Time to ask questions, and I have gotten no answers. I have bugged them enough that they have acquiesced and are sending me return labels. However, If these batteries are working as designed, I don't need replacement.

I need them to accept charge once at 75% ish or lower no matter how long they have been dormant.

Feedback appreciated. (craved?) Especially from folks who might have Li Time batteries.

 

[Pappion]

The BMS FETs have a inherit Body Diode, it still conducts when FETs are off, in one direction, albeit with a 0.7v to 1.2v drop. The 2 Banks of FETs are placed Back to Back, if both banks are off it is definitely off. If 1 Bank is off, it can have the Body Diode voltage drop in one direction, and be off in the other direction. So the drop from 13.6 to 13.05 makes sense.
Your meter is likely 10M ohm, to little loading. FET leakage currents might throw off your voltage measurements.

 

[TacomaJoe]

I recall an issue with Ampere Time batteries (before they changed the name to LiTime) but don't recall specifics. I did find this video:

Could be some issues have followed the rebrand.

 

[OTRwSolar]

Oops.

 

[Q-Dog]

It really just sounds like the cells inside the battery are unbalanced and the BMS is shutting off charging when one cell gets overcharged. Use the battery, and charge it as slowly as you can stand, and the cells will eventually balance.

 

[KS_Kampers]

I'm going to do a complete discharge on the battery I have on the bench today. I don't have a low amp charger, so I will likely charge it with what I have for now and monitor the Ah charge it takes.

I need to debug / setup my Victron smart shunt. It appears to want to reset the meter every day so I lose the continuity of measurements right now. I note the Victron manual is not very verbose. Does anyone actually do decent tech writing any more? (Looking at you too, Garmin).

 

[Q-Dog]

Keep the 2 big bateries together ... discharge and charge them together and you cut the charger output to each battery in half. Put a load on the system while charging and you cut the amperage into the batteries even more. The longer you can keep the batteries charging at their upper voltage range, the sooner the cells will balance.

 

[Ampster]

I would say that's controlling charging

I think the point that @LakeHouse was making was that the BMS is a safety device that should be considered as a last resort for control purposes. The inverter or charge controller should be the primary control device because either of those devices can control current, voltage and when the charging stage transitions from Constant Current (Bulk) to Constant Voltage (Absorb).

 

[KS_Kampers]

Keep the 2 big bateries together ... discharge and charge them together and you cut the charger output to each battery in half. Put a load on the system while charging and you cut the amperage into the batteries even more. The longer you can keep the batteries charging at their upper voltage range, the sooner the cells will balance.

Thanks. I want to characterize each battery first, to make sure that the capacities are relatively equal. Then I'll charge them individually. Great idea using a load while charging to control speed of charge. Then put them in parallel and repeat the characterization test.

Dang. Being OCD hurts.

 

[Tulex]

I think the point that @LakeHouse was making was that the BMS is a safety device that should be considered as a last resort for control purposes. The inverter or charge controller should be the primary control device because either of those devices can control current, voltage and when the charging stage transitions from Constant Current (Bulk) to Constant Voltage (Absorb).

I agree with the main purpose of the BMS being a safety devise. And stand alone, that's what it is.
In my situation, I don't have that option. Being closed loop with my 15k, the BMS does all the controlling. I read how you got the 12k to accept values, I can't get the 15k to do that. If I remember, any value I got the inverter to take, it changed my BMS values. So whatever my values are in the BMS are what it's going to do.

So yes, the BMS is first a safety devise, but also can control the charging.

 

[Ampster]

but also can control the charging.

Yes in that situation as noted earlier, the BMS does indirectly control charging. What I mean by "indirectly" is that the BMS in closed loop, monitors the battery, and controls the Voltage and Current by telling the inverter what Voltage and Current to supply.to the batteries.The BMS has no direct control over Current or Voltage.