diy solar

diy solar

Building a 24 V battery bank: Using 12V batteries is much cheaper, but is it worth it over using 24V batteries?

BjornM

MacGyver's apprentice
Joined
Oct 7, 2023
Messages
157
Location
Sweden
I'm building a 24V 400 Ah LiFePo4 battery bank. My first choice would be to get 2 x 24V 200 Ah and do 2p. I've found a unit that looks good at LiTime.

Then I saw that 4 x 12V 200 Ah in a 2s2p config is a lot cheaper. But is it worth the savings? Then I will have the potential risk of getting unbalanced batteries over time. Not sure about how much of a problem this is in real life. LiTime make it seem like a non issue. They even have examples of 4s4p configs without using any external battery balancers.

If I make sure to do some battery maintenance once a year, will that take care of it? I'd fully charge each of the four batteries separately.

Separate question: if you do 2s2p, what are the pros and cons of connecting the two midpoints together? What I mean is you first do two pairs of 2p that you then connect in series, so technically I guess you call it 2p2s.
 
if you do 2s2p, what are the pros and cons of connecting the two midpoints together?
I wouldn't connect the midpoints of 2 2S batteries.

When wired properly, the difference between 2S2P and 2P2S is negligible. I would read the fine details of each BMS and see it there are limitations that they present.
 
$2120 vs $2970, so $850 more.
Why not really build your own and save another $500-$1000 while having more capacity and a great appreciation of being able to easily troubleshot?

Do you have regulations that require a pre-built battery vs raw cells?

Do you already have a 24v inverter? Why not 48v?
 
I just click on the 12v 400amp batteries vs the 24v 200amp batteries and the 12v 400amp is more expensive. are you sure you are comparing like for like?
View attachment 171233
Thanks for helping me out. I'm in Sweden, so I went to their German shop.

Here is the 12V 200 Ah, pack of four, 2000 EUR = $2120


Here is the 24V 200 Ah, pack of two, 2800 EUR = $2970


Another option would be 24V 100 Ah, pack of four, 2470 EUR = $2620

So the question remains: will balancing be a problem or not, and if not, I'll get the cheapest option. Space is no problem. Extra cables no problem.
 
An active balancer should keep things under control, only downside is they all seem to review as being audibly noisy in operation. $850 is quite a price difference.
 
Why not really build your own and save another $500-$1000 while having more capacity and a great appreciation of being able to easily troubleshot?

Do you have regulations that require a pre-built battery vs raw cells?

Do you already have a 24v inverter? Why not 48v?
I listened to Will Prowse's YouTube channel, where he said that the Chinese budget 12V/24V packs are now getting so cheap it makes no sense to build from raw cells anymore.

But for sure I'd be willing to do that if it saves me money. I have an electronics background. I'm just not familiar with what brands of raw cells are good, what BMS is good, etc. I suspect that there are only a few of factories making these, I just need to learn where to source them. If you could give me some pointers it would be appreciated.

My setup if very simple. No need for heaters or low temp charging cutoff, because the batteries are indoors at 25 C/77 F. I have a 24 V system (inverter and other equipment).
 
I listened to Will Prowse's YouTube channel, where he said that the Chinese budget 12V/24V packs are now getting so cheap it makes no sense to build from raw cells anymore.

But for sure I'd be willing to do that if it saves me money. I have an electronics background. I'm just not familiar with what brands of raw cells are good, what BMS is good, etc. I suspect that there are only a few of factories making these, I just need to learn where to source them. If you could give me some pointers it would be appreciated.

My setup if very simple. No need for heaters or low temp charging cutoff, because the batteries are indoors at 25 C/77 F. I have a 24 V system (inverter and other equipment).
Post 5 and 7 should give you more info.

I'm pinging @Steve_S to provide contact info for Amy, I think this would be your best route for your location.
 
An active balancer should keep things under control, only downside is they all seem to review as being audibly noisy in operation. $850 is quite a price difference.

which balancers are those? I have not experienced noise from HA01/02 balancers.
 
Building your own battery packs is not hard, a lot of us do that here.
We always recomend that you use Native Voltage Batteries, ie 12, 24 or 48 V. Placing Lithium Based batteries in Series more often than not causes issues.
Using Battery Packs in Parallel increases your AH, the Fault tolerance by providing fallback in the event another pack in the "Bank" and in addition to that Packs within a Battery Bank divide & share both load & charge which reduces strain on the system

The rules for Paralleling Packs into a Bank are fairly simple.
Keep wires from Battery Terminals to Common DC Busbars the same length. Keep them together as much as possible.
Fuse each battery pack for the appropriate Amperage, taking into account possible Surge demand Handling.
They can be placed on Shelves, onto a Rack, or shelves, Do keep them off the ground. *Northerner Tip !

