diy solar

diy solar

Calculate amps through each bus bar?

My BMS (REC Active BMS) has a 2 amp active balancer, and I will usually be charging and discharging at low C, so I'm not too concerned about keeping them balanced.

The other point to make here is if you are using Canbus connected equipment like a Victron Multiplus as the charge source that 2A active balancer is actually a lot more powerful then it seems.

This is because as you start to get close to your desired peak cell voltage the BMS will tell the charger to ramp down on its amperage output to 2A per cell - 8A total.

The BMS will then actively divert up to 2A from the highest cell to the lowest cell; so the lowest cell is getting 4A now and the top cell close to 0A. This cycle repeats as the cells converge on the charge set point.
 
Thanks for your report, @jwelter99 glad to hear the 3P is working well for you. Like you, I'm not too worried about paralleling 4 (in your case 3) cells. Lots of high-knowledge builders (e.g. Battleborn, electric car companies, etc) do parallel many more cells, and let the BMS monitor groups of cells, rather than individual cells.

Regarding when you said you "did add all 3 bus bars crossing between cells middle top" (which was also suggested by @grizzzman), I think I'm missing something. Please help me understand. OK, so I have 4 groups of 4 cells. Each group of 4 will be parallel to each other. Then I have to serial them up. Group 1 to group2, then group 2 to group 3, then group 3 to group 4. If I understand you correctly, both of you and @grizzzman are suggesting extra bus bars for the serial connection between group 2 and group 3 (what you refer to "middle top". But with this layout, there will only be a single connection between group 1 to group 2 (bottom left to top left) and group 3 to group 4 (top right to bottom right). To sum up, with a battery that has 3 serial connections, you suggest one of the serial connections have much higher capacity than the other two. Can you please explain why I'd want this?

In your original diagram with the upside down "U" at the top it will work. The current through that one cross-connect bus bar will be whatever the draw on the battery is. It keeps the topology that between every parallel group of cells you have a single connection to the next group. In your case these points carry 100A.

The other opinion is that you want the lowest resistance across the topology. So any open locations for bus bars should be populated as they lower the overall resistance losses across the pack.

This is why the linear topology is the best as it allows all the cells in parallel to also each have a bus bar to the next parallel cell. So the current through any of the bus bars will be reduced to the total draw / parallel cells. In your diagram case 25A.
 
The other point to make here is if you are using Canbus connected equipment like a Victron Multiplus as the charge source that 2A active balancer is actually a lot more powerful then it seems.

This is because as you start to get close to your desired peak cell voltage the BMS will tell the charger to ramp down on its amperage output to 2A per cell - 8A total.

The BMS will then actively divert up to 2A from the highest cell to the lowest cell; so the lowest cell is getting 4A now and the top cell close to 0A. This cycle repeats as the cells converge on the charge set point.
Great point, and yes, I have mostly Victron equipment, which is one of the reasons I went with an REC Active BMS, which speaks "Victron". One of the reasons I decided to go with a 4P4S setup, rather than 4S4P is I want just one BMS "to be in charge" and communicate the charge needs of the bank to all my Victron equipment. This, as you say, will allow my balancing to be more effective. Multiple batteries, each with a BMS requesting different things of the charge equipment sounds like a nightmare. That isn't the only reason I landed on 4P4S, but this isn't a thread about that, so I won't get into it here.
 
After taking in all the helpful comments in this thread, I've slightly changed my battery design. I think I'll be spring clamping the cells, and using wire and lugs rather than solid bus bars, or braided straps, and adding more series connections compared to my initial design. I've attached a diagram of what I'm intending to do my guesses about how much amperage will flow through each connection when pulling 100 amps. Some of my thinking was informed by my other thread about strain relief of terminals. https://diysolarforum.com/threads/cell-terminal-strain-relief-bus-bars-and-compression.27439/

Screen Shot 2021-08-27 at 11.28.15 PM.png
 
After taking in all the helpful comments in this thread, I've slightly changed my battery design. I think I'll be spring clamping the cells, and using wire and lugs rather than solid bus bars, or braided straps, and adding more series connections compared to my initial design. I've attached a diagram of what I'm intending to do my guesses about how much amperage will flow through each connection when pulling 100 amps. Some of my thinking was informed by my other thread about strain relief of terminals. https://diysolarforum.com/threads/cell-terminal-strain-relief-bus-bars-and-compression.27439/

View attachment 62007

This looks reasonable but before you commit to it what material are you using for the bus bars?

