Making a small pack 5S24P

leffex

1 kW
Joined
Apr 8, 2012
Messages
443
Location
Sweden
Just some photos of my small pack I want to share

How to assemble: Soldering
exactly how: tesla style. One side is fused with wire capable of 5-15a max(deciding), the other more robust fastened with thicker copper wire. Both sides uses twined copper wire as busbar. Hot glue gun to make pack sturdier and also additional plastic and soft material inbetwen + manually activated fan cooling.
Size: 18,5x30x7,5 cm
Numbers of cells 5s24p
Cell type: Used tested Panasonic 18650PF
Usage: max 150A, 6,25 per cell
Total wh: 1221 (18,5Vx66AH)

YyDyZKi.jpg


WVVmrlI.jpg


u8gE9oh.jpg
 
Work table:
o3WyCNJ.jpg


1 of 5 pieces
8rObp1F.jpg


5 pieces fits inside one
OHpeWNe.jpg


Finished pack:
ONDnoex.jpg




6ZqiWck.jpg


Module pack stats:
18,5cm*7,5cm*30cm = 4162cm^3 Volume = 4,162 L
5s*24p 2,9PF nominal 1221wh
1221wh / 4,162 L = 293wh/L

Next mission, what to do. Test pack for voltage drop at set load and monitor temperature
 
what cell holders did you use?
 
Something like cylindrical or 6squared plastic cell holders. The ones I used was 1x4p so I added them together and broke some to make smaller pieces were needed

These are very alike but not the ones I have
https://vruzend.com/product/hexagonal-18650-battery-holders-3-cell/

I first thought not of using any but then its pretty much better in all ways to use plastic cell holders than not to
 
Another pack mocked up for photos:

7MzkxJc.jpg

q5RSl0X.jpg


Module 2 pack stats:
18,5cm*7,5cm*30cm = 4162cm^3 Volume = 4,162 L
5s*26p 26F nominal 1196wh
1196wh / 4,162 L = 287wh/L

Power rating 26p x 5,2a = 135,2A

I'm thinking about what these modules can give in relation to power. I don't want the cells to get warm and they are only 26p for a normal usage of 1,7kw/h or 65w / km. That means the power from the full battery will be average 1700w used per hour and 1/4th of the energy will come from module 2 consisting of 1196wh module which is 1196/4 = 299wh. That makes out to average load of 1700w/90v = 19a shared load for 26 parallel cells. hmmm it will work. easy :D
 
i found the cell holders on ali: https://www.aliexpress.com/item/32919687859.html

i have orderd a set and see if they can take self-destructing cells without catching fire or melt too bad that it can lead to a catastrofic ending for a battery. that is going to be a fun -but smelly- afternoon.
 
I searched for a box that fits the space I want to use and put my battery into. I can put 4 of them into my space so I started with ordering 2 boxes to try.

Ordered another two from here:
https://www.aliexpress.com/item/33004388619.html?spm=a2g0o.cart.0.0.77773c00w03GYq&mp=1


cwah said:
How do you find the right size box?

Also to find out how big your box or pack should be try to decide its Voltage and AH or how any cells you need.

each cell is 1,8cm wide and 6,5cm in height and a little wider with a plastic cell holder depending if you use one or not . If a box is 20cm in width and 20cm length and 8,5cm in height then 20/1,8=11,11 cells can fit on one side and the other is the same so another 11 (because we can't split cells) Maximum potential is 11x11 cells in that box so 121 cells. If you pack them as tight they can be maybe a few more cells may find its way into the pack.

First I was thinking of making a box from building material. Plastic squared tubes but easiest way of doing it is to get a box that is in perfect size already and go from there even though it may cost you some.
 
Just tested module one 5s24p with panasonic PF cells.
They have a delta voltage between cells in the module below 0,11v at 3,95v and 150a load. Maybe I didn't balance them so it could even be lower. (Delta voltage between cells at no load was 0,005v)

However. Total pack voltage at testing was 19,75v or 3,95v per cell. At a load test of 150a or 100-150A the voltage sag came to a maximum of 1,91v. That is almost 10% voltage sag at a very high load.

