Parallel and series connections, what comes first?

[chaka]

Most of us when first staring out buy lipo packs prewired in series and simply add packs in parallel to gain more capacity. The trouble is when our packs get close to 80% depleted some of the cells lose their capacity to deliver the amps needed and you get individual cells taking a voltage dive which triggers the low voltage cutoff, if you have one installed. My thinking is if the cells are wired in parallel first before being connected in series the cells will be better able to handle amp loads when close to 80% depleted. My reasoning for this is that instead of asking 1 cell to deliver a given amperage at a given voltage we are asking a pack of cells to do the same task. To me this seems inherently more stable but time will tell once I re-wire my system into a 2p8s instead of 8s2p. This is probably old news since most larger systems I see are wired similarly in EV's. Even if my thinking is flawed and it offers no extra stability it greatly simplifies the balance wiring harness which is nice if you are stuffing 120 18650 cells into a skateboard.

 

[dnmun]

your thinking does not account for the laws of physics. the current flows serially and there is no current flowing in parallel. use large conductors for the serial connection and you can use small conductors for the parallel connections since they carry only the balancing current when the BMS is balancing the pack.

 

[chaka]

your thinking does not account for the laws of physics. the current flows serially and there is no current flowing in parallel. use large conductors for the serial connection and you can use small conductors for the parallel connections since they carry only the balancing current when the BMS is balancing the pack.

Your comment only seems to back my reasoning, and please forgive my ignorance of laws. To me it would seem that a pack that is wired in series before the parallel connection is still asking individual cells to deliver the needed amps under load (Yes I understand the amperage is split among the several serial packs wired in parallel). In the end I suppose my thinking only offers lower resistance, due to the lack of balance leads between the parallel cells. I have noticed my packs are much more stable when I leave my balance leads connected in parallel under load. In the end wiring in parallel first simplifies the balance leads and construction which is good enough reason for me to adopt this configuration. To be clear, the Tesla Roadster cells are wired in parallel prior to the series connection. I wanted to add, current does flow in parallel when one cell starts to lose its capability to meet the demand under load, as a pack reaches 80% discharge the C rating will drop starting with the weakest cell.

 

[Voltron]

Classic avatar pic btw Chaka. And I think you're right..current will flow thru the parallel links as a low cell tries to equalize with the others around it. From the diagrams I've seen, parallel then series is better for continuing to supply current as a low cell is giving out. Series first makes it easier to identify that dying cell, as it's not hidden in the middle of a parallel group.
For visualizing, just to bounce some numbers, with the 120 cells, if you went 20s, then grouped them 6p, and one cell goes bad, one whole 20s string goes down, leaving you with 1/6 less capacity. But it's going to be easy to figure out which string, and which cell went bad. Going 6p first, then 20s, one cell going bad is supported by the others, leaving one group of the 20 at 5/6 capacity. Better for still delivering power, but harder to figure out which of the 6 in the group it is without isolating them from each other.

 

[chaka]

Classic avatar pic btw Chaka. And I think you're right..current will flow thru the parallel links as a low cell tries to equalize with the others around it. From the diagrams I've seen, parallel then series is better for continuing to supply current as a low cell is giving out. Series first makes it easier to identify that dying cell, as it's not hidden in the middle of a parallel group.
For visualizing, just to bounce some numbers, with the 120 cells, if you went 20s, then grouped them 6p, and one cell goes bad, one whole 20s string goes down, leaving you with 1/6 less capacity. But it's going to be easy to figure out which string, and which cell went bad. Going 6p first, then 20s, one cell going bad is supported by the others, leaving one group of the 20 at 5/6 capacity. Better for still delivering power, but harder to figure out which of the 6 in the group it is without isolating them from each other.

Haha, thanks Voltron! I come from a downhill longboarding background so I tend to ride aggressively and if I am not careful I will start to encounter voltage drops in individual cells at 60% discharge. This stopped happening when I left my balance wires connected in parallel while under load.

From what I understand, a cell's voltage will drop when it cannot meet the discharge demand. When a cell has a partner to share the capacity they recover better after burst currents. On the Tesla Batteries all the cells are connected to a busbar via fuse wires. I suppose the fuse wires are there to eliminate a bad cell before it can destroy the whole pack. In a 20p system with 2500mAh cells the pack would go from 50,000mAh to 47,500mAh which would lower the overall discharge capacity of the entire system signalling that something is wrong.

 

[dnmun]

and how large are the sense wires that connect it in parallel? what did you show that i did not explain originally?

 

[chaka]

and how large are the sense wires that connect it in parallel? what did you show that i did not explain originally?

They are 20amp fuse wires which, according to you, are useless. It really baffles my mind wondering why Tesla would waste the time and expense to build them in such a way, laws of physics being what they are.

To add: the fuse wires consistently blow at 25amps. What I am wondering is whether they are designed to blow to eliminate a bad cell or for over-current. I am thinking it is for a failing battery since there is a 100amp fuse on each pack of 22p15s Tesla packs. It would be interesting to see what kind of current the rest of the pack would deliver to a dyeing cell and guard against that.