! Building a Battery Pack with Paralleled Cells within is generally NOT advised unless you can be certain that you have properly fully Matched & Batched cells (not just Voltage & IR matched as most vendors do). It is CRITICAL for a properly performing pack to have cells which have a consistent IR (Internal Resistance) throughout the entire "Working Voltage Range" otherwise paired - mismatched cells can & often do cause issues... Consider that in an EV, the cells MUST pbe perfectly matched, for a reason... There are some "geniuses" that argue otherwise, be cautious, they can make compelling hooey up & do so. See Voltage Charts Resource below.

FACTS to hang onto...
LFP has an Allowable Voltage which is from 2.500->3.650 Volts per cell. This is "the" voltage range where no harm/damage is caused. This is no different than any other battery chemistries "Allowable Voltage", od course every chemistry is a bit differewbt.
The "Working Voltage" range which is what every manufacturers tests & is responsible for delivering the Actual Specified AH, This Working Range is from 3.000-3.400 Volts per cells - The Flat Voltage Curve.

* Believe it or not, every degree Celsius does make a difference in how cells charge & discharge, as well as their SOC (slightly) and this is perfectly normal with this chemistry, LYFP & LMFP (both $$ variants) do not... FYI, those chemistries will "start" to become publicly available in Q2-2024. I spent 3 years doing various evaluations & tests while keeping the battery packs @ 10C, 15C, 20C, and with various "shelf" configurations. On average there was a 1.5C difference between each pack on a stacked shelf, with the lowest temp near the floor, this does have a negative effect which can be observed, as the battery packs will try to hold slightly different voltages. I now have 6 Packs in Parallel on one shelf @ 92cm above the floor. Battery packs performed better above 20C but below 10C, there is a noticeable decrease in AH, it does recover with temp however.
*-? My powerhouse is a hyper insulated structure with a frost protected slab foundation with R20 Foam Insulation.

BMS'. There are a great may of them out there, some good, many not bad and a few terribles. There is VAST experience here will all of those !
My suggestion is to use JKBMS with Active Balancing (Vendor below sells them as well as all other bits like Silicone wire). A general rule is 1A Active Balancing per 100AH of capacity.

CHARGE PROFILE:
Absorb/Bulk:
27.6 for 45 minutes (3.45vpc) (some call this boost)
Equalize: OFF
Float 27.5V (3.437vpc)
MIn Volts: 21.2 (2.650vpc) (Also a good cutoff for Inverter)
Max Volts: 28.6 (3.575vpc)
Rebulk Voltage: 25.6 (3.200vpc)
End Amps: (Allows for full Saturation at set Float Voltage, then as A drops to this point to trigger the transition to FLOAT mode.)
(*1): End Amps is calculated from the LOWEST AH Battery Pack in a Bank. IE: (100AH X 0.05 = 5A 280AH X 0.05 = 14A.)
NB: EndAmps = TailCurrent. Different manufacturers use different terms for the same thing.

Here is a couple of my resources you will find handy.
Luyuan Tech Basic Lifepo4 Guide V1.0A
General LiFePO4 (LFP) Voltage to SOC charts/tables 12/24/48V

Known Good Reliable Vendor who does provide EVE Factory Tested Fully Matched & Batched cells and related equipment, including battery cases & more. They have an assembly service, of course at cost extra $.
Money Saving Tip: If you order everything from 1 Vendor, ie: cells, bms, wires etc, they add it into the other shipping boxes. There is a Min Fee per box + weight etc and that adds up really fast (shippers = gougers)


You did not mention if you are using an AIO or Component build, here are two of my Logical Drawings that may help as well. These examples show Battery Bank of 4x 12V/4S but same for all voltages up to 16S/48V.

Parallel System-setup PNG.png Parallel System-with AIOs setup PNG.png

Hope this information helps and answers several questions.
Good Luck from a Fellow Northerner. (ick, it is coming and won't be nice, hope your ready for it)
 
Some reasons for building a 24v or 48v "battery-bank" out of 12vXXXah batteries, especially where the battery vendor supports such configurations:

1. quick reconfiguration if a single battery gives grief (even temporary). the more batteries in the battery bank, the better the chance to reconfigure around the problem battery, until you can return to your desired configuration. It's like having spares, and little to no downtime for any given event.

2. more batteries usually means more BMS's providing power to the (big) loads. No design survives 1st contact with the family of power users ...

In my opinion, choose batteries with:
- the highest bms amp output you can, as this is a critical limiting factor affecting you forever; I've seen 100-, 200- and 250-amp output so far
- ability to serial/parallel in many ways; verify that vendor supports any proposed battery-bank configuration.

Finally, work through the vendor's return policies, and figure out how easy it will be to return a battery, or if it is even possible. Don't be suckered by a quick, low sales price, where they are just dumping the battery on the market. Find vendors who have legit sales/service location(s) near you, as you might have to return one for a valid reason. This means you need to understand their support (email, phone, both) & return process (who pays to return, etc.). What do the forums & threads indicate for that vendor ...

With my 24v system, I now have 4 LiTime 12v300ah batteries in 2S2P config (busbars), each with 200+ amp-out BMS's. When one battery was giving me grief, and it looked like I was going to have to return it, I quickly reconfigured battery-bank around 2 batteries, and isolated the problem battery. It did get returned, and LiTime did all the heavy lifting of cross-shipping, shipping labels, etc. ... was very pleased with their response.

