Building 100 AH LiFePo4 Battery Box

Hello everyone,

I started acquiring parts for this project just over a year ago, but what with three days off due to snow days (!) I finally decided to get started on assembling some pieces together.

Technically, I started assembly about a month or so ago when I got my 100 AH LiFePo4 batter cells--four of them.  The first step was to balance the cells.  That meant checking each individual cell voltage to see if there was any discrepancy.  Three of the cells read 2.999 to 3.000 volts, which is about perfect for shipment.  But one cell read 2.990 volts.  This might not sound like much, but over time it will affect the lifespan of the batter pack.  So I connected all four cell in parallel for a week and checked again--and the one cell had gone up to about 2.994 volts.  So I connected them back and let them sit--for a month (I got involved in other projects).  After I checked, all four cells are now reading 2.998-2.999.

After I decided to start assembling, I got to cutting out placement holes in the ammo can and put in the basic exterior components.

I will attach some pictures soon.

Eric


Edited to add pictures

Hello Eric,

I highly recommend to get a Battery Management System that can balance the cells.
It will protect against over-charging and -discharging and balance your cells when charging is complete.

Lead-Acid batteries handle over-charging without problems, they just get warm.
LiFePo4 cells will get damaged.

Sebastian,

Absolutely, I have a BMS in the parts list.

Okay, good.

The USB chargers are nice.
I want some on my (2.5KWh) battery bank as well, but it has 16 cells in series and I have yet to find a USB-PD (Power delivery) module that supports voltages of 60V (58V when charging stops).

Sebastian,

I have plans on some day making something much larger--perhaps a 300 AH 24 Volt system.  This would be sized like a large rolling tool case and it would have a good AC inverter attached.

In the meantime, I saw this build on a YouTube video and I like the compact nature but still having a relatively high capacity.  Also, I was wanting to build the battery itself instead of buying the actual battery outright.  This piqued my interest as I could have a virtually unlimited supply of electrical charge for all sorts of small electrical devices in a fairly compact case--and these ammo cans are really nice to work in.

After this is built, I want to acquire a set of solar panels that can charge this in a reasonable amount of time.

Also, for the record, this was supposed to be an 80 AH system as per the build on YouTube but the 80 AH cells were not available.  The next size up was 105 AH which is what I got.  They look like they will still fit so that is what I will be working with.  And to boot, the 105 AH cells were actually slightly less expensive than the 80 AH cells!

Eric

For the outputs I  included a bunch of USB--Six USB type A and three USB type C.  On the lower right I have a cluster of four Anderson Power Pole connectors.  On the lower left I have an XT60 connector which I plan to use as the main charging port.

On the back side I decided to include two 3 Watt lights so that I could use the box as a light source.  I might put lights on the top so that I could shine the light up on the ceiling as an area light, but we will see how crowded things get inside.

The whole system will be circuit broken at 30 amps so that I can keep things well under the standard "C" rating for the battery cells.


Eric

One component that I have that I have not hooked up yet is a Hall Effect charge indicator.  This may be a bit superfluous as the BMS will already accomplish most of this and even connect via Bluetooth to my phone, but I like having an actual gauge on the box itself.  I will probably attach this on the top of the case, but at this point I don't know exactly where as my options are limited.  Its not impossible that I will place it at the rear in the vicinity of the light switches.  Of course, if I do this, this affects the future placement of any lights that I may place on top.  Also, I don't want to weaken the lid any more than I have to.  An alternative placement might be somewhere on the side of the box which at this point has no perforation as of yet.

Eric  

I would probably not go into that effort of adding a coulomb meter and just use a digital voltmeter that uses very little power, since you already have the same functionality in the BMS.
That way you can still get a rough idea of the state of charge, but it is not quite as complex (and expensive).

Sebastian,

Interesting thought.  Technically, the coulomb meter is for measuring state-of-charge which is challenging in lithium batteries as the discharge profile is so flat that voltage is not a reliable indicator of charge level.  I am not actually certain that the BMS has that specific functionality, but I can check in easily enough.  I already have the coulomb meter and it was not at all expensive so that part doesn’t bother me.  I do, however, have reservations about cutting unnecessary holes into the case.

I will look into things and get back to you.

