DIY Electric Car Forums banner

Battery split question

6K views 38 replies 7 participants last post by  Functional Artist 
#1 ·
Hello I am very new electric conversion and ready to work on one. I have bought pretty much all the parts and in the process designing location of everything.
Below is the battery layout. There are 12 batteries total and use as 6S2P and is split to 8 back and 4 front.
The back battery will run all the way to the front and will that cause voltage difference since it is 10 ft?
Do you think this is a good idea?
I do wish to max out at 1000amp
Thanks you,
Ryan
 

Attachments

See less See more
1
#4 · (Edited)
Below is the battery layout. There are 12 batteries total and use as 6S2P and is split to 8 back and 4 front.
The back battery will run all the way to the front and will that cause voltage difference since it is 10 ft?
There is voltage drop in every conductor, and there will be more drop in the long (10 foot) connection between modules than in the shorter connection, but since it is a series connection between pairs of modules that won't matter to the balance between modules - the current goes through each of them, and that's the important part.
 
#31 ·
Hey, would love to follow up on this discussion since I have a similar question. Am considering a battery pack of 192 total cells arranged in 2p96s configuration, made up of 16 modules each in the 2p6s configuration in a VDA355 form factor. I have room for all 16 modules under the hood, and really like the idea of a single module for simplicity of the enclosure, BMS, and cooling. I will be liquid cooling using some sort of cooling plates. However, because I am also designing for roughly even weight distribution front to back in the finished car, I am considering splitting the battery pack into two. Mostly likely choice would be to keep the 2p96s configuration, put 12 modules in pack 1 under the hood, and 4 modules in pack 2 in the rear.

With a single pack, I would be able to use a radiator and cooling loop for the single battery pack all under the hood. HV cables that would run the length of the car would likely be those between the charge port and charger (mounted in the rear) and battery pack (mounted in front), and from battery pack (front) to motor and controller (rear). With a split pack, I would have to run additional cooling lines from front to back (assuming I put both packs into a single cooling loop, which I think is easiest, and to keep the battery cooling loop separate from motor cooling loop, even though pack 2 would likely be much closer to the motor itself), and would also need to run additional HV cables from front to back to connect both packs (again, in series). I think I should be able to use a single 96-cell capable BMS to monitor and manage both packs, treating it as a single 2p96s pack from the BMSs perspective. (I have noticed that some people use two BMSs units for two packs in a lead/follow configuration, but am not really sure why...)

My questions are:
  • aside from some slight voltage drop across the long pack-interconnect HV lines, any downsides to two separate packs in series like this (aside from the added complexity of just having two)
  • would this setup make it any easier or harder to identify bad modules?
  • could a single BMS be used as I suggested?
  • would you expect the cooling to be effective in this setup? I would expect the larger pack to get hotter than the smaller pack, but also rely on the radiator and coolant pump to do their best to not just dissipate heat but to keep the temperature difference between the packs to a minimum since they are linked by a single coolant loop
  • anything I have forgotten to consider here? Fatal flaws with the idea of a split pack in series config?

thanks
 
#7 ·
The parallel connections are done by bus plates. Because there is one pair of threaded sockets for each pair of 8 Ah cells, in the 6P modules the plates have three rows of holes (at the ends) and six rows of holes (3 for positive one cell group plus 3 for negative of the next cell group); to reconfigure these modules to 5S12P you would replace those plates with plates twice as long (six rows of holes at the ends and 12 rows of holes for the other plates). With 5S12P modules you would then connect all modules in series for the same 30S12P overall pack. This would simplify the BMS substantially, but if the seller doesn't offer different sizes of bus plates (they don't appear to) you would need to custom-build the longer plates.
 
#9 ·
I have consider that but it seems like copper busbar prices are too high that's why I followed this diagram by Orion bms and It seems like it will work for me. If it does not seem like a good idea to follow orion bms I'll find a way for longer busbars.
Rectangle Slope Font Parallel Pattern
 
#20 ·
#21 ·
#26 ·
kW/kWh is a kW/kWhanadian term, apparently

Everybody but Brian writes his kW/kWh as "C"

Fires start and pack life gets totally trashed when you "hope".

For Lithium, COOLED batteries can generally do 3C continuous, 5C peak. Meticulous engineering using a compute farm of over 1000 GPUs, like the kids at Tesla and SpaceX have, get you that 8C, maybe 10C.

There's no magic. Hope gets you nothing in engineering. Ever cross a bridge, thinking the cable was designed for bicycles but you're hoping it wouldn't snap with a car going over it if you go quick enough? That's some Wile E Coyote thinking going on, dude.

Live within the current limit or buy the second pack by force to act again as one limited pack. You don't get Tesla amps (or even 1000 of them) just because you wrote it on the internet after buying half the battery and a "1000 amp" controller.
 
#27 ·
So, if with proper cooling (air) and a 1000amp capable busbar. Is it worth/capable to possibly get up to 500amp comfortably?

The cells are rated to discharge 200a continuous and able to pulse 400a. So with 12 in parallel each cell shouldn't have too much load on them right?
 
#29 ·
So, if with proper cooling (air) and a 1000amp capable busbar. Is it worth/capable to possibly get up to 500amp comfortably?
10C (now that you know what "C" means, including that it makes no sense dimensionally) is wildly inappropriate, and 3C is optimistic by comparison with well-designed production EVs... so why would 5C be appropriate without liquid cooling? The module cases are plastic, so air cooling will be very ineffective.