Hope this helps ...
 
Steve, nice diagrams. Maybe they can be a bit confusing as balancers/BMS are mentioned in your description but they are not shown or labeled in your diagrams. One diagram shows 3 leads from negative posts to bus bars, but not from a single point / BMS or if a separate balancer is not part of the BMS function.

I just read another thread where the new builder doesn't understand how to connect his JK BMS but your diagrams would not help him, as an example.

Just my suggestion as another northernish Canuck getting ready for winter.
 
which balancers are those? I have not experienced noise from HA01/02 balancers.
I have the four port version of those with lcd display, three of them. They all emit a high pitched squeal. It's not the end of the world, but if I had to keep them indoors near me, I wouldn't be happy.
 
Steve, nice diagrams. Maybe they can be a bit confusing as balancers/BMS are mentioned in your description but they are not shown or labeled in your diagrams. One diagram shows 3 leads from negative posts to bus bars, but not from a single point / BMS or if a separate balancer is not part of the BMS function.

I just read another thread where the new builder doesn't understand how to connect his JK BMS but your diagrams would not help him, as an example.

Just my suggestion as another northernish Canuck getting ready for winter.
The BMS documentation provides wiring diagrams, so follow the manufacturers docs. I've posted several as I've written a few of those manuals. There are also several diagrams provided by folks, using BMS + External Active Balancer. The JK Docs (previous version) explains the wiring for those models, the latest updated manual for the new models also does.

This is also one of my 24V diagrams, These show the B- & B= off the same cell. Most BMS' put the B+ same cell with + out.
24V Basic Prismatic Configuration.jpg
 
Last edited:
Building your own battery packs is not hard, a lot of us do that here.
We always recomend that you use Native Voltage Batteries, ie 12, 24 or 48 V. Placing Lithium Based batteries in Series more often than not causes issues.
Using Battery Packs in Parallel increases your AH, the Fault tolerance by providing fallback in the event another pack in the "Bank" and in addition to that Packs within a Battery Bank divide & share both load & charge which reduces strain on the system

The rules for Paralleling Packs into a Bank are fairly simple.
Keep wires from Battery Terminals to Common DC Busbars the same length. Keep them together as much as possible.
Fuse each battery pack for the appropriate Amperage, taking into account possible Surge demand Handling.
They can be placed on Shelves, onto a Rack, or shelves, Do keep them off the ground. *Northerner Tip !

! Building a Battery Pack with Paralleled Cells within is generally NOT advised unless you can be certain that you have properly fully Matched & Batched cells (not just Voltage & IR matched as most vendors do). It is CRITICAL for a properly performing pack to have cells which have a consistent IR (Internal Resistance) throughout the entire "Working Voltage Range" otherwise paired - mismatched cells can & often do cause issues... Consider that in an EV, the cells MUST pbe perfectly matched, for a reason... There are some "geniuses" that argue otherwise, be cautious, they can make compelling hooey up & do so. See Voltage Charts Resource below.

FACTS to hang onto...
LFP has an Allowable Voltage which is from 2.500->3.650 Volts per cell. This is "the" voltage range where no harm/damage is caused. This is no different than any other battery chemistries "Allowable Voltage", od course every chemistry is a bit differewbt.
The "Working Voltage" range which is what every manufacturers tests & is responsible for delivering the Actual Specified AH, This Working Range is from 3.000-3.400 Volts per cells - The Flat Voltage Curve.

* Believe it or not, every degree Celsius does make a difference in how cells charge & discharge, as well as their SOC (slightly) and this is perfectly normal with this chemistry, LYFP & LMFP (both $$ variants) do not... FYI, those chemistries will "start" to become publicly available in Q2-2024. I spent 3 years doing various evaluations & tests while keeping the battery packs @ 10C, 15C, 20C, and with various "shelf" configurations. On average there was a 1.5C difference between each pack on a stacked shelf, with the lowest temp near the floor, this does have a negative effect which can be observed, as the battery packs will try to hold slightly different voltages. I now have 6 Packs in Parallel on one shelf @ 92cm above the floor. Battery packs performed better above 20C but below 10C, there is a noticeable decrease in AH, it does recover with temp however.
*-? My powerhouse is a hyper insulated structure with a frost protected slab foundation with R20 Foam Insulation.

BMS'. There are a great may of them out there, some good, many not bad and a few terribles. There is VAST experience here will all of those !
My suggestion is to use JKBMS with Active Balancing (Vendor below sells them as well as all other bits like Silicone wire). A general rule is 1A Active Balancing per 100AH of capacity.