Eric

Ok, it’s been a busy couple of weeks so I have not gotten much done on my battery box yet.  Also I have been waiting for a couple of odds and ends to come in.  I will not have a ton of space once I put in the battery pack and that makes putting in a fuse box difficult.  So instead, in an attempt to minimize lots of extraneous wires in a limited space, I plan to make all of my connections connect to a two posts (one positive, one negative), and then each post will connect to main power.  There will still be some internal fuses and circuit breakers, but these will be inline, automotive types that I will deliberately wire for easy access.  I decided that since I was using a post connection, it had better be the best connection that I can get so I ordered a pair of copper bolts to use as the main posts.  All connections will connect at these bolts before connecting to main power.  I also ordered a pack of copper washers to place between ring connectors in case the connectors can’t make a solid connection to each other.  

Here is a picture of my copper bolts/main terminals.



Eric

I got up early this morning and started wiring the internal connections in the box.  Most of the wiring is now done, but the exact placement may change just a little bit in an attempt to tidy things up.  Overall, I tried to make all of my positive lines all connect on the right side to a common terminal—a copper bolt.  And likewise, the negative connections line up on the left to another copper bolt.  Each bolt then connects to the main circuit of the battery.

Here are some pictures:

Eric

Ok, I will try to explain what is going on in this (mess) picture!

Up near the front I have all of the front port connections, notably USB ports, but also Anderson Power Pole and XT60 connectors.  The XT60 will be the main charging point.  All of those positive connections are routed to the right, turn back, and loop back up to the front to get an automotive style fuse.  From there they lead to a common connection point which is the copper bolt.  At the moment I have not yet decided how I want to connect the bolt to the box.  I have thought about putting a little bracket in and then bolting it to the side, but that would leave a live electrical connection on the outside.

At any rate, the top wire then runs through a long loop and goes through a 30 amp in-line circuit breaker.  From there the line goes to the center switch (rated to 30 amps).  The  line coming out of the switch then connects to the battery positive (at the moment, it is not in place).

The other side of the battery (negative side) gets more complicated.  The Battery Management System (BMS) sits on the right side of the battery positive.  There are a series of thin wires that connect the BMS to each battery cell.  This is so that the BMS can read the exact voltage of each cell and charge each cell to get them individually back up to full charge.  

In the back side of the BMS, there are two main cables.  The one on the left connects to the main switch.  The one on the right connects to the main loops around and connects the main negative terminal.  That negative terminal then has several connections that branch out to outside electrical connections.

At the moment, this is all roughed in, a very loose fit.  All of blue painters tape is strictly temporary and will eventually come out.  Its purpose is to temporarily hold things in place while I cut wires to length.

The orientation of the BMS is not as I had originally intended as the little BMS cell monitoring cables are terribly short.  In fact. For a moment I wondered if I even had enough length to attach at all.  I eventually flipped the BMS around and there is plenty of cable length, but now the main power cables from the BMS were two short and pointing in the wrong direction.  Fortunately, lengthening the power cables is easy and with a little bit of maneuvering I had things in place.

I still have a few things to tidy up.  I need to shorten some cables and tie other cables together to make things look better, but I think it is coming along.

Thoughts?

Eric

Yours almost looks professional.
Instead of lengthening the power cables, you could extend the sensing wires. Less expensive and doesn't add (almost any) weight.

I am just using standard solid wiring (probably 12 AWG), which should be fine for the peak load of 27A and which will liquefy at 200A, acting as a fuse.

Sebastian,  

I wish I could lengthen my sensing wires.  I was a little surprised when I found just how short they were.

And thank for the almost-professional comment.  Right now I think the internals are a bit messy.  Now before I finish I will clean this up, but I am still making adjustments.

By the way, what do you think about the copper bolt-as-common-terminal idea?  And, how would you attach the bolt?

Eric

The copper bold is a nice touch, but I think a steel bolt would suffice. The washer part of the terminals have more than enough surface area to conduct the current up and down the stack.
As to mounting it, if you don't want it floating around, maybe a large patch of double sided tape holding a block of wood that acts as a mounting point for the bold? It could even have space for the bold-head countersunk on the backside.

Sebastian,

Not a bad idea for the bolt mount.  I might try something like that.

Eric

Doing this project really added to my collection of small electrical connectors.  Keeping track of all of these connectors was becoming difficult to impossible so I took steps to organize my things.  This really is a necessity for me as I get tools that seem to grow legs and walk off on their own.  A part of this is my general propensity to be perfectly happy in a disorganized environment—until I need something and can’t find it!  Another reason is that each member of my family has at times used my tools but then put them wherever they thought appropriate.

