DIY Electric Car Forums banner
1 - 11 of 11 Posts

·
Registered
Joined
·
15 Posts
Discussion Starter · #1 ·
Hello everyone, I have stated before that I have built many simple EVs in the early days. Mostly AGM batteries and early LIPO packs with a ceiling of 156VDC, those are always pretty simple being AGM and low count series LIPO. I am doing a personal vehicle project and want a larger than normal module but, I do not want to fall into the trap of pretending to know more than I actually do and doing something expensive, self-defeating, and dangerous. I have been researching but cannot pinpoint where the balance point is in a series/parallel aspect. So, I am hoping someone out there can give me constructive input. I am looking at constructing a battery module rated at 400 volts utilizing Headway 16Ah 40160cells. I would love to stack (3) 400VDC+/- in parallel, but that would require 125 cells in series and I am thinking that is not a good number. Are 125 cells in series a feasible number? I mean that's a lot of cells in series. A better way to do this would be? and how about this many cells in series for a BMS? And any ideas on a good BMS through your experience?
 

·
Registered
Joined
·
290 Posts
I am using the thunderstruck BMS, and it is good for up to 384 cells. Basically each group of 96 cells needs one BMS controller (BMSC), and 3 BMS satellites (BMSS). The cells are wired in consecutive groups of 12, but it does not matter how they are configured - the software lets you map them out however you want. 125 is not optimal, and I think that the chips these BMSes uses is pretty universally based on using 12 cells. If you were willing to live with 120 cells, you could get away with 2x BMSCs and 3x BMSSs. That would be a $1725 BMS, adding the 5 extra cells would bring it to an even $2000.

The Orion BMS handles up to 180 cells on a single unit, so that might honestly be a better fit for what you are doing. Looks like they run about 1350$ for a 120 cell version, and 1450$ for 144cells. That does not include the wiring harnesses, which did come with the thunderstruck BMS.

One thing to think about too is how much space you have. The Orion is all in one big enclosure, and would not have fit inside my hacked OEM battery box. Not sure if there are other options out there, but maybe someone running an Orion system will chime in.

Here is some light reading:
 

·
Registered
Joined
·
15 Posts
Discussion Starter · #3 ·
I am using the thunderstruck BMS, and it is good for up to 384 cells. Basically each group of 96 cells needs one BMS controller (BMSC), and 3 BMS satellites (BMSS). The cells are wired in consecutive groups of 12, but it does not matter how they are configured - the software lets you map them out however you want. 125 is not optimal, and I think that the chips these BMSes uses is pretty universally based on using 12 cells. If you were willing to live with 120 cells, you could get away with 2x BMSCs and 3x BMSSs. That would be a $1725 BMS, adding the 5 extra cells would bring it to an even $2000.

The Orion BMS handles up to 180 cells on a single unit, so that might honestly be a better fit for what you are doing. Looks like they run about 1350$ for a 120 cell version, and 1450$ for 144cells. That does not include the wiring harnesses, which did come with the thunderstruck BMS.

One thing to think about too is how much space you have. The Orion is all in one big enclosure, and would not have fit inside my hacked OEM battery box. Not sure if there are other options out there, but maybe someone running an Orion system will chime in.

Here is some light reading:
Carl,

Thank you for your input, this is just what I was looking for! I used to live in Lake Oswego, then Tualatin, and eventually moved out to Alaska after ending a converted Skoolie bus trip from Ohio to Panama. I actually have not necessarily an unending amount of room for my battery modules but, I do have an immense amount of free space to utilize for this project which is actually one vehicle utilized in four different configurations at four different times to achieve four different goals then it all gets put into a completely different vehicle for a fifth iteration as a different project happening over the space of 2 years. The upside of the project car portion is that all I need for myself is the drivers seat which leaves me trunk area, under hood area (now housing a V-12), rear seat area, and passenger area.
 

·
Registered
Joined
·
923 Posts
Cells in series just get to higher voltages, dangerous but for many use cases necessary.