CHARGE PROFILE:
Absorb/Bulk:
27.6 for 45 minutes (3.45vpc) (some call this boost)
Equalize: OFF
Float 27.5V (3.437vpc)
MIn Volts: 21.2 (2.650vpc) (Also a good cutoff for Inverter)
Max Volts: 28.6 (3.575vpc)
Rebulk Voltage: 25.6 (3.200vpc)
End Amps: (Allows for full Saturation at set Float Voltage, then as A drops to this point to trigger the transition to FLOAT mode.)
(*1): End Amps is calculated from the LOWEST AH Battery Pack in a Bank. IE: (100AH X 0.05 = 5A 280AH X 0.05 = 14A.)
NB: EndAmps = TailCurrent. Different manufacturers use different terms for the same thing.

Here is a couple of my resources you will find handy.
Luyuan Tech Basic Lifepo4 Guide V1.0A
General LiFePO4 (LFP) Voltage to SOC charts/tables 12/24/48V

Known Good Reliable Vendor who does provide EVE Factory Tested Fully Matched & Batched cells and related equipment, including battery cases & more. They have an assembly service, of course at cost extra $.
Money Saving Tip: If you order everything from 1 Vendor, ie: cells, bms, wires etc, they add it into the other shipping boxes. There is a Min Fee per box + weight etc and that adds up really fast (shippers = gougers)


You did not mention if you are using an AIO or Component build, here are two of my Logical Drawings that may help as well. These examples show Battery Bank of 4x 12V/4S but same for all voltages up to 16S/48V.

View attachment 171323 View attachment 171324

Hope this information helps and answers several questions.
Good Luck from a Fellow Northerner. (ick, it is coming and won't be nice, hope your ready for it)

Wow, thanks for taking the time to help out, very helpful. I'm a bit overwhelmed by all the info and stuff to read through. Will write a more complete reply as soon as I have some comments/questions. My current lead acid battery bank is starting to fail after 8+ years of use. I will definitely buy a LiFePO4 bank to replace it. Only question is if it will be DIY or off-the-shelf. Not sure how many more months my current battery lasts, so I feel a bit pressed for time. As now we're heading for winter, and in Canada and Sweden it is the season you DON'T want an electric failure. My setup is very similar to your left schematic. Off-grid home with solar power. More to come...
 
Some reasons for building a 24v or 48v "battery-bank" out of 12vXXXah batteries, especially where the battery vendor supports such configurations:

1. quick reconfiguration if a single battery gives grief (even temporary). the more batteries in the battery bank, the better the chance to reconfigure around the problem battery, until you can return to your desired configuration. It's like having spares, and little to no downtime for any given event.

2. more batteries usually means more BMS's providing power to the (big) loads. No design survives 1st contact with the family of power users ...

In my opinion, choose batteries with:
- the highest bms amp output you can, as this is a critical limiting factor affecting you forever; I've seen 100-, 200- and 250-amp output so far
- ability to serial/parallel in many ways; verify that vendor supports any proposed battery-bank configuration.

Finally, work through the vendor's return policies, and figure out how easy it will be to return a battery, or if it is even possible. Don't be suckered by a quick, low sales price, where they are just dumping the battery on the market. Find vendors who have legit sales/service location(s) near you, as you might have to return one for a valid reason. This means you need to understand their support (email, phone, both) & return process (who pays to return, etc.). What do the forums & threads indicate for that vendor ...

With my 24v system, I now have 4 LiTime 12v300ah batteries in 2S2P config (busbars), each with 200+ amp-out BMS's. When one battery was giving me grief, and it looked like I was going to have to return it, I quickly reconfigured battery-bank around 2 batteries, and isolated the problem battery. It did get returned, and LiTime did all the heavy lifting of cross-shipping, shipping labels, etc. ... was very pleased with their response.

Hope this helps ...
Thanks, very helpful. Your scenario is almost exactly like mine, and it is good to hear that LiTime is easy to deal with. Right now I'm looking at a DIY solution with raw cells. The fallback will be LiTime. I'm also tempted to get packs that can handle 200A instead of the standard 100A. If one of the strings of a 2s2p fails, then 100A only gives you 2.4 kW to the inverter, wouldn't work in my case.
 
Building your own battery packs is not hard, a lot of us do that here.
We always recomend that you use Native Voltage Batteries, ie 12, 24 or 48 V. Placing Lithium Based batteries in Series more often than not causes issues.
Using Battery Packs in Parallel increases your AH, the Fault tolerance by providing fallback in the event another pack in the "Bank" and in addition to that Packs within a Battery Bank divide & share both load & charge which reduces strain on the system

The rules for Paralleling Packs into a Bank are fairly simple.
Keep wires from Battery Terminals to Common DC Busbars the same length. Keep them together as much as possible.
Fuse each battery pack for the appropriate Amperage, taking into account possible Surge demand Handling.
They can be placed on Shelves, onto a Rack, or shelves, Do keep them off the ground. *Northerner Tip !

! Building a Battery Pack with Paralleled Cells within is generally NOT advised unless you can be certain that you have properly fully Matched & Batched cells (not just Voltage & IR matched as most vendors do). It is CRITICAL for a properly performing pack to have cells which have a consistent IR (Internal Resistance) throughout the entire "Working Voltage Range" otherwise paired - mismatched cells can & often do cause issues... Consider that in an EV, the cells MUST pbe perfectly matched, for a reason... There are some "geniuses" that argue otherwise, be cautious, they can make compelling hooey up & do so. See Voltage Charts Resource below.