My wandering tools, along with a scattering of miscellaneous connectors made it necessary to corral all of these items—which to me is like herding cats!  But my solution was three-fold.  First and foremost, I purchased several 5.56 magazine plastic ammo cans, the same type that serves as the housing for my battery box.  I like this size as it is big enough to stash it bunch of stuff without being so big that I will lose tools in them.  Further, the construction is surprisingly sturdy for being plastic, has two sturdy locking clasps and even has a water resistant seal in the lid.  They cost $40 on Amazon (and everywhere else I looked), but I consider that worth it if I can avoid losing tools.

Secondly, I picked up a little tray organizer on a random trip to Menards one day.  This was great for storing large numbers of item that came in small numbers (such as there being one tray portion devoted to switches—two of them).  Finally, I have also purchased a collection of plastic jars with screw on lids for items that might come in the hundreds (ring terminals).  

All in all, I have really cut down the clutter of my work station.  I especially like the ammo cans as they can really hold quite a lot, but are both highly portable (they have good handles) and they take up little space when not used as they are stackable!

Thought I would share.


Eric

MAKE SURE TO INSULATE BETWEEN THOSE CELLS! Any movement between any of them will cause that thin blue layer to rub through and then the cells short in-between themselves. Yep I found out the hard way and it cost me! They also have to be compressed or they will start expanding with use and cause other shorting issues, again learned through experience!  

Larry,

Will do.  Thanks for the advice.

Eric

Ok, progress!!

At this point I can get basic functionality out of my front panels.  The wiring is starting to get crowded but it is workable.

I still need to install a master breaker which will be 30 amps.  That is a fairly simple addition to what I already have in place.

In addition to the main circuit breaker, I also need to install the lights and set up and configure the BMS.  I also need to do an initial charge.

But at the moment, I have basic circuitry working.

Eric

Ok, it’s been a while since I last posted.  Today I learned a humbling, frustrating lesson.

As much as I love these projects, the components often leave something to be desired.  In this case it was the BMS—Battery Management System.  The BMS itself is fairly decent quality (Daly), but the instructions are pretty abysmal.  Adding to that, they are written in Chinese and translated (very poorly) into English and sometimes leave out important details.  For instance, the BMS is designed to communicate via phone through an app.  Ok, that’s simple enough, but the App Store has several apps for this specific purpose and there is no mention of exactly which app I should use.  Naturally I use the one at the top of the list.  

That app in fact recognizes my device and even looks like it’s going through a connection/uploading/downloading process—that never ends.  I tried adjusting various parameters on my phone but to no avail. Watched several YouTube videos but they all instantly connected and then demonstrated some other problem.  They did all keep mentioning going back to the App Store and I saw that the little icon looked slightly different than what I downloaded.  So I went back to the App Store and downloaded a third, ostensibly identical app and BAM!  It worked.

If anyone at home tries this, just keep downloading the apps.  Maybe even download all and figure out which one is the proper one.


Eric

Hi Eric,

How did this project turn out? It looks like fun. More than what I need right now but someday. Any Idea what it has cost so far?

I'm looking for a unheated portable 12v replacement LifePO4 and found this thread. I know Battle Born are supposed to be good but $$$. I can get 5 with the same 'specs' on amazon for the price of one Battle Born. I also wouldn't mind having a BMS that I could replace. Maybe a new thread is in order.

Hi Les,

So I am not *Quite* finished yet.  I just need to make a couple of finishing connections and then I will have it finished.  But I get this way sometimes.  I get the majority of the project finished and then I don't want to finish because then the project is done and I have no project left!

But that is no excuse.  The reality is that by now all of the electrical connections work except for the lights which I have not wired in yet--that will come very shortly.  Overall I have really enjoyed the project.  One note:  I originally purchased the batteries from Ali Express as they were the only supplier of 80AH batteries.  But they were out so they gave me 100AH batteries.  In the meantime, the order was taking so long and I had no communication, nor would they answer my emails that I thought they took my money and ran.  I bought again from Amazon for a trivial amount more.

And then they came--all at once!

The Amazon came on time as stated.

The Ali Express came--but I ended up buying two sets as I didn't think that my first purchase went through.

So now I have a total of three sets of 100AH, 12V LiFePo4 cells.

This makes me wonder what I will do with the other sets.  One option is to build another copy of my present box and give it to each of my kids.

Another option is to build a 24V cell.

I know that at some point I want to make a 320AH 24V battery.

We will see what happens!



Eric

Thanks, Eric, for posting your build here as it provides not only good detail and education for those like me looking to do the same, but also inspiration to take the plunge.  As you noted, it seems like new sources of (vendors) of components are coming online daily and although one needs to be cautious with fidelity of vendor an'd quality of product, it's nice to have options.  