If it's the crappy bus bars that come from China then it is worth thinking about.

But if you are making your own out of something like 3/4 wide by 1/8 thick copper you will easily have 100A current carrying capacity and I'd not worry about introducing the cables as well.

As it's going in a boat you will find bus bars will last much longer then even really good cables/lugs simply because eventually moisture will get into the cable no matter how well sealed and quickly corrode the fine strands of wire - and it is hard to see this until it's too late.
 
After taking in all the helpful comments in this thread, I've slightly changed my battery design. I think I'll be spring clamping the cells, and using wire and lugs rather than solid bus bars, or braided straps, and adding more series connections compared to my initial design. I've attached a diagram of what I'm intending to do my guesses about how much amperage will flow through each connection when pulling 100 amps. Some of my thinking was informed by my other thread about strain relief of terminals. https://diysolarforum.com/threads/cell-terminal-strain-relief-bus-bars-and-compression.27439/

View attachment 62007

Maybe the + and - connections of the pack should similarly tap two points of their busbars.

You have each connection to positions "1" and "4" of the 4p. What if you alternated between "1" & "4" with "2" & "3"?

I think with those two changes it perfectly balances draw on each pack. There would also be zero voltage across busbars between "2" and "3", except for balancing.
 
This looks reasonable but before you commit to it what material are you using for the bus bars?

If it's the crappy bus bars that come from China then it is worth thinking about.

But if you are making your own out of something like 3/4 wide by 1/8 thick copper you will easily have 100A current carrying capacity and I'd not worry about introducing the cables as well.

As it's going in a boat you will find bus bars will last much longer then even really good cables/lugs simply because eventually moisture will get into the cable no matter how well sealed and quickly corrode the fine strands of wire - and it is hard to see this until it's too late.
Those china bars are same as 1 awg..

Mine are 20mmx2mm. That is 40mm2 which is 1 awg.

They are more than enough for 100 amp ?
 
Those china bars are same as 1 awg..

Mine are 20mmx2mm. That is 40mm2 which is 1 awg.

They are more than enough for 100 amp ?

Mine were much smaller, and the material was some sort of tin and not copper. 13mm by 1.2mm.
 
Maybe the + and - connections of the pack should similarly tap two points of their busbars.

You have each connection to positions "1" and "4" of the 4p. What if you alternated between "1" & "4" with "2" & "3"?

I think with those two changes it perfectly balances draw on each pack. There would also be zero voltage across busbars between "2" and "3", except for balancing.
Nice improvement, thanks. Here is the latest.
Screen Shot 2021-08-28 at 9.04.01 PM.png
 
Looks to be getting overly complex. In the first iteration the power flows through 15 bars and related connections no matter the path. Trace a few and count. This puts the balance as equal for all cells. If you want more capacity or lower resistance I would stick with the first and just beef up the bars.
 
Looks to be getting overly complex. In the first iteration the power flows through 15 bars and related connections no matter the path. Trace a few and count. This puts the balance as equal for all cells. If you want more capacity or lower resistance I would stick with the first and just beef up the bars.
I appreciate the simplicity argument. But there are a few advantages to this approach. Two requirements I have that others may not are I've decided to use wire instead of solid bus-bars, (see this thread if you want a discussion of why). Also, I'm designing this battery for a 400 amp load (converting my boat to 24v was prohibitively expensive, and a DC-DC converter is not powerful enough for many of my existing loads). My first diagram would have had each series interconnect carrying up to 400 amps, and some parallel interconnects carrying 300 amps. Plus, they would have to be tightly bent wires. My latest design allows much smaller wires, and the bends in the wire are less severe. No series wire sees more than 200 amps, and no parallel interconnect carries more than 100 amps. So that is the rationale behind that decision.
 