FACTS to hang onto...
LFP has an Allowable Voltage which is from 2.500->3.650 Volts per cell. This is "the" voltage range where no harm/damage is caused. This is no different than any other battery chemistries "Allowable Voltage", od course every chemistry is a bit differewbt.
The "Working Voltage" range which is what every manufacturers tests & is responsible for delivering the Actual Specified AH, This Working Range is from 3.000-3.400 Volts per cells - The Flat Voltage Curve.

* Believe it or not, every degree Celsius does make a difference in how cells charge & discharge, as well as their SOC (slightly) and this is perfectly normal with this chemistry, LYFP & LMFP (both $$ variants) do not... FYI, those chemistries will "start" to become publicly available in Q2-2024. I spent 3 years doing various evaluations & tests while keeping the battery packs @ 10C, 15C, 20C, and with various "shelf" configurations. On average there was a 1.5C difference between each pack on a stacked shelf, with the lowest temp near the floor, this does have a negative effect which can be observed, as the battery packs will try to hold slightly different voltages. I now have 6 Packs in Parallel on one shelf @ 92cm above the floor. Battery packs performed better above 20C but below 10C, there is a noticeable decrease in AH, it does recover with temp however.
*-? My powerhouse is a hyper insulated structure with a frost protected slab foundation with R20 Foam Insulation.

BMS'. There are a great may of them out there, some good, many not bad and a few terribles. There is VAST experience here will all of those !
My suggestion is to use JKBMS with Active Balancing (Vendor below sells them as well as all other bits like Silicone wire). A general rule is 1A Active Balancing per 100AH of capacity.

CHARGE PROFILE:
Absorb/Bulk:
27.6 for 45 minutes (3.45vpc) (some call this boost)
Equalize: OFF
Float 27.5V (3.437vpc)
MIn Volts: 21.2 (2.650vpc) (Also a good cutoff for Inverter)
Max Volts: 28.6 (3.575vpc)
Rebulk Voltage: 25.6 (3.200vpc)
End Amps: (Allows for full Saturation at set Float Voltage, then as A drops to this point to trigger the transition to FLOAT mode.)
(*1): End Amps is calculated from the LOWEST AH Battery Pack in a Bank. IE: (100AH X 0.05 = 5A 280AH X 0.05 = 14A.)
NB: EndAmps = TailCurrent. Different manufacturers use different terms for the same thing.

Here is a couple of my resources you will find handy.
Luyuan Tech Basic Lifepo4 Guide V1.0A
General LiFePO4 (LFP) Voltage to SOC charts/tables 12/24/48V

Known Good Reliable Vendor who does provide EVE Factory Tested Fully Matched & Batched cells and related equipment, including battery cases & more. They have an assembly service, of course at cost extra $.
Money Saving Tip: If you order everything from 1 Vendor, ie: cells, bms, wires etc, they add it into the other shipping boxes. There is a Min Fee per box + weight etc and that adds up really fast (shippers = gougers)


You did not mention if you are using an AIO or Component build, here are two of my Logical Drawings that may help as well. These examples show Battery Bank of 4x 12V/4S but same for all voltages up to 16S/48V.

View attachment 171323 View attachment 171324

Hope this information helps and answers several questions.
Good Luck from a Fellow Northerner. (ick, it is coming and won't be nice, hope your ready for it)

I've chatted with Amy Wan at Luyan about an order, and she is very pleasant to deal with. Offered to take any cost with import duties and tax at the Swedish border, which is a big plus, because this is a big unknown.

I'm making a 8s2p of 280Ah cells, which will give a 24V system with 13.4kWh capacity. Two BMS (one per string).

The BMS I'm looking at handles max 200A discharge, but only handles 2A equalization current, and the cells are 280Ah, so according to your rule of thumb I should use a BMS with 3A. The spec sheet doesn't mention a largest cell size. Will this work?


Grade B cells are significantly cheaper, and on top of that she gave me a discount to $88/cell. Amy says they are the same EVE model as grade A (although it doesn't say on the page), and that they will be matched for capacity, internal resistance and voltage. Any comments?

My first choice, $122/cell

Grade B, $88/cell

Then the question is whether to enclose in a box or not. I have built a shelf specifically for my current batteries, with a door that close them off from access, but still allow for air circulation. If I can get away with it, I'll just skip the box and tie the cells together with fiberglass tape and let them sit on the shelf. Tape the BMS on to the cell pack or nearby. I've seen that I have to allow the cells to physically expand, which I assume the insulation sheets will allow for? Any thoughts or tips?

Cells, class T string fuses, BMS, cables. Anything I've missed?
 