I think it was Larry K or someone elsewhere noting that the BMS, under proper circumstances, *could* be mounted outside of the box....something I've been considering for batteries that will be largely stationary and indoors.  One thing I'm wondering about in this regard that may also address your 'extra cells' issue:  Can an external BMS that is designed for, say, 16s be used externally across two boxes of cells, each containing 8 cells?  It appears that most builds attempt to ensure that the series terminal connections are all of the same length and 'resistance'....but can this rule be broken to have two different packs.  Would the wires from the BMS still be able to reach all of the cell terminals necessary or does it need to be a pretty tight fit?  Again, thanks for posting your efforts here......very helpful!

Regarding the external BMS.....

That is going to be tricky.  If we were talking about lead chemistry, this might work.  But Lithium's great strength is also one of its great weakness.


Lithium is known for having a flat voltage discharge rate.  This is an excellent quality when powering a device.  But the problem is is measuring the amount of electricity that has been given up for that power.  Voltage is useless.  Is the battery 80% full?  Is it 20% full?  Voltage does not really help as the voltage discharge is so flat that the discharge tells us nothing.  Instead, for Lithium we need a BMS (or a small charge status indicator, but that's a different point).  The BMS "counts" the charge leaving the battery and then subtracts that figure from the the total capacity of the battery.  Really, a battery build like this one should be tested--charge to maximum followed  by a discharge to about 30% (max safe discharge), calibrated and then done again.  This is a time consuming process but gives you a good idea of just how much in your battery.

If the same BMS were used on two or more battery packs, it would never get used to just how much is going into and out of the battery.

I hope this does not burst your bubble, but maybe it helps.


Eric

Eric Hanson wrote:

Lithium is known for having a flat voltage discharge rate.  This is an excellent quality when powering a device.  But the problem is is measuring the amount of electricity that has been given up for that power.  Voltage is useless.  Is the battery 80% full?  Is it 20% full?  Voltage does not really help as the voltage discharge is so flat that the discharge tells us nothing.

This doesn't seem to be true for the lithium batteries in my ebikes. I refer to the following chart quite frequently before I charge or store a battery.

There are at least 6 different kinds of lithium batteries so maybe in some cases voltage is useless?

Ok I'm decoding this voltage thing for myself. The voltage chart above is for Lithium Ion batteries whose voltages are a bit higher and drop off more, faster,
than LiFePO4 batteries.

During charging or discharging it is not accurate to determine state of charge from voltage. After a rest of 30 minutes or so voltage should give you a pretty good idea of state of charge with Lithium ion. LiFePO4 stays much more flat overall but mostly in between 100 and 75 percent state of charge using the voltage chart I found below.

The balancing part of the BMS needs the voltage of every cell in order to balance them. So accurate that somewhere I read the wires going to each cell have to be the same lengths no matter how far from the BMS each cell is to keep the voltage readings accurate.

Here is a state of charge voltage chart I found specifically for LiFePO4 batteries.

Now I apologize for hijacking this thread in my confusion and it seems another about the different lithium batteries might be in order.

Hoping other's projects here are moving along positively.  A few questions as I'm sloooooooowly assembling some components.  One question refers to BMS specs that seem confusing although I may be misreading something.  Most LiFePO4 assembled batteries are accompanied by the notation of, for example, 4s4p, meaning in this case that such a battery can be connected to up to 3 other similar batteries in series configuration (for higher voltage) and 3 other batteries in parallel configuration (for higher amp-hour rating).  However, the BMS is also rated as being 4s, 8s, 16s, etc. denoting the possible number of cells within the battery that can be series connected to obtain the voltage of that battery pack.  What is not clear to me is whether or not it is possible to have 4 series connected cells plus a second set of 4 series connected cells (all of same cell type/age) in the same box with the two different "4-packs" connected to each other in parallel to raise the amp-hour capacity of that final single battery.  When regarding just the cells themselves, can a single BMS be configured to control both series connected cells and parallel connected cells within one battery?  Thus, the outcome would be 8 cells (3.2V) divided into 2 "4-packs" (series connected) in which those 2 packs are connected in parallel to double the amp-hour rating of the battery.  All of this is to try to maximize space utilization with prismatic LiFePO4 cells.  Thanks!

Edited to add....Eric, looking forward to your update!

John,

Thanks for updating this thread and kicking me in the butt to get things finished!! I kinda got distracted.

But back to your question, I am afraid that the BMS is very particular about its configuration.

