diy solar

diy solar

24v Setup

Texasbaird

New Member
Joined
Jan 8, 2020
Messages
32
Will, first of all, I appreciate your videos. They have been a real blessing to a novice like myself and provided me the courage to take on a solar project. I will elaborate more when I post my design for more experienced people to comment on. Here is where i am today.

To those more experienced..
I recently order 16 fortune 100Ah LiFePO4 3.2v batteries which will arrive today. I want to configure them as a 24v battery at ~200Ah. You recommend that I should first put them in parallel and once balanced, then connect them in series.

Is the process and layout correct in the attached pdf?
 

Attachments

  • 24v Battery setup.pdf
    48 KB · Views: 81
The drawing shows 4S for a 12v system you want 8S for 24v. If you cells are now and the voltage is all the same I would not worrie about balancing them. put into your config for use and charge them as you would be using them. for 8s stop at 28v
 
You show a 400Ah 12 Volt 4P4S cell configuration
1580852648914.png

You want a 200Ah 24V 2P8S configuration.

1580853681455.png
 
1580855381425.png

1580855440308.png
Series-first vs parallel-first. There are near religious arguments over which is better.

I prefer parallel first because of the single BMS simplicity. However, that means a much higher current BMS (Expensive) so there are times when series-first makes sense.
 
Last edited:
One way to use a low current BMS in a high Current set up is to use Wills trick of inserting a DC contactor in the negative path
1580856131034.png
 
Last edited:
Will, first of all, I appreciate your videos. They have been a real blessing to a novice like myself and provided me the courage to take on a solar project. I will elaborate more when I post my design for more experienced people to comment on. Here is where i am today.

To those more experienced..
I recently order 16 fortune 100Ah LiFePO4 3.2v batteries which will arrive today. I want to configure them as a 24v battery at ~200Ah. You recommend that I should first put them in parallel and once balanced, then connect them in series.

Is the process and layout correct in the attached pdf?

As far as process goes, there is good enough and there is better. However, both of those are trumped by what you have the equipment to do.
(and there are near religious arguments on this too)

I think your process was just about connecting (not balancing).

Connection Process:
The connecting part is rather straight forward:
  1. 1) Check the voltages on all the cells.
  2. If they are all within .1 V or less, connect pairs in parallel and then connect the 8 pairs in series. - Done
  3. If you have outliers,
    • connect the ones that are close in parallel and then connect the outlier through a 3-5 ohm resister and let it set overnight.
    • once they have all settled to the same voltage, connect pairs in parallel and then connect the 8 pairs in series.
Note: If you have outliers, you may want to balance them first.​
Balancing process:
For what you are doing, Top balancing is probably best. (Note: if you have new, matched cells, there is an argument that you don't need to balance at all). The best way to top balance is to hook them ALL in parallel and then charging them to 3.65v and waiting for current to drop. Once the current drops to something very low, disconnect them from the supply, take it all apart and then build the 2P8S system.

Corrected below
Notes:
Technically, you need a power supply that can limit the current and voltage. In the case of 16 parallel 100Ah cells, it would take 320 amps to just get to a .2C charge rate so current limit is not likely to be a problem, but the supply has to manage the current to something the supply can handle till the voltage hits the 3.65V level.

Make sure to disconnect the power supply once the voltage is at 3.65V and current gets to near zero.... do not let them sit at 3.65 for an extended time.

Now, before people start jumping all over me about how wrong I am about balancing: Let's keep that 'discussion' over on this thread
 
Last edited:
As far as process goes, there is good enough and there is better. However, both of those are trumped by what you have the equipment to do.
(and there are near religious arguments on this too)

I think your process was just about connecting (not balancing).

Connection Process:
The connecting part is rather straight forward:
  1. 1) Check the voltages on all the cells.
  2. If they are all within .1 V or less, connect pairs in parallel and then connect the 8 pairs in series. - Done
  3. If you have outliers,
    • connect the ones that are close in parallel and then connect the outlier through a 3-5 ohm resister and let it set overnight.
    • once they have all settled to the same voltage, connect pairs in parallel and then connect the 8 pairs in series.
Note: If you have outliers, you may want to balance them first.​
Balancing process:
For what you are doing, Top balancing is probably best. (Note: if you have new, matched cells, there is an argument that you don't need to balance at all). The best way to top balance is to hook them ALL in parallel and then charging them to 3.65v and waiting for current to drop. Once the current drops to something very low, disconnect them from the supply, take it all apart and then build the 2P8S system.

Corrected below
Notes:
Technically, you need a power supply that can limit the current and voltage. In the case of 16 parallel 100Ah cells, it would take 320 amps to just get to a .2C charge rate so current limit is not likely to be a problem, but the supply has to manage the current to something the supply can handle till the voltage hits the 3.65V level.

Make sure to disconnect the power supply once the voltage is at 3.65V and current gets to near zero.... do not let them sit at 3.65 for an extended time.