The JKbms with 2A Active Balancing is fine for the 280's, my packs all use those BMS'.
The B cells are Voltage & IR Matched and still quite reasonable at that price point. They are fine if you build a straight 8S Battery Pack with it's own BMS.
I have A Cells, B Cells, Used EV Cells in my bank. I have done numerous tests including extreme Thrash tests because of my situation, reliability is critical, especially when it's -35 outside. The A Cells certainly do perform better, with slightly faster charge because the cells are IR matched so very little balancing is required, they do give me the 280AH. My B cells are more finicky when finishing charge but it is not significant enough to be a problem. The Used EV cells are actually pretty darn good after I tweaked them. In "MY CASE" having 6 packs in a bank irons all that out and I do get 100% of my capacity... Yes, I've drained the entire bank to Low Volt Disconnect in tests.

I built my own boxes, I cannot ever recommend having the cells/terminals etc in "open air" - that's just begging for trouble! Even a Covered shelf with a clear cover in front... be safe !

I use MRBF Fuses because they are perfectly fine for <36V. These are cheaper than Class-T (as good) and make for a much cleaner install (depending on the individual of course LOL) You can see below, that the (+) Busbar is offset to allow the MRBF fuse to clear over it's (-)

Blue Sea MRBF fuses made by Eaton/Bussman (for 200A BMS, use 250A fuse)

! WARNING ! There has been a LOT of Horse Pucks tossed about Binding/Compressing cells. It's been bad enough that people could have gotten to Physical Fistacuffs over it. The cells are to be BOUND snugly to prevent their expansion from twisting the terminal blocks. Do NOT put them under pressure !!! In fact, if you build the boxes below they fit the cells "snugly" and that's enough... Alternately Amy sells the Steel Cases for a more Pro finish, she also has Stackables that sit one ontop of the other... Not shown yet I think.

H4454fb0b4ae044059f1b227cda4f87afN.jpg_720x720q50.jpg


<--- Box Link:

I built all my own boxes out of 3/4" - 19mm Plywood with a clear polycarbonate cover. See attached below. I am using Pike Industries BusBars shown below. https://pikeind.com/industries-busbars/

24v-280ah-jpg.120043

280ah-battery_box-jpg.30611
 
The JKbms with 2A Active Balancing is fine for the 280's, my packs all use those BMS'.
The B cells are Voltage & IR Matched and still quite reasonable at that price point. They are fine if you build a straight 8S Battery Pack with it's own BMS.
I have A Cells, B Cells, Used EV Cells in my bank. I have done numerous tests including extreme Thrash tests because of my situation, reliability is critical, especially when it's -35 outside. The A Cells certainly do perform better, with slightly faster charge because the cells are IR matched so very little balancing is required, they do give me the 280AH. My B cells are more finicky when finishing charge but it is not significant enough to be a problem. The Used EV cells are actually pretty darn good after I tweaked them. In "MY CASE" having 6 packs in a bank irons all that out and I do get 100% of my capacity... Yes, I've drained the entire bank to Low Volt Disconnect in tests.

I built my own boxes, I cannot ever recommend having the cells/terminals etc in "open air" - that's just begging for trouble! Even a Covered shelf with a clear cover in front... be safe !

I use MRBF Fuses because they are perfectly fine for <36V. These are cheaper than Class-T (as good) and make for a much cleaner install (depending on the individual of course LOL) You can see below, that the (+) Busbar is offset to allow the MRBF fuse to clear over it's (-)

Blue Sea MRBF fuses made by Eaton/Bussman (for 200A BMS, use 250A fuse)

! WARNING ! There has been a LOT of Horse Pucks tossed about Binding/Compressing cells. It's been bad enough that people could have gotten to Physical Fistacuffs over it. The cells are to be BOUND snugly to prevent their expansion from twisting the terminal blocks. Do NOT put them under pressure !!! In fact, if you build the boxes below they fit the cells "snugly" and that's enough... Alternately Amy sells the Steel Cases for a more Pro finish, she also has Stackables that sit one ontop of the other... Not shown yet I think.

H4454fb0b4ae044059f1b227cda4f87afN.jpg_720x720q50.jpg


<--- Box Link:

I built all my own boxes out of 3/4" - 19mm Plywood with a clear polycarbonate cover. See attached below. I am using Pike Industries BusBars shown below. https://pikeind.com/industries-busbars/

24v-280ah-jpg.120043

280ah-battery_box-jpg.30611
Thanks, lots of good info.

Yes my packs will be protected on all sides, but not tightly wrapped in a case. The inter-cell busbars are of the flexible type, so I don't think the terminals will be affected at all from cell swelling.

How did you pack the cells together before inserting into the boxes, tape? How is the BMS mounted?

Re cell compression, I'll follow your guide Luyuan_Tech_Basic_LiFePO4_guide_1a.pdf on page 8, and try to make a snug package to keep parts from moving around.

Re fuses, a 3.2V cell with 0.25 mOhm IR could in theory release a 12.8kA current. But the 0.5 m AWG0 cable I will use to connect to the main busbars (1m in cable total) will up the resistance to maybe 0.60 mOhm. Then the current would be 5,3kA. Looks like the Blue Sea MRBF can handle 5kA @ 32 VDC and even more as voltage drops, so it should be fine.