Here is the deal:

My BMS is a 4S type.  This means that it connects to each cell individually in order to read each individual cell voltage which is absolutely critical in this type of system.  The BMS will—to a degree—try to balance the cell voltages to get them exactly identical.  They will sorta take energy from a higher cell and send it to a lower cell.  This can happen passively in the background as is the case with mine.  Active balancing does this much quicker.

But if you doubled your cells in parallel and connected each pair to the BMS as though it were one battery, the BMS would not know which cell it was talking to.  In fact, it would just assume that there was only one battery.  This becomes problematic when charging and discharging, especially to the extreme operating ends (near fully charged and nearly fully discharged),  In this case, the BMS may allow charging when at the max point because the less-charged cell would lower the pair’s overall voltage.  The higher charged cell would still get overcharged.  And near the empty point, the BMS may allow further discharge because the higher cell raises the overall voltage and the lower voltage cell drops into dangerously low to a state where it can no longer be charged again.

If you want to go with a 4s2P system, those BMS devices are easily available .

Does this help?





Eric

Les,  you nailed it.

John Weiland wrote: When regarding just the cells themselves, can a single BMS be configured to control both series connected cells and parallel connected cells within one battery?

A BMS is sold by the number of series cells it can control, and an amperage output rating. They never know how many parallel cells there are (short of fancy high end ones that ask you to punch in numbers, they don't *know* otherwise) and have no way of controlling them. The post earlier talking about the BMS "knowing" the battery state of charge must be talking about a particular model with a coulomb counter and some sort of display or Bluetooth connection to give a percentage. Most cheap BMSs don't have any of that.

Because they don't know any better you can configure numerous packs In to one large logical pack. With high output cells it's common to use some sort of cell to cell fusing so one bad cell doesn't take the rest of the parallel string with it.

Part of the reason you tend to see overkill connects on battery builds is usually because people are trying to push their build to extremes. Your wiring only has to be as heavy as your largest load demands, and fusing appropriate for the wire. If you only used very light loads, the voltage drop is extremely low, and you won't end up with much difference even if you added a few cables 10x longer than needed. At that point the cost per length is a factor. A heavy enough load will cause more voltage drop and could cause more drift between cells if one or more of the connections convert energy in to heat. Then there is managing cell drift once it occurs.

I have the original battery to my e bike which I got cheap because the tide came in at the beach and messed things up. The cell configuration went down one side then up the other, with cells at the bottom (pack middle) corroded. I lopped off 3 sets of cells (6P) to convert it from 13S to 10S, used thicker wire to bridge the gap, and mostly use it on lower power draw stuff like my hedge clippers or DIY slow cooker. A lot can be done with creative thinking if your goals aren't too demanding.

Daniel,

Nice example.  

Regarding the fancy models--that is actually what I have connected to my system.  The BMS has leads that connect to each battery cell to measure each cell individually.  The BMS connects to my phone via Bluetooth where I can see the exact voltage of each cell.  And by exact, I mean to the 1/1000 volt.

But I am also experimenting with a concept design so I might be overdesigning it.


Eric

Eric H. wrote: "My BMS is a 4S type.  This means that it connects to each cell individually in order to read each individual cell voltage which is absolutely critical in this type of system. "

I understand this 'side' of the designation for cell connections, but once you have the battery completed, isn't there some additional designation for the use of that new battery build to participate in a string of batteries...either in series or parallel connection?  Thus, if your battery is 12V and you want to combine it with other similar builds for higher voltage (series) or higher amperage (parallel), does your BMS come with a designation for how many batteries can be in either of those strings? Or is this not a function of the BMS?  When I see commercial batteries for sale, they typically have a designation like "4s4p" and I was unsure as to how this might related to the BMS's control of the internal cells vs the battery string.  And I totally understand how if the BMS is not "seeing" those series-connected cells properly, it won't be able to keep them balanced.

I think the rest is clear....I would need a more sophisticated BMS if I wanted to safely control/balance cells that are combined in different packs, but within the same battery case.  Thus, if the goal were to combine 4x50 Ah cells (series) in one pack and 4 cells in a second pack (series)....and THEN connect these two 12V packs in parallel to obtain a final 12V 100 Ah battery, that may require a fancier BMS, correct?  I may in the end just opt for a build similar to your own, with 4 cells positioned on their side.  I'm narrowing down now which cells I hope to purchase for the first attempt. And definitely want to get a BMS with active cell balancing and Bluetooth.

Thanks again for this and all other respondants...please let me know if my thinking is wonky here.  Hoping to get started once I can open up my workspace to 'true' spring temperatures, which still are a few days away!

