DIY Electric Car Forums banner
1 - 8 of 8 Posts

·
Registered
Joined
·
69 Posts
Discussion Starter · #1 ·
Greetings to all.

I am new to the EV world and I am starting on my electric motorcycle conversion. I acquired some a123 26650 batteries, used, but in good condition still. They were built in packs of 40V for electric buses.

Anyways, I am planning my pack design as I have almost all of the components to start building the pack.

I have attached the layout of what my thoughts were. With my tight dimensions I have with my bike, I am doing a 28S 17P pack. I am more concerned about the individual connections to make sure I have the right concept.

Please let me know your guys thoughts of what I can do. I am doing 17 "layers" of the 28 cell blocks. a 4 x 7 block is the 28 cell block layer so to say. The top view is what each block looks like and have those stacked 17 times to achieve my desired dimensions.
 

Attachments

·
Registered
Joined
·
6,200 Posts
It looks like that's actually a 7S17P4S configuration (strings of four cells in series are connected with 17 strings in parallel then four of those 4x17 sheets are connected in series), unless there are somehow connections vertically between layers at every cell junction. What is the plan for a BMS?
 

·
Registered
Joined
·
69 Posts
Discussion Starter · #3 ·
The middle block has them wrapped like a snake where the positive end is on the top left corner and the negative is the bottom right. making the 4x7 area a 28S if I am looking at it correctly. The attached document may help as the green is the series connection, the teal is the parallel connection and then you have your positive and negative leads off the ends.

The BMS I have is ZEVA modules. I have 3 of the 12 unit modules only using 28 out of the 36 available spots for this configuration.
 

Attachments

·
Registered
Joined
·
6,200 Posts
The middle block has them wrapped like a snake where the positive end is on the top left corner and the negative is the bottom right. making the 4x7 area a 28S if I am looking at it correctly. The attached document may help as the green is the series connection, the teal is the parallel connection and then you have your positive and negative leads off the ends.
Yes, but if a 4X7 layer (in series) is horizontal (so that middle view is from the top), then the drawings do not indicate whether you have vertical connections between cells other than on the faces with the terminals. You can connect all of those BMS wires to the cells in the top layer, and they will be connected to the individual cell level in the top layer, but unless you have vertical connections every cell, 16 layers of the cells won't be connected to the BMS except every fourth cell.

If you unfolded the snake to make a 17 cell high and 28 cell wide sheet, positive on the left and negative on the right, would there be vertical paralleling connections for every column of cells, or only only every fourth column where they were exposed on the left and right sides of the folded block? In terms of your colour coding, are there 35 of the teal paralleling connections, or just 14 (seven on the positive face and 14 on the negative face)?

Unless you build it as 4 cell wide by 17 cell high sheets, then join those into the whole block, it's hard to see how you could have all of the connections in the middle. I suppose you could set up a forest of 5 by 7 vertical wires or bars, then connect 4x7 layers of cells to them, but it doesn't seem practical.
 

·
Registered
Joined
·
69 Posts
Discussion Starter · #5 ·
Yes that makes sense and I just realized that mistake. Thank you for clarifying. I have updated the drawing with attached document. Could I basically make one large layer of 28S 17P and have extra nickel strip for the fold where the dashed lines are? In between each fold would be an insulated, non-conductive material to prevent shorts. Then have the bms wires attached to each 28S row essentially.
 

Attachments

·
Registered
Joined
·
6,200 Posts
I have updated the drawing with attached document. Could I basically make one large layer of 28S 17P and have extra nickel strip for the fold where the dashed lines are? In between each fold would be an insulated, non-conductive material to prevent shorts.
Electrically, yes - that makes sense. Mechanically, I don't know what will work with your fabrication method, but the folding idea would get the desired electrical configuration into the desired package shape. :)

I'm not sure what the plan is for structural support. You can't just stack the cells with nothing holding them together other than the electrical connections - some sort of framework is appropriate.

Then have the bms wires attached to each 28S row essentially.
It would be a BMS connection to each of the 28 columns, each column being 17 cells connected in parallel.
 

·
Registered
Joined
·
69 Posts
Discussion Starter · #7 ·
That was my next step was for the framing of the module. Once I fold the pack, I'll make measurements to make a cage like structure around it with either angle iron or some other fabrication method. I appreciate the feedback as well greatly! Building larger packs come with a lot of complexity when you break it down by any means.
I also heard of the a123 26650 battery cells I am using requires a vent when spot welding the tabs on? Is this true or?
 

·
Registered
Joined
·
6,200 Posts
That was my next step was for the framing of the module. Once I fold the pack, I'll make measurements to make a cage like structure around it with either angle iron or some other fabrication method.
I doubt that holding just the outer edges of the stack is adequate, especially with a rectangular (rather than hexagonal) stacking pattern. I suspect there would still be mechanical stresses on the electrical connections, but who knows - maybe it would work.

I also heard of the a123 26650 battery cells I am using requires a vent when spot welding the tabs on? Is this true or?
I have no idea about that. Modern EV battery modules connect cells with ultrasonic welding.
 
1 - 8 of 8 Posts
Top