Estimated voltage sag levels at lower amperage levels:
10a load = 0,13
25a load= 0,32
40a load= 0,51
50a load=0,63
75a load=0,95
100a load=1,27

Estimated average usage 50a

Pack module temperatures. No difference at first test for 5 minutes and 2km. A slight difference in temperature outside the module box of 3 Celsius degrees higher than surrounding temperature after second run made 10 minutes later.

End voltage of each cell was 3,895 with delta between cells at no load at 0,006


vigXSte.jpg

*edit added photo - 20190920

Next test will be on Saturday 21st of September and 8km in distance times two.
 
you need way more copper on your bus if you wat better performance. same goes for the inbalance during load. you need way more beefy series connections.
 
flippy said:
you need way more copper on your bus if you wat better performance. same goes for the inbalance during load. you need way more beefy series connections.

What happends when I double or tripple the bus bars? Better performance...

Pack module cable are short 10awg 20cm and battery cables are oversize awg 2, 25mmsq^2 to the controller.

From each cell group there are at least 4 copper wires going to the next cell group apart from the copper connecting all cells of course as seen in the pictures. The copper awg I would estimate to be at least 12awg

Thanks for your input and help
 
Let's crunch a few round numbers. This isn't intended to be exact. 24P is 24 * 3 Ah for a 72 Ah cell. 150A discharge is slightly greater than 2C discharge rate. Let's look at the battery specs:

8808930F-6C4F-4589-9DAF-FBE1BB81C4FC.jpeg

Now I've never seen a chart using "lt", presumably load test. I'm going to assume lt = C, which is capacity since it fits the graph for a max 10A or slightly greater than 3C discharge. So the cell specs call for a .22V voltage drop between .2C and 2C rate. For 5S, that's 1.1V.

Interpolating the test results for .2C load, .2*72 = 14A. The observed pack voltage drop at 14A, about .2V, so the perfect world drop is 1.1+.2=1.3V versus the observed 1.9V, a .6V delta.

Plus it's really greater than 2C due to rounding. The factory data for a single cell is probably impossible to achieve in a 100 cell configuration. I'd say there is little room for improvement. I'd be pretty satisfied.
 
BlueSeas said:
...So the cell specs call for a .22V voltage drop between .2C and 2C rate. For 5S, that's 1.1V.

Interpolating the test results for .2C load, .2*72 = 14A. The observed pack voltage drop at 14A, about .2V, so the perfect world drop is 1.1+.2=1.3V versus the observed 1.9V, a .6V delta.

Plus it's really greater than 2C due to rounding. The factory data for a single cell is probably impossible to achieve in a 100 cell configuration. I'd say there is little room for improvement. I'd be pretty satisfied.
That was interesting facts to know!


I did a longer test drive today. Two times 8 km. Power used was about an hour of charge of 95vx5.5a so I used about 522wh of energy. 32wh/km.

Pack module 1 was not hot or anything after "feeling it". On this test drive however I did put the voltage meter on ONE of my 500 cycles RC-lipo multistar 12c rated 4s1p10AH connected in 2p as 20AH. Delta measured from it was 2,16v. They can't be compared directly because of 5 series vs 4 series but in this case the PF is the better of the two for reasons of course.

measured max delta voltage at 150a load
PF, 0,38v delta / cells (66ah)
RC, 0,54v delta / cells (20ah)
calculated estimates:
RC, 0,27v delta / cells (40ah)
RC, 0,38v delta / cells (28,5ah)
 
Making another pack. This one has 25p because the cells should measure up to 66ah and so one more is needed in each group beacuse of lower cell capacity.
Difference with this module is that I'm trying more copper in the series connections. 8 instead of 4-5

oepJA4h.jpg
 
At last a new year and time to use those batteries.

I have in total build three of these batteries and here they are in all their glory.

WEJsj83.jpg


The first was a 24p which was then changed to 25p for the two last packs. There's a 2s Sanyo Ga 20p ready to start its life soon in the left corner as it get connected. Other than that it will be interesting to se how this will work. First version module has 5x 12awg twined copper wire between each series connection and the newer batteries have 8x.
 
Back
Top