John Weiland wrote:Eric H. wrote: "My BMS is a 4S type.  This means that it connects to each cell individually in order to read each individual cell voltage which is absolutely critical in this type of system. "

I understand this 'side' of the designation for cell connections, but once you have the battery completed, isn't there some additional designation for the use of that new battery build to participate in a string of batteries...either in series or parallel connection?  Thus, if your battery is 12V and you want to combine it with other similar builds for higher voltage (series) or higher amperage (parallel), does your BMS come with a designation for how many batteries can be in either of those strings? Or is this not a function of the BMS?  When I see commercial batteries for sale, they typically have a designation like "4s4p" and I was unsure as to how this might related to the BMS's control of the internal cells vs the battery string.  And I totally understand how if the BMS is not "seeing" those series-connected cells properly, it won't be able to keep them balanced.

I think the rest is clear....I would need a more sophisticated BMS if I wanted to safely control/balance cells that are combined in different packs, but within the same battery case.  Thus, if the goal were to combine 4x50 Ah cells (series) in one pack and 4 cells in a second pack (series)....and THEN connect these two 12V packs in parallel to obtain a final 12V 100 Ah battery, that may require a fancier BMS, correct?  I may in the end just opt for a build similar to your own, with 4 cells positioned on their side.  I'm narrowing down now which cells I hope to purchase for the first attempt. And definitely want to get a BMS with active cell balancing and Bluetooth.

Thanks again for this and all other respondants...please let me know if my thinking is wonky here.  Hoping to get started once I can open up my workspace to 'true' spring temperatures, which still are a few days away!

The issue you're up against, is that most of the commonly available BMS on the market are not designed to communicate with other BMSs, and operate as if they're autonomous/stand-alone.  There are some that can, and those tend to be on the higher price range.  Most consumer grade BMS won't have any way to network, and/or compare capacity, so connecting 2 or more batteries in series (or parallel), means that your weakest link doesn't go away, but might cause a failure of the entire system with little or no warning.

For my "DIY" powerwall system I built, the BMS itself doesn't actually communicate with the other BMS (I think?), but they're networked through an interface board that allows them to have different addresses so they can be daisy-chained and the charge controller can talk to and see both of them at the same time, and IT manages their charging/discharging, faults/warnings, etc.

These are the ones that I have used, although their stocking levels vary widely:  https://www.basengreen.com/product/diy-case-for-48v-battery-packenclosure-accessories-usa/

As sold, they can be ganged up to a total of 16 units in parallel and each individual BMS is only responsible for keeping the cells inside its box balanced/charged/etc..  I started with 2 of them here:

and added 2 more later:


In this example, the first inverter/charger (closest to the forefront) is the master controller and controls both the 2nd inverter/charger, and handles battery communication.  The communication is fairly basic, as I didn't choose to go with the same manufacturer/brand of inverter/charger and battery system, but it works well enough to make happy with the results.

On smaller scales, I have seen other BMS that were truly designed for packs made up of multiple banks of series-connected cells, but I don't remember the vendors for those, as I've not been active in my dormant hobby of robotics since before COVID, and I fear many of those places are long gone now.

Barring a true multipack BMS, one would need to have a separate BMS for each string of series-connected cells to be safe, and thereafter, they might not really need to communicate between banks with each other, as each would at least address the individual cell safety concerns sufficiently.

Hope that adds clarity to the discussion?

If memory serves, I think Andy on the YouTube channel OffGrid Garage does test some "Frankenstein" combinations of batteries, and IIRC, as long as there were BMS running on all of them, with direct cell monitoring, they seemed to play nice with each other, and there was at least protection when one battery or another ran out of charge before the others - nothing bad happened as a result, because the BMS halted the discharging when needed, even if it meant not getting all the energy out of the battery in question.

John Weiland wrote:Eric H. wrote: "My BMS is a 4S type.  This means that it connects to each cell individually in order to read each individual cell voltage which is absolutely critical in this type of system. "

I understand this 'side' of the designation for cell connections, but once you have the battery completed, isn't there some additional designation for the use of that new battery build to participate in a string of batteries...either in series or parallel connection?  Thus, if your battery is 12V and you want to combine it with other similar builds for higher voltage (series) or higher amperage (parallel), does your BMS come with a designation for how many batteries can be in either of those strings? Or is this not a function of the BMS?  When I see commercial batteries for sale, they typically have a designation like "4s4p" and I was unsure as to how this might related to the BMS's control of the internal cells vs the battery string.  And I totally understand how if the BMS is not "seeing" those series-connected cells properly, it won't be able to keep them balanced.