The cells are rated to discharge 200a continuous and able to pulse 400a. So with 12 in parallel each cell shouldn't have too much load on them right?
An interesting spec. No reasonable person would claim that an 8 Ah cell can handle 200 A continuous and 400 A intermittently; those are the specs from the retailer (Battery Hookup) and are given by various retailers, but the cell manufacturer isn't even known, and I really doubt that they would claim those rates if you could find them (the images of these cells that I have seen have labels without any indication of the manufacturer).

Apparently these "SPIM08HP" cells are widely used for DIY e-bikes. If getting incredible discharge rates was just a matter of buying hobby-grade cells, all the EV manufacturers would use them... but of course they don't. I don't have any reason to believe that there's anything wrong with the cells, but I also don't have any reason to believe that they can perform any better than production EV cells... so my guess (and it's just a guess) is that it might be reasonable to plan on 3C peak without active cooling, and much less continuous.
 
#28 ·
"Proper" air cooling CANNOT give you better pulse performance than published. It's a limitation you're stuck with.

Busbar has nothing to do with what the battery can deliver, only with what the busbar can deliver, unless it was deliberately undersized in its original application. The 5C rule of thumb applies, regardless, as a cell limitation.
 
#32 ·
You need to be able to shut off all voltages coming out of the front pack and the rear pack for safety reasons.

That precludes running 96 wires willy nilly all over the car to a single BMS, since those wires always have voltage on them and are usually not fused. So, you'll need a BMS master in one box connected with galvanically isolated communications to a satellite BMS in the second box.

You also need two contactors inside each battery box. One for plus, one for minus of its HV battery connection.

Interbox HV cables should be run separated, so a responder with a chop saw does not melt the sawblade with 2000 amps running through it. Ideally protected from road debris damage (including innocent-looking rocks, gravel, dropped car parts, and traction grit).

HV charge port cables need to run to a box with 2 more contactors applying voltage to the HV battery in the proper sequence if you are planning DC charging.

Depending where you live, you likely will need a coolant heater in the loop for the batteries as well.

Any heat source connected in series will heat the downstream heat source. So, no, you will not be equalizing the temperatures.

You can run the rear battery box's coolant loop through the motor (not the other way around) if you want to be frugal about number of loops.
 
#33 ·
super helpful, thanks. Yeah, separate contactors and BMS...this is why i'm trying to avoid separate packs. I live in southern california, so hoping to skip coolant heating, but have also been thinking about at minimum designing cooling loop in which I can easily add heater later, in case I move back to a cold place. I will keep messing with the design to see if I can maintain the single batter pack approach, but I may give up and take on more complexity, we'll see.

On a related note, I have been looking for guidance on battery box enclosure design, but not finding much. Easiest thing for me will be to design a sheet metal box that can be folded and welded. I think 18 gauge (1.2mm thick) steel sheet is reasonable to create a rigid box that will hold all the components. BUT, this is going in a car, so I'm thinking about crash and impact safety. My gut tells mem that 18 gauge steel is too thin for attaching mount points and other fasteners that may be needed on the enclosure. Eyeballing videos online from Electric Classic Cars and Zero EV, I think they're using thicker sheet metal. Any guidance on how to select a sheet metal thickness, or other considerations in enclosure design?
 
#36 ·
I'd point out that structural portions of a car body use doublers and triplers and that car body sheet metal that is holding the appreciable weight of a battery module could either tear out a fastener or buckle under load and sudden loads like potholes and trees. The car bodies of today use high strength steel to withstand wind loads...the real structure is under the skin.
 
#37 ·
I'd point out that structural portions of a car body use doublers and triplers and that car body sheet metal that is holding the appreciable weight of a battery module could either tear out a fastener or buckle under load and sudden loads like potholes and trees. The car bodies of today use high strength steel to withstand wind loads...the real structure is under the skin.
Yes I agree, I was just addressing metal thickness

A battery box, just like a car body is just a facade
...& must be properly supported

The 18g steel outer "skin" of the battery box that I built, for the mini-Jeep is basically, a protective cover
...for the battery module, battery cables & connections

So, I used (2) pieces of 1" x 1" x 1/16" angle steel to "double" (& re-inforce) the (front & rear) bolting flanges
Handwriting Rectangle Font Material property Parallel

These (2) "cross-members" also, support the weight of the battery module & the box itself
...& then, by connecting to the frame
...the whole unit is then supported by the entire chassis of the vehicle
Wood Bumper Automotive exterior Rectangle Gas

My gut tells mem that 18 gauge steel is too thin for attaching mount points and other fasteners that may be needed on the enclosure.
Yes, like Remy mentioned just double or triple the steel, in the areas that may need extra strength
...but, doubling the thickness of the entire box, would just add unnecessary weight ;)
 
#38 ·
Really helpful thanks! I'm thinking of following similar approach. Considering battery box wall thickness, areas that need reinforcement and thicker steel eg around mounting points, then separately designing either a way to bolt into existing engine mounts from donor chassis, and/or adding additional subframe elements to existing chassis to support battery box. New subframe elements would be beefier, eg box member or possibly tubular steel, TBD. Not there yet.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top