Any paralleling if necessary should be limited to 1S groups at the "bottom level" one series string to get to high V afterwards

or maybe two for redundancy, but then you get issues with unbalanced flows
 

·
Registered
Joined
·
923 Posts
These days the term LiPo usually means small pre-wired packs used for model hobbyists

very unsafe, fire prone.

Same li-ion chemistries 3.6-3.8V nominal, packaged in hard shell casing, single cells as small cylindricals or larger pouches are a bit safer.

Safest of all LFP also 10x the lifespan, but then you lose some energy density.
 

·
Registered
Joined
·
15 Posts
Discussion Starter · #6 ·
John,

thank you for taking the time to give your input! Are you perchance saying I should not parallel any other 400 volt series strings together on top of the first string? Due to the possibility of unbalanced flows? In charging mode or output mode? Or both? And why would that happen? If I envision cells 1 - 63 as the bottom and cells 61 - 126 as the top, Are you implying that the top half of the string would discharge and charge faster than the bottom? I can see that happening, 1 - 63 would be trying to re-charge 63 - 126 as 63 - 126 is trying to discharge, would that be a correct way to envision what you might be saying? And if so what is your thoughts on getting to my target voltage safely? Again thank you for your input!
 

·
Registered
Joined
·
15 Posts
Discussion Starter · #7 ·
I am very open to be taught the correct way to view how a pack should be built. I have been searching for an in-depth tutorial on building a lithium pack but most of what I find in rudimentary and does not go into depth on whys and why nots
 

·
Registered
Joined
·
290 Posts
What you dont want is multiple strings in parallel:

Best to have just one string that is made up of electrically joined cells. The pairs in the bottom example will act as though they are one string of cells with double the capacity per cell.

However, there are problems to contend with here that you would not have to deal with if you simply used single cells that were the capacity you wanted to start with. With multiple cells you have multiple possible current paths. If one cell shorts, then the other cells will have a closed circuit (and you will have a fire). If a cell fails as an open circuit, now the whole pack current that used to be divided up will all flow through a smaller number of cells.

Building your own battery packs is a big undertaking. Used OEM battery packs are out there, and the design teams at major car makers have done all the hard work for you. Large format LFP cells like John61ct is a fan of used to be the only solution available, but the cost is high, and finding a reputable supplier is a problem.
 

·
Registered
Joined
·
6,491 Posts
What you dont want is multiple strings in parallel:

Best to have just one string that is made up of electrically joined cells. The pairs in the bottom example will act as though they are one string of cells with double the capacity per cell.
Very true. Every production EV pack is built this way.

However, there are problems to contend with here that you would not have to deal with if you simply used single cells that were the capacity you wanted to start with. With multiple cells you have multiple possible current paths. If one cell shorts, then the other cells will have a closed circuit (and you will have a fire). If a cell fails as an open circuit, now the whole pack current that used to be divided up will all flow through a smaller number of cells.
That's valid, but that's also why a BMS is used which presumably shuts off the contactor when things go wrong. While there are some hybrids with only a string of single cells, nearly every production battery-electric vehicle has at least two cells in parallel.

Tesla has a fusible wire on each cell, which is probably a good idea with the dozens of cells which they put in parallel, but everyone else uses two or three big pouch cells in parallel and simply connects them together.
 

·
Registered
Joined
·
923 Posts
Are you perchance saying I should not parallel any other 400 volt series strings together on top of the first string? Due to the possibility of unbalanced flows? In charging mode or output mode? Or both? And why would that happen?
Because physics. No way the two strings can have precisely the same resistance at all SoC / voltages. Both modes.

Just do all your paralleling at the lowest 1S level.

If I envision cells 1 - 63 as the bottom and cells 61 - 126 as the top. Are you implying that the top half of the string would discharge and charge faster than the bottom?
This is series connection, not parallel, nothing to do with the above scenario.
 
1 - 11 of 11 Posts
Top