I think the rest is clear....I would need a more sophisticated BMS if I wanted to safely control/balance cells that are combined in different packs, but within the same battery case.  Thus, if the goal were to combine 4x50 Ah cells (series) in one pack and 4 cells in a second pack (series)....and THEN connect these two 12V packs in parallel to obtain a final 12V 100 Ah battery, that may require a fancier BMS, correct?  I may in the end just opt for a build similar to your own, with 4 cells positioned on their side.  I'm narrowing down now which cells I hope to purchase for the first attempt. And definitely want to get a BMS with active cell balancing and Bluetooth.

Thanks again for this and all other respondants...please let me know if my thinking is wonky here.  Hoping to get started once I can open up my workspace to 'true' spring temperatures, which still are a few days away!

I was digging into this further and I think I can help mesh the current state of available tech with a better explanation of how it works.

In the multi-pack batteries, the better/safer ones will use what's called "protected" cells {this is primarily used for lithium polymer 18650 cells, but can be done with the pouch cells too}. For that, each cell will have a tiny protection circuit board to guard against the basic worst case scenarios, like excess current, over voltage, over discharge, etc, where the purpose is to just disconnect the cell if threatened.

Everywhere that will sell bulk lithium polymer cells I've shopped at, offers the unprotected cells with a warning and disclaimer about needing to know what you're doing, if building packs, and most of them also sell protected cells too. With protected cells there's much less hazard using them in packs, and one can get away with using a single BMS on something like a 4s4p, even though each cell won't get a monitoring lead. The downside is that with time/use, the cells probably won't stay balanced, and the pack will wear out prematurely.

The biggest downside to protected cells, is that they tend to have much lower peak current capacity than unprotected cells, so you're gonna need to consider the peak current demand of the application/load, to insure you can even run it (some things won't even start if their peak/starting current demands can't be met, and those that do may cause enough of a voltage sag it could be destructive too). Actually, you should plan for/allow for the peak current goals in the design stage, no matter which parts you end up using, or you might be wasting your time and money.

On commercially available consumer gear (especially on the cheaper ones) , most manufacturers don't care about that, it just means they can sell replacements sooner, & boost their profit margins. On the top end of the scale, there will be individual cell monitoring, and also active balancing, to keep all the cells as close to each other - all the time.

One way you can get close to the same results, is by choosing your BMS to include Bluetooth communication for management of the battery, and as you add others in parallel, using the same model BMS will give you the ability to monitor the entire battery - all the banks, even though you might have to switch between them on the app by serial number. At least with LiFePO4 batteries, the age is much less of a factor, and you can get away with adding new batteries to an older battery system.

I can suggest/recommend JK BMS, if you are building from scratch, as those are well made, decently capable, and maybe more importantly, they have very good compatibility with lots of inverters, if/when you get to a point where you need that.

Hopefully this is no longer clear as mud now

If you were to go completely rogue, and just wire the cells together as a pack, depending on which battery chemistry, it might cause accelerated wear on the pack (LiFePO4) or end up causing a fire as any one of the cells got overcharged with no monitoring or intervention (LiPoly/LiMnCo).

On the original Tesla battery design, there wasn't any cell monitoring of the hundreds of 18650 cells in the pack, but they did have a liquid-cooled jacket to keep them at a stable temperature and each of the cells were connected to the system arranged in a grid (series AND parallel at the same time), with a small wire that acted like a small fuse, so any cell that got overzealous enough, would take itself out of the equation and it wouldn't change anything but the overall capacity of the pack until/unless all the cells in the same parallel row failed. I don't know if the current generation has retained that design or not.

Allen Jackson wrote:....... At least with LiFePO4 batteries, the age is much less of a factor, and you can get away with adding new batteries to an older battery system.

I can suggest/recommend JK BMS, if you are building from scratch, as those are well made, decently capable, and maybe more importantly, they have very good compatibility with lots of inverters, if/when you get to a point where you need that.

Excellent continued discussion and I thank you Allen for your time and expertise on my questions....hoping all of this is of use to the OP as well!

