diy solar

diy solar

4 batteries: two of them drain to charge the others???

I'd be tempted to reconfigure the 4 banks into 2 of 6P4S. Would add stability to the runners.
And still provide redundancy. (And a bunch of spare parts).

Hard to tell from the pic the difficulty, though.

But maybe a better approach would be to just replace the cells with new ones and sell the old ones off as grade B.
Worth a cost appraisal, anyway, vs electrician time + opportunity cost.
Though if things are kind of working now, probably just make this your next upgrade when the time comes.

Good luck!
 
If I wasn’t selling the boat, I would replace all the cells and start from scratch using 6 BMS units and make 6 batteries of 2P 4S.

And I would spend a little more time trying to figure out how the batteries are charging with 5A when the engines are on?? Where they are connected? And why two batteries drain into the others…

But since the system is working pretty good, I’m confident it will surpass other people’s standards. I’m just a bit nitpicky finding every single small issue.
 
If I wasn’t selling the boat, I would replace all the cells and start from scratch using 6 BMS units and make 6 batteries of 2P 4S.

And I would spend a little more time trying to figure out how the batteries are charging with 5A when the engines are on?? Where they are connected? And why two batteries drain into the others…

But since the system is working pretty good, I’m confident it will surpass other people’s standards. I’m just a bit nitpicky finding every single small issue.
Cool. That's interesting that you are more towards multiple packs vs. fewer bigger packs. I know there was some chatter toward that decision, but I never saw anything compelling except for in the case of older, damaged cells that need more troubleshooting.

Do you think it's plenty viable now to make a 4p4s out of new cells? Given that a backup bank is also available.
And particularly for an oversized bank that will mostly stay in the 20-80% zone.

Interested in your thoughts, thanks!
 
Cool. That's interesting that you are more towards multiple packs vs. fewer bigger packs. I know there was some chatter toward that decision, but I never saw anything compelling except for in the case of older, damaged cells that need more troubleshooting.
From the advice I’ve gotten from multiple people here is that it’s always better in the long run to go with 2P packs and try to avoid anything bigger. They have all sorts of reasons and proved it regarding how different cells draw at different lengths, and how it can be dangerous with runners. (My original idea was to make a giant 12P pack and everyone convinced how dangerous that was.)


Do you think it's plenty viable now to make a 4p4s out of new cells? Given that a backup bank is also available.
Brand new A grade cells shouldn’t be a problem for years. Keep in mind, at some point you do need to take them apart and test them then rematch them. What’s the likely hood of you doing that? For me, I can say that no way will I ever go through the month-long process of capacity testing 48 cells ever again. Or even breaking down my system so is rather just build it in the most stable way. Then 7-8 years down the line just buy new cells which will be cheaper and better than anything we have today.
And particularly for an oversized bank that will mostly stay in the 20-80% zone.
Why 20-80%? I mean you could get it to last 10 years or longer like that… but by that point, I’d want to upgrade. So I’m not too concerned about longevity.
Interested in your thoughts, thanks!
 
From the advice I’ve gotten from multiple people here is that it’s always better in the long run to go with 2P packs and try to avoid anything bigger. They have all sorts of reasons and proved it regarding how different cells draw at different lengths, and how it can be dangerous with runners. (My original idea was to make a giant 12P pack and everyone convinced how dangerous that was.)

Brand new A grade cells shouldn’t be a problem for years. Keep in mind, at some point you do need to take them apart and test them then rematch them. What’s the likely hood of you doing that? For me, I can say that no way will I ever go through the month-long process of capacity testing 48 cells ever again. Or even breaking down my system so is rather just build it in the most stable way. Then 7-8 years down the line just buy new cells which will be cheaper and better than anything we have today.

Why 20-80%? I mean you could get it to last 10 years or longer like that… but by that point, I’d want to upgrade. So I’m not too concerned about longevity.
Good info thanks. I'm planning on a 4p4s, so I'll keep that in mind and be ready to split it into two batteries if it gets troublesome.
 
Good info thanks. I'm planning on a 4p4s, so I'll keep that in mind and be ready to split it into two batteries if it gets troublesome.
One thing I’ve learned, especially working with damaged B or C grade batteries is that you want lots of them because they will fail.

If you have just 2 batteries, when one goes out, that’s 50% of your system. In my case I have 4 batteries and if one drops, I still have 75% power left.

Also keep in mind this is a safety thing too. Because the amperage of whatever you are using then gets divided by whatever batteries remain. Say you are using a 3,000w inverter. At 12v that means your system is drawing 250A theoretically. (In practice it would be up to 20% more)

In my 4 battery system, each battery is only providing just 75A of power. Easy peasy. But if 2 of them drop out, suddenly that goes up to a whopping 125A per battery or more!! Most BMS can’t even handle that kind of amps even if they say they are rated for it. Unless you have 4/0 gauge cables, things will get hot fast. So having more batteries helps if one or even two drops out.

Example: a 6-battery system of my 48 cells would be 2P 4S. (It’s what I should have done if I had the money.)
Each of the 6 batteries are handling just 42A each. If one drops, it only goes up to 50A, still a low number.
 
Back
Top