Rebuild of a 36V/12Ah lead-acid battery

thanar

10 mW
Joined
Feb 28, 2018
Messages
22
Hello,

I have an old 3S 12Ah lead-acid battery pack that's too heavy for my needs, and too old for being any good. It's being used on a 250W nominal e-bike. Instead of replacing the lead-acids which would cost around €160, I am thinking of building a pack with Sony Konion US18650V3 2250mAh cells, arranged in a 3P10S fashion (6.7Ah), since I only use the e-bike for around 5km daily and would really like a MUCH lighter battery. I will also probably need a new charger of course. That would cost me €80 for the cells, €10 for the BMS and around €20 for a new Lithium charger, for a total cost of JUST €110!

Question is:
Given that US18650V3s are rated at 10A maximum, would it be OK to use them in such a configuration along with a 35A max 10S BMS, or should I add another row of cells in parallel, making it 4P10S? Would a 250W nominal e-bike motor ever draw over 30A of current? The lead-acid battery case is equipped with a 25A fuse that never failed.
 
Personally, I would go for more cells in parallel if you can afford it. Reducing the C rate on the cells will help them live longer.
 
I have used 10S2P pack to carry as a lightweight backup. The cells are Sanyo GA cells (3500 mah). The range is about 10 miles on flats (7Ah capacity). 10P3S should be sufficient for your needs, but more capacity is always good. As the cells age, they lose capacity, and you will be stressing the cells and will have to deal with large voltage sag and internal resistance. A 10S4P pack is very light, but if you absolutely want 10S3P, it should work just fine for a while.
 
999zip999 said:
How are you going to connect the batteries do you have a spot welder ?
I have grown quite adept at using a DIY solenoid relay spot welder.
 
I finally opted for a 4P design, mainly because my battery holders are 5x4!
Here's a couple of pictures:
View attachment 1
Please let me know what you think of my spot welding:
IMG_7074.JPG
 
Stop doing the parralel strips first. Honestly,.does nobody get this basic fact that the parralel strips need to be LAST?

FIRST you make the series strips as they carry the current. When that is all done you can make a parralel connection for balancing.

Also have the parralel strip go over the side so you can makr the wire on the side and have a flat top.
 
I don't think it makes a different which you do first, parallels or series connections; are you saying that the current having to go through an extra 0.15mm of nickel will add to the total resistance of the battery? By how much?
Frankly, I do the parallel strip first because it makes the whole pack much more robust and you need robust packs when putting together something for a vehicle. Series connections will be over each cell, not just one per cell-group as in the photo. The wires will be connected on three spots, in between the four cells.

Will upload more photos as soon as the extra strip I ordered arrives.
 
Seems like I forgot something... but then again, why not leave it as is? What's the dirrerence? High currents go through series connections anyway! (Talking about not adding half the parallel strips).


 
thanar said:
I don't think it makes a different which you do first, parallels or series connections; are you saying that the current having to go through an extra 0.15mm of nickel will add to the total resistance of the battery? By how much?
Frankly, I do the parallel strip first because it makes the whole pack much more robust and you need robust packs when putting together something for a vehicle. Series connections will be over each cell, not just one per cell-group as in the photo. The wires will be connected on three spots, in between the four cells.

Will upload more photos as soon as the extra strip I ordered arrives.
you are stacking welds for starters. the actual connection is pretty poor by comparison to a direct connection. plese bear in mind that battery resistance is measured in milliohms, so the series connections are a very important factor as those add up in series having 16S battery means you can have up to 16 crappy connections adding resistance. and resistance means more losses, higher voltage drop under load and poor performance. especially in high power and low cell count packs like this were every milliohm counts.
in many batteries the connections have more resistance then the actual cells. this causes heating, uneven balancing and is just bad practice in general.
the notion of pack rigidity is simply flase, the strength comes from the cell holders. you might want to put a dab of hot glue in the corner of 2 cells and the cell holder. that increases the rigidity even more.
it is also safer during constructon as any possilbe short is limited to the batteries you short, not the enitre parralel group.

also, if you extented the parrelel strip you could have all the balance wires on the side where they dont rub or get shorted....

just FYI: i build EV batteries for a living.
 
Looks very nice. Should work very well for your needs. Are these new cells ? Is it well balance before you hook up the bms ? For your use I don't know how much S frist or P frist matters much. They sale premade S and P connections on one strip one piece. So only have to weld to cells. Less connections
How are you going to finish the packaging and protecting the end connections ?
 
Cells are new LGs at 2250mAh/10A continuous. They all measured 3.60-3.61V when put together. I know of the S/P strips, have ordered some, but take some time. The pack was covered with thin foamy material, leaving the BMS cooling plate outside. It was then put into the original 3x12V lead-acid battery case, leaving lots of empty space of course. I used additional foamy material to make the pack fit tightly so it won't move around. I am using the original external connections as before. XT60 and spade for charging are hidden inside the large battery box.

IMG_7210.JPG
 
I was thinking of that red cardboard they use with a thin yellow flat fiber glass board on top of the terminals ? I seen it on packs but don't know where to go it.
 
I have been testing a couple nickel strips I got from different sources. Testing voltage drop @1A, both strips 0.15mm thickness, 8mm width. Results vary a lot:

  • Strip I got from Banggood, EU warehouse: 0.3V@1A (2m length) -> 150mΩ/m
  • Strip I got from random eBay seller: 0.09V@1A (1.5m length) -> 60mΩ/m
  • 2.5mm^2 copper cable: 0.009V@1A (1.20m length) -> 7.5mΩ/m

Does that make sense?
What puzzles me more is the fact that the best nickel strip I got has a tad lower resistance than an AWG23 copper cable. Why are we using nickel strips then? Just connect the cells with AWG20 cables!
 
Did you check the nickel for real-ness?

Sandpaper a bit and drop it in salt water for 2 days. No rust is good.

Ps: you cant weld 20awg wire. Soldering is bad Mmkay.
 
Cooper can tyrn green over time in a battery connection. Plus some spot welders we're having trouble with copper. How about copper strips plated with nickel or some other plating material. Would be custom or time involved.
 
999zip999 said:
Cooper can tyrn green over time in a battery connection. Plus some spot welders we're having trouble with copper. How about copper strips plated with nickel or some other plating material. Would be custom or time involved.

You cannot spot weld copper because spot welding actually requires high resistance which produces heat, so as the metals will melt momentarily and bond together. Copper has very low resistance, so when trying to spot weld copper the current will just go through it and nothing will happen.
 
Yes I have read that I do not have a spot welder. I had read of people nickel plating copper but I didn't know just a nickel plating would add enough resistance to weld. Plus the third photo down it almost looked like you had used copper strips in the photo or the color of the strips in the photo. Or and I guess I've kind of misread things and what my eyes saw ?
 
Back
Top