2P4S or 4S2P and why?

[Keith_PDX]

I am going to convert my van to lifepo4. I have 8 150ah cells that I am going to build a 12v pack from. I'm staying 12v as I already have 400w of solar and 12v Magnum inverter. I've done quite a bit of research and some opinions vary quite a bit. With that said, what is the consensus with this group?

 

[Strobi-1]

I was getting ready to post the exact same question! Sorry to hijack... 4S2P (creating cell pairs, then wiring the cells in series) feels right, but I purchased my cell used I won't be able to monitor individual voltage to make sure I'm not over charging a single cell. Because of that I considered wiring them as 2 individual 4S1P and then running those two sets in parallel. I think this would give me better over sight, but doesn't seem as tidy (are there any other concerns doing this?). My 3rd thought was to do the latter initially and when I see that everything is good to go, switch to a 4S2P setup. Again, sorry to hijack ? but I leave for a trip in 2 days, so today is the decision day. Any thoughts would be greatly appreciated! Thanks, Andreas

 

[Sojourner1]

Not sure of your question other than you will have a 12v 300ah battery with 8 150ah cells which would be 4s2p.
(4s) 4 x 3.2v=12.8v
(2p) 2 x150ah=300ah

My bank is 4s5p (12v 500ah) 20 cells.

 

[Strobi-1]

Anyone see an issue with either setup of attached photo?

 

[Sojourner1]

I was getting ready to post the exact same question! Sorry to hijack... 4S2P (creating cell pairs, then wiring the cells in series) feels right, but I purchased my cell used I won't be able to monitor individual voltage to make sure I'm not over charging a single cell. Because of that I considered wiring them as 2 individual 4S1P and then running those two sets in parallel. I think this would give me better over sight, but doesn't seem as tidy (are there any other concerns doing this?). My 3rd thought was to do the latter initially and when I see that everything is good to go, switch to a 4S2P setup. Again, sorry to hijack ? but I leave for a trip in 2 days, so today is the decision day. Any thoughts would be greatly appreciated! Thanks, Andreas

Anyone see an issue with either setup of attached photo?

It would be the bottom of those 2 pics, the top one I'm not sure what is going on. Put 4 in series to make your 12.8v then parallel them.

 

[grizzzman]

You would need two BMS,s on the second pic and one on the first. (The second is two 12 volt batteries connected together)

 

[Strobi-1]

You would need two BMS,s on the second pic and one on the first. (The second is two 12 volt batteries connected together)

Yes Grizzman, I'd agree with you, but a bunch of guys like Will keep saying that on a solar system although preferable, a bms is typically not required. You just need to do some additional maintenance every few months. This is my first exposure to a system like this. I've built 18650 packs in the past that absolutely required a bms. So I was going to go with the theory of no bms required. Any personal experience of yours that would go against that theory? Thanks!

 

[Sojourner1]

On a perfectly match set of cells it's probably not needed. If you're not using a bms are you still going to be able to read the individual cell voltages? If not how are you going to know if the pack cells start to go astray other than keep checking with a meter.

Your solar controller/ charger (appropriate settings) is the first line of protection, the bms is the last line of protection.

 

[Keith_PDX]

From what I have read, parralel before series so the parallel cells can equalize. Provides some safety and the ability to easily measure voltage. I am leaning towards this..
2P4S

 

[CheezWiz]

You know, I always thought it was S that always came first, in other words 4S2P was the same as 2P4S. However based on this conversation and what I have found elsewhere I am enlightened. Here is how I am considering doing my pack:

(d) will get you the self balancing advantages of (a) while allowing you to protect the conjoined twin from catastrophic failure of their other half. You can also manage this pack with a single 4S BMS.

 

[Strobi-1]

On a perfectly match set of cells it's probably not needed. If you're not using a bms are you still going to be able to read the individual cell voltages? If not how are you going to know if the pack cells start to go astray other than keep checking with a meter.

Your solar controller/ charger (appropriate settings) is the first line of protection, the bms is the last line of protection.

If you make two individual 4s1p packs, just leave balance wires attached to each and periodically check them by plugging in one of those little multi cell battery checkers. Just to be clear, I really don't know that much and am going by what I think is a good idea because other people have stated it would work. So please let me know if my thinking is flawed. ? Thanks

 

[Keith_PDX]

As you indicated, most people tend to think 4P2S and 2P4S are the same but they're quite different electrically.

I personally think it comes down to the single fact that in a 2P4S build your cells will stay better balanced under charge and discharge in relation to each other. The 2 in parallel will always stay at the same voltage and the BMS or balancer will keep the 4 pairs in series within balance with a standard 4S balance scheme.