Fair enough. Consider four separate batteries with four separate BMS. Connect the 4x 4s batteries in parallel. Fuse each battery at 125 amps and connect to a bus to feed 400 amps to the loads. Can still be made in the compact configuration shown connected in 4 smaller series groups with the included bus bars to connect.

With 400 amps running you are looking at 2/0 or 4/0 wire in case there is a connection issue and power decides to mostly go through one of the parallel connections.

Just some random thoughts. Your call.
 
Fair enough. Consider four separate batteries with four separate BMS. Connect the 4x 4s batteries in parallel. Fuse each battery at 125 amps and connect to a bus to feed 400 amps to the loads. Can still be made in the compact configuration shown connected in 4 smaller series groups with the included bus bars to connect.

With 400 amps running you are looking at 2/0 or 4/0 wire in case there is a connection issue and power decides to mostly go through one of the parallel connections.

Just some random thoughts. Your call.
I appreciate the suggestions, even if I don't do them, because it makes me think through why I'm doing what I'm doing more carefully, and it often leads to changes in my design. I thought about 4P4S vs 4S4P pretty carefully, and there are a bunch of reasons 4P4S will work better for me, using a single REC BMS that can talk to all my Victron stuff. This probably isn't the forum thread to go into those reasons, so it stays on topic.
 
Just use buss bars, the extra wires are unnecessary, and make it more complicated and less reliable. Having multiple wires connecting between 2 bus bars could cause current loops. As a general rule, you never want to give electricity more than one path between 2 points. Only one path or a current loop can form.
To your original question, if the total amperage drawn is 100A, then every buss bar will be carrying 100A, with each of the 4p batteries contributing ~25A to that total. The current will not be even throughout the bar, but will total 100A at the point the bar spans from one 4p group to the next 4p group. This will not cause any issues.

FYI, I have a 3p4s pack in my boat in a very similar configuration to what you are planning. Its only been installed about 6 months, but I love it and has been completely worry free.
 
Just use buss bars, the extra wires are unnecessary, and make it more complicated and less reliable. Having multiple wires connecting between 2 bus bars could cause current loops. As a general rule, you never want to give electricity more than one path between 2 points. Only one path or a current loop can form.
To your original question, if the total amperage drawn is 100A, then every buss bar will be carrying 100A, with each of the 4p batteries contributing ~25A to that total. The current will not be even throughout the bar, but will total 100A at the point the bar spans from one 4p group to the next 4p group. This will not cause any issues.

FYI, I have a 3p4s pack in my boat in a very similar configuration to what you are planning. Its only been installed about 6 months, but I love it and has been completely worry free.
I appreciate your perspective, but I've decided that flexible connections are a better solution for me than solid bus-bars, as discussed thoroughly in this thread: https://diysolarforum.com/threads/cell-terminal-strain-relief-bus-bars-and-compression.27439

Very glad to hear your 3P4S pack is working well for you. Where are you located, and what type of boat?
 
Currently in Norfolk Virginia. I left San Francisco in 2018, and will complete my circumnavigation next year. Boat is a Morgan 382.

If going the flexible route, that's fine. But your drawing shows 2 wires connecting 2 bus bars together. Don't do that. Only have one electrical path to connect 2 bus bars. Also, do not use copper braid on a boat. There is lots of information about that in SSB radio discussions elsewhere on the Internet. Copper braid sucks up moisture, and salt, from the air. It will quickly turn green and fall apart. Tinned copper braid would be better, but still not a good idea. If you need that kind of flexibility, Insulated tinned wire, with lugs that had adhesive line heat shrink is best.