Now, before people start jumping all over me about how wrong I am about balancing: Let's keep that 'discussion' over on this thread
 
@FilterGuy thanks again for the in-depth response. I have put everything in parallel and they are all testing at 3.292v. I am going to put them in series and add the BMS. Does this BMS config look correct?
1580930127881.png
 
What BMS are you using. having both B0 and B- on the negative seems unusual but without looking at the BMS spec I can't tell.
 
Oh..I think it just clicked with me. In your drawing, is the B- the big wire going from the battery to the BMS? If so then your drawing looks correct. But please let me know the BMS.
 
What BMS are you using. having both B0 and B- on the negative seems unusual but without looking at the BMS spec I can't tell.
Here's the wiring diagram that came with it and the BMS itself.
 

Attachments

  • BMS.jpg
    BMS.jpg
    144 KB · Views: 30
  • BMS wiring diagram.jpg
    BMS wiring diagram.jpg
    75.2 KB · Views: 30
With cells in parallel you cannot balance or manage each cell because they are paired, the pair will always show one voltage so imbalances can & will occur. Also if one cells poops out the pack becomes useless till fixed. It is the worst mix one can do but it works. I know I have one like that, never again and I will likely split it when I get my other two packs up & running. BTW, this is the method some companies use to "flog" cells by putting a good cell next to a weak cell and paralleling them together and making packs to go with BMS included for cheap.... Often used with B-Grade or used cells.

The ideal solution is two 8S packs with their own BMS and then paralleled. If one pack poops out the other keeps going till you get sorted out. An extra $100 is minimal insurance or fault tolerance. BUT I would also recommend a smart BMS like Will reviewed with Bluetooth so you can actually tweak & monitor your cells. Those BMS' have no internal balancing function though but that's usually OK if you have independent cell management.

I personally use Chargery BMS which has soft / passive balancing and quite configurable & programable but it is a more complex system with external relays and such. For my application and the way I set my fault tolerance requirements (offgrid & remote) it's the best solution for me.

BTW, this is the ShunBin pack whih is series-parallel and a fine example of why not.... (links in my signature)
lfp-pack-layout-jpg.6401
 
With cells in parallel you cannot balance or manage each cell because they are paired, the pair will always show one voltage so imbalances can & will occur. Also if one cells poops out the pack becomes useless till fixed. It is the worst mix one can do but it works. I know I have one like that, never again and I will likely split it when I get my other two packs up & running. BTW, this is the method some companies use to "flog" cells by putting a good cell next to a weak cell and paralleling them together and making packs to go with BMS included for cheap.... Often used with B-Grade or used cells.

The ideal solution is two 8S packs with their own BMS and then paralleled. If one pack poops out the other keeps going till you get sorted out. An extra $100 is minimal insurance or fault tolerance. BUT I would also recommend a smart BMS like Will reviewed with Bluetooth so you can actually tweak & monitor your cells. Those BMS' have no internal balancing function though but that's usually OK if you have independent cell management.

I personally use Chargery BMS which has soft / passive balancing and quite configurable & programable but it is a more complex system with external relays and such. For my application and the way I set my fault tolerance requirements (offgrid & remote) it's the best solution for me.

BTW, this is the ShunBin pack whih is series-parallel and a fine example of why not.... (links in my signature)
lfp-pack-layout-jpg.6401
Like I said in an earlier post "Series-first vs parallel-first. There are near religious arguments over which is better. "

Neither is wrong. It all comes down to what you are comfortable with and what the specific circumstance dictates.

I always start with new, matched cells and I don't have a big concern about them aging differently, so I lean toward parallel-first. (A lot of my installs are for other people that won't be checking cells so the ability to check them would be wasted anyway)

If you have poorly matched cells, or you don't believe matched cells age similarly, perhaps series-first is better.

BTW: My next install is series-first because the owner bought two lower amp BMSs rather than one higher amp BMS.
 
The ideal solution is two 8S packs with their own BMS and then paralleled. If one pack poops out the other keeps going till you get sorted out. An extra $100 is minimal insurance or fault tolerance. BUT I would also recommend a smart BMS like Will reviewed with Bluetooth so you can actually tweak & monitor your cells. Those BMS' have no internal balancing function though but that's usually OK if you have independent cell management.

I personally use Chargery BMS which has soft / passive balancing and quite configurable & programable but it is a more complex system with external relays and such. For my application and the way I set my fault tolerance requirements (offgrid & remote) it's the best solution for me.

BTW, this is the ShunBin pack whih is series-parallel and a fine example of why not.... (links in my signature)

My concern with series first is the added complexity of multiple management systems. Maybe (probably) this is a concern born of ignorance, but with two or more BMS' it seems much more complicated (or maybe not possible) to manage and communicate with external components (Inverter, SCC, battery protect, relays, etc), and makes centralized monitoring more complex (if you use a BMS where monitoring is a feature, like the chargery).

I haven't thought a lot of this through yet, these are just my initial concerns. Am I making much ado about nothing, are these non-issues or solveable problems? or are these downsides/tradeoffs realistic?
 
Back
Top