I'm on a tight budget. If I had more money, I'd double the capacity, get BMS displays for monitoring and proper boxes, but I think my current setup will work. This is my first version. I suspect I'll be learning and experimenting as I go along.
 
The JKbms with 2A Active Balancing is fine for the 280's, my packs all use those BMS'.
The B cells are Voltage & IR Matched and still quite reasonable at that price point. They are fine if you build a straight 8S Battery Pack with it's own BMS.
I have A Cells, B Cells, Used EV Cells in my bank. I have done numerous tests including extreme Thrash tests because of my situation, reliability is critical, especially when it's -35 outside. The A Cells certainly do perform better, with slightly faster charge because the cells are IR matched so very little balancing is required, they do give me the 280AH. My B cells are more finicky when finishing charge but it is not significant enough to be a problem. The Used EV cells are actually pretty darn good after I tweaked them. In "MY CASE" having 6 packs in a bank irons all that out and I do get 100% of my capacity... Yes, I've drained the entire bank to Low Volt Disconnect in tests.

I built my own boxes, I cannot ever recommend having the cells/terminals etc in "open air" - that's just begging for trouble! Even a Covered shelf with a clear cover in front... be safe !

I use MRBF Fuses because they are perfectly fine for <36V. These are cheaper than Class-T (as good) and make for a much cleaner install (depending on the individual of course LOL) You can see below, that the (+) Busbar is offset to allow the MRBF fuse to clear over it's (-)

Blue Sea MRBF fuses made by Eaton/Bussman (for 200A BMS, use 250A fuse)

! WARNING ! There has been a LOT of Horse Pucks tossed about Binding/Compressing cells. It's been bad enough that people could have gotten to Physical Fistacuffs over it. The cells are to be BOUND snugly to prevent their expansion from twisting the terminal blocks. Do NOT put them under pressure !!! In fact, if you build the boxes below they fit the cells "snugly" and that's enough... Alternately Amy sells the Steel Cases for a more Pro finish, she also has Stackables that sit one ontop of the other... Not shown yet I think.

H4454fb0b4ae044059f1b227cda4f87afN.jpg_720x720q50.jpg


<--- Box Link:

I built all my own boxes out of 3/4" - 19mm Plywood with a clear polycarbonate cover. See attached below. I am using Pike Industries BusBars shown below. https://pikeind.com/industries-busbars/

24v-280ah-jpg.120043

280ah-battery_box-jpg.30611
I just watched this video (link below), and it got me a bit worried about grade B cells. In my discussion with Amy I got the impression that grade B only had slightly lower/higher capacity, voltage and/or internal resistance. But there seems to be something more fundamental wrong with some of the grade B cells. All it takes is to draw 100W from a cell and it starts showing weird behavior, which in my case is 1,6kW total. I will easily hit that threshold when I run power tools etc, and I don't want the BMS to cut out from detecting a low cell voltage. Do you know if Luyuan does a stress test of the cells before they are shipped?

 
I used Fibratape (Fiberglass reinforced tape) and taped the cells into 4 cell blocks, I also included a pull string, in case I had to pull them out. That was a Boo Boo, it made them a tad too snug and made that really hard to do.
30609-c0da85ea3c4dc174035846b0935c794a.jpg


Note that the BMS' come with 10cm 7AWG Ultrafine wire silicone wire. I found the leads to be too short for what I needed for my setup, so I extended them using a Crimped Butt Connector an 4AWG Silicone wire. 2x 7AWG = 4AWG.

Battery Cables are something that folks do trip over a little. I'll keep this simple, there are different grades of wire, from Coarse to Ultra Fine wire, there is Pure Copper, Tinned (for marine & $$) and then the "dubious". You can purchase any kind, premade to specific lengths but if that is the case then be cautious and stick with Top Quality Manufacturers & Vendors... STAY AWAY from Auto supply houses ! You can't imagine the crap they do flog ! Fine for an old Diesel Truck but NOT solar !

I use Ultra Fine Wire "Welding Cable" which has the most strands per square mm. When sizing your wire, you have to account for the worse case scenario. That means 1 Battery Pack pushing the absolute Max Amps that your Inverter could conceivably pull. For example, my Samlex EVO can deliver up to 12,000W pulling just under 500A to do so. Now my BMS' will prevent that BUT as we all know "shit happens" and only a fool does not take such into account... hence fuses, breakers & all that goodness to keep our butts safe. I myself use only Royal Excelene 4/0 (120 mm2) Ultra Fine Welding with from Southwire, see the link below for ampacity & thread count info etc... http://cabletechsupport.southwire.com/en/tile/10/spec/70305/?country=International I am sure you can get this in EU or very similar, it is not all that unique.
 