I've got components lined up now hopefully for purchase this week (contingent on the income tax equation for previous FY not being too brutal!).  One question now pertains to the fact that I'm looking at some 100 Ah EVE cells that are 'unprotected'.  These cells have a max continuous discharge rate of 1C, but I'm not finding a max 'surge' discharge rate.  That said, I was looking at a Daly BMS that has a 200A cut-off and now am unsure if this is an unsafe configuration for these unprotected cells.   If I don't know whether the cells can tolerate a 200A burst discharge for a few seconds, do I need a BMS with a lower over-current cut-off limit?  With this first time build, I'm hoping to get my feet wet experimenting with the cells, the BMS, the organization with the chosen case (ammunition case) and all incidentals like bus bars, cell spacer and wrap sizing, and ultimately, getting it all to work properly.  For this first build, I hope to skimp a bit on the BMS capability (no Bluetooth)  Later, I hope to up the game to Bluetooth app monitoring of cells/battery, possibly even jumping straight to 48V build at that time.

As you and others have noted, it's fairly crucial to work backwards from the peak load one wishes to power.  In the case of this first build. I only hope to use it for portable 12 power....powering 12V lights, trail cameras, etc. wihtout an inverter, but adding an inverter for items that would not necessarily need large surge amperage.  Later builds or configurations would be more thoughtfully designed for powering well/irrigation pumps, sump-pumps, etc....using inverters for DC-AC conversion.  And also to add that I have kept my eye on the JK brand BMS and hope to use them in a Bluetooth configured build later on....I currently love using Bluetooth apps to monitor solar controllers, battery states via BMS (commercially purchaed batteries) on golf carts and off-road EVs.  Hopefully looking forward to beefing up the off-road EV (48V) to Ah capacity to assist in emergency home power provision through a current transfer switch on our utility power pole....would really like to relegate the petrol generators to other reduced duty at this point.

Anyway, all a great discussion and I'm learning so much now.  If possible, could you please expand on your line above that I placed in bold text:  I was under the impression that string connection of LiFeP)4 batteries was best done with batteries all of the same make, same age, same state of charge, etc. was crucial for tolerable functionality.  But your comment suggests that, with proper monitoring, older LiFePO4 battieries might be added to a series or parallel string???...   Truthfully, I'm thankful for being alive at a time when some of this has become so much more plug-n-play than earlier.  Still a lot of learning and caution needed around the power being generated, but solar right now, when paired with the new storage paradigms, is just mesmerizing me with its capabilities!  Thanks again!...

Specifically, adding a new bank of cells to run parallel with an older bank of cells (as long as each bank has their own BMS), doesn't present the dangers that the manufacturers warn about when they caution to not mix batteries of differing ages. That's mostly a CYA for warranty claims, if you started each batch with new cells that were matched when they started their service.

That's something that Andy tested on the OffGrid Garage, where he debunked the claim of doom, gloom, and self-destruction of the battery banks if one failed to follow those warnings.

On a different note. I don't think there is such a thing as a protected LiFePO4 cell - at least I've never seen any, and it's not really needed for those type of cells, because the major risk is prematurely killing the battery vs. spontaneous combustion of a sort that requires a class D fire extinguisher to put out (most folks have never seen or heard of those - they're white and very expensive).

The protected cells are pretty much limited to the fire-hazard chemistry cells, to reduce the legal liability of selling them to folks who don't know better... OK, so I'm waxing cynical - that's to reduce both the risk of fire and the potential damage of thermal runaway, especially when used for something like a vape device...

The peak current output of that model cell (EVE 100Ah) is generally tied to the warranty offered by the vendor and/or the expected lifecycle of the cells, but you can get away with occasional short peaks of 2C (200 amps) for a second or so, and possibly even 3C momentarily, but the limiting factor is how fast the cells are heating up, and possibly delaminating from the rapid heating. Keeping them both cooler and under compression will go a long way toward prolonging them if you are expecting to abuse them, but even then, you're talking about the difference between 6000-8000 cycles and 3000-5000 cycles maybe - unless you just start welding with them in a near dead short.

Short (2C) bursts up to 1 second shouldn't be a problem. EVE does publish the data for those cells, but I don't read Chinese and don't have a quick reference to the translated version of that model on hand right now. I do know they've become more conservative over the past couple of years, as their marketing claims of longevity have risen and the product gets older, so there's more real data collected about their durability, but I'm personally willing to baby my cells and only use the top 50%,just to get 10K-12K cycles from them - by then, there will probably be a better battery choice anyway (10+ years down the road).

OH and, don't change your BMS over that limit unless you know that you're going to be wringing the life out of your cells frequently - the cells are pretty tough and you're more likely to need/appreciate the headroom later.

Allen Jackson wrote: but I'm personally willing to baby my cells and only use the top 50%, just to get 10K-12K cycles from them...

Should read "but I'm personally NOT willing to baby my cells and only use the top 50%, just to get 10K-12K cycles from them"