They make buss bars that have a bend in them so they have some give, and also slotted holes instead of round. If I were concerned about movement, that is the route I would take. I am not sure if you mentioned what cells you are using. I am using Calb cells, purchased through official distribution. They are not prone to expansion like the commodity cells from Alixxx. In a solid frame, my cells do not move.
 
Last edited:
Currently in Norfolk Virginia. I left San Francisco in 2018, and will complete my circumnavigation next year. Boat is a Morgan 382.
Sounds great! We went Hawaii—>Seattle—>Mexico last year and this, i
and plan to continue on this fall, aiming to go “the wrong way” to Europe by 2023.

If going the flexible route, that's fine. But your drawing shows 2 wires connecting 2 bus bars together. Don't do that. Only have one electrical path to connect 2 bus bars.
Can you elaborate? And link to resources/discussions? I’ve seen so many people’s packs with multiple connections. Multiple links allows me to build a perfectly balanced pack, and reduce wire size, which is important because I’m designing for a 400 amp constant load and using tinned wire as interconnects)

Also, do not use copper braid on a boat. There is lots of information about that in SSB radio discussions elsewhere on the Internet. Copper braid sucks up moisture, and salt, from the air. It will quickly turn green and fall apart. Tinned copper braid would be better, but still not a good idea. If you need that kind of flexibility, Insulated tinned wire, with lugs that had adhesive line heat shrink is best.
Yup, decided not to go with braid for those reasons.

They make buss bars that have a bend in them so they have some give, and also slotted holes instead of round. If I were concerned about movement, that is the route I would take.
If I were to use bus bars, and not have multiple links, as you recommend, some bus bars would be carrying 400 amps. Seems like something to avoid.

I am not sure if you mentioned what cells you are using. I am using Calb cells, purchased through official distribution. They are not prone to expansion like the commodity cells from Alixxx. In a solid frame, my cells do not move.
I’m using EVE 280ah grade A lfrom Amy, who has an excellent reputation. And I think they require some extra coddling as far as fixturing since they are not as physically robust as Calb/Winston/etc. That’s why I am spending so much time effort on low strain interconnects, and a very supportive battery box with spring compression.
 
A possible downside to 4p4s that's been discussed on the forum is that if one cell does decide to short itself out, the other three dump into it. Of course just one shorting out makes itself hot and overheats others. That alone isn't supposed to start a fire or cause failure of other LiFePO4 cells, but if intimate thermal contact it would overheat others. Other lithium chemistries "vent with fire." LiFePO4 vents below self-ignition temperature of the electrolyte. But, overheated busbars, sparks, etc. could ignite the pack. At least one forum member had a LiFePO4 pack ignite.

Packs made with flashlight battery sized cells, like those in laptops, have spot welded strips connecting many in parallel and the sheet metal strips double as fuses, so one shorted cell is isolated.
Perhaps ideal for 4p4s would be suitable fuses for all parallel connections. That would be 12 fuses. Cheap fuses that is possible, but fuses good for 3.4V 20kA interrupting would get expensive.

For these reasons, it would seem 4s4p (with one fuse per 4s) has fewer failure modes. It does require 4 BMS.

Loops - ground loops are a thing, and in our lab we're measuring high microamps to milliamps of 60 Hz current in chassis structure. That wouldn't bother most machinery, but for sensitive instruments it does. Source is a combination of loops picking up magnetic fields in the room, and leakage current from power lines to ground, such as due to EMI filters having capacitors to ground. In lab, back when fluorescent lamps were common, that was a 4' long single conductor which made a magnetic antenna.

Small loops of wire are a mesh, not such a problem. I don't think 4" or 1' spans are any issue for these low frequencies (which aren't RF).
What can be a problem is if one connection becomes open so current crowding into remaining connection is excessive and burns it out.
If a single connection becomes high resistance, voltage drop across it makes it hot. At some point, voltage drop is enough that BMS disconnects load due to imbalance between cell voltages.

I wouldn't worry about loops (redundant paths) in battery packs for noise reasons. Only reasons involving excess current or cell failures.
 
Back
Top