I just watched this video (link below), and it got me a bit worried about grade B cells. In my discussion with Amy I got the impression that grade B only had slightly lower/higher capacity, voltage and/or internal resistance. But there seems to be something more fundamental wrong with some of the grade B cells. All it takes is to draw 100W from a cell and it starts showing weird behavior, which in my case is 1,6kW total. I will easily hit that threshold when I run power tools etc, and I don't want the BMS to cut out from detecting a low cell voltage. Do you know if Luyuan does a stress test of the cells before they are shipped?
Heads Up, SFK has a Rancid Reputation around here... Most of us have NOTHING good to say about them. I'll leave that there.

Using 280AH cells as example;
A-Cells will always deliver the specified Amp Hours from the Working Voltage Range (3.000-3.400). These are Factory Validated & can be matched to other A Cells. When these cells are further tested by outsiders, from 2.500-3,650 they usually test out around 290AH +/- a bit.

B-Cells DO NOT deliver the specified AH from the working range, they are typically under by a few percent. That is Factory Tested and then classed as such. They will typically deliver between 270-280. When tested from 2.500-3.650 they may test up to 285AH - You can expect at least 5% to 10% (max) less capacity.

IF less the cells test lower than 10%, most manufacturers will recycle or at worse debadge and go to "grey market". For example CATL will Recycle the cells - they learned a long time ago...

The Factory Testing involves some stress tests... I do not know any vendor Stress Testing cells... that's a labour nightmare and more...

I have only had 1 Battery Pack cutout when under heavy load (Mig Welder actually) when it was down to 2.700Vpc and one cell dropped to 2.500 when my cutoff was 2.600. A Mig hits hard ! There's a spike to see. If your bank is that low, then you better charge before plugging tools in.

Something else which is obvious but not... When you have multiple packs in one bank, they do divide & share the loads/charge proportionately
and that really reduces the chances of an LVD. If your inverter demands 100A and that's split with 4 packs, they only have to deliver 25A ea, which barely would wake them up.

Note when I say proportionately.. If you have a bank of LFP batteries, they can be different AH without issue ! "Simple Description" Say you have a 100AH, 200AH & 300AH in a bank and you are pulling 60A, the 100 will deliver 10A, the 200 will deliver 20A and the 300 will deliver 30A. This is generally how it works and if observed during operation you will see this happening and the Amperage will float a small bit but stays pretty consistent. The same thing happens in reverse when Charging.

Hope that helps.
 
Heads Up, SFK has a Rancid Reputation around here... Most of us have NOTHING good to say about them. I'll leave that there.

Using 280AH cells as example;
A-Cells will always deliver the specified Amp Hours from the Working Voltage Range (3.000-3.400). These are Factory Validated & can be matched to other A Cells. When these cells are further tested by outsiders, from 2.500-3,650 they usually test out around 290AH +/- a bit.

B-Cells DO NOT deliver the specified AH from the working range, they are typically under by a few percent. That is Factory Tested and then classed as such. They will typically deliver between 270-280. When tested from 2.500-3.650 they may test up to 285AH - You can expect at least 5% to 10% (max) less capacity.

IF less the cells test lower than 10%, most manufacturers will recycle or at worse debadge and go to "grey market". For example CATL will Recycle the cells - they learned a long time ago...

The Factory Testing involves some stress tests... I do not know any vendor Stress Testing cells... that's a labour nightmare and more...

I have only had 1 Battery Pack cutout when under heavy load (Mig Welder actually) when it was down to 2.700Vpc and one cell dropped to 2.500 when my cutoff was 2.600. A Mig hits hard ! There's a spike to see. If your bank is that low, then you better charge before plugging tools in.

Something else which is obvious but not... When you have multiple packs in one bank, they do divide & share the loads/charge proportionately
and that really reduces the chances of an LVD. If your inverter demands 100A and that's split with 4 packs, they only have to deliver 25A ea, which barely would wake them up.

Note when I say proportionately.. If you have a bank of LFP batteries, they can be different AH without issue ! "Simple Description" Say you have a 100AH, 200AH & 300AH in a bank and you are pulling 60A, the 100 will deliver 10A, the 200 will deliver 20A and the 300 will deliver 30A. This is generally how it works and if observed during operation you will see this happening and the Amperage will float a small bit but stays pretty consistent. The same thing happens in reverse when Charging.

Hope that helps.
Good to know about SFK. I don't mind if a cell is 10% under in capacity or has a bit higher IR, I can live with that. But a cell that takes a nose dive in voltage at 100W (as shown in the video) would ruin the whole pack. I really hope that it is a rare abnormality, and that he is making it seem more common than it really is. It is inconvenient to have to wait another two months of shipping for a replacement of a faulty cell.

I agree that if you have many cells in your bank, then stuff like that doesn't matter as much. I'm more exposed, because right now 16 x 280Ah is all I can afford.

Interesting to see your cell assembly method and selection of cables, good info. No more questions at the moment. Thanks for the support. I'll keep you updated on my project.

Oh yes, one more question. A few competitors of Luyuan surfaced on Alibaba when I searched for "LF280K EU stock", for example QSO. What other sources besides Luyuan would you recommend? Always good to compare before you make a deal.

 

diy solar

diy solar
Back
Top