With 4S2P what you basically have is two individual battery packs that only paralleling the output voltage of each pack. This is very similar to what people do with two 12 volt lead acid batteries in parallel. The difference with LifePO4 is that one pack may have different individual cell voltages in relationship to the other pack and unless you have networked bms or networked active cell balancers your probability of getting full capacity in and out if each 4s pack is lowered. Simply having 2 BMS or 2 active cell balancers (1 on each series string) that due not communicate to each other will never yield the ability to charge and discharge the whole bank uniformly in relation to time. The system becomes significantly more complicated than it needs to be.

Each design has its own positives and negatives in regards to fault tolerance and redundancy while maintaining overall system voltage that warrant I guess a completely different design discussion.

 

[Keith_PDX]

You know, I always thought it was S that always came first, in other words 4S2P was the same as 2P4S. However based on this conversation and what I have found elsewhere I am enlightened. Here is how I am considering doing my pack:
View attachment 883
(d) will get you the self balancing advantages of (a) while allowing you to protect the conjoined twin from catastrophic failure of their other half. You can also manage this pack with a single 4S BMS.

Do you have an implementation of (d) to show? I am having a hard time wrapping my head around how you do the green lines with a single 4s bms. Link or otherwise is fine. Thanks!

 

[CheezWiz]

Do you have an implementation of (d) to show? I am having a hard time wrapping my head around how you do the green lines with a single 4s bms. Link or otherwise is fine. Thanks!

I do not have it built yet, but basically the green wires will keep the parallel batteries balanced with each other. So connecting a 4S bms to either cell in the pair will yield the measurement from the pair. The green wires are very small because the currents necessary to keep the cell pair balanced are small and it will melt through should a cell decide to short. At that point, the BMS should shut the whole thing down as the voltages would be all weird. I need to test of course. I got the idea for that configuration from the book that Jehu links at the end of his videos.

 

[Craig]

If yo do 2p4s you will have 4 batteries to get out of balance If you do 4s 2p you will have 8 batteries to get out of balance. when you paralel the batteries first they become one. when you parallel the batteries after the series both banks will have same total voltage after equalization but each battery in the series can be different. they will still add up to 12 volts but each cell may be different say something like

2.95+2.99+3.03+3.03= 12

2.9+3.3+3.0+2.8 =12

Both sets equal 12 but batteries are all over the board.
It probably would not go this far out of whack I was just trying to make an example.
each bank would need its own BMS

 

[CheezWiz]

As to the OPs question as to which is better. There is some science out there that answers that.
In this book, starting on page 170, they explain it. I am able to read most of the section when logged into my Google account and basically it seems the science shows that P-S configurations tend to be more reliable and reduce costs associated with BMSs,

Battery Management Algorithm for Electric Vehicles

This book systematically introduces readers to the core algorithms of battery management system (BMS) for electric vehicles. These algorithms cover most of the technical bottlenecks encountered in BMS applications, including battery system modeling, state of charge (SOC) and state of health...

books.google.com

You will also notice that is where I got that handy diagram above!

 

[SolarRat]

The first thing I'd do is 8p overnight to balance them. Then whatever is easiest and cleanest for the 12v pack with no bms. I'd check the balance a few months later and see if it needed a BMS then.

I kinda like the 2p4s because if one battery was a tiny bit low or high it has another battery to "help" it. That same battery in a 4s would drag all 4 down. That's just my theory but in real life it probably doesn't matter one bit.

 

[Sojourner1]

For new folks reading it can be confusing when talking about cells with in a battery and calling them batteries. It might make it easier and not so confusing to call them cells.
12v = 4 x 3.2v "cells" make a "battery".

 

[grizzzman]

Yes Grizzman, I'd agree with you, but a bunch of guys like Will keep saying that on a solar system although preferable, a bms is typically not required.
You just need to do some additional maintenance every few months. This is my first exposure to a system like this. I've built 18650 packs in the past that absolutely required a bms. So I was going to go with the theory of no bms required. Any personal experience of yours that would go against that theory? Thanks!

I really like Will due to his enthusiasm but.....His advice is at times, not the best. Especially with the Dippen, your feet in it crowd. Save your future grief.
Buy a BMS. Il give you an example. The second trip out with my pac I had a grounding issue causing the solar controller to spike to 17 volts. THE ONLY reason I found it is I had a voltage logger hooked up to my system. Found it when I got home and downloaded the file. With a BMS as insurance, I did not need to worry about it clearly damaging the pac.

 

[CheezWiz]

Considering the investment I am making in my cells and how little a BMS can cost relative to those cells (see Wills latest video), why not use a BMS? Low cell voltage can absolutely damage these cells.