My 2nd Fuse link 18650 battery build

takyka

10 mW
Joined
Apr 1, 2015
Messages
32
Location
Hungary
Hello,

After almost three years of my first battery taken into use, I started to build a new battery.
https://endless-sphere.com/forums/download/file.php?id=215011&mode=view
I consider the 1st attempt as a great success. It is a 6S4P NCR18650PF pack soldered with fuse links at the negative terminal. Provides energy for a kick scooter. Seeing 40A peak (10A/cell) and around 15A continous. Since the building, the battery have around 100 cycles, used between 4,1-3,0V. Not measured, but no sign of degradation in real life use. Also, the difference between cells are not more than 30mV from each other, so no need of balancing. (Cells are protected by cell level LVC only)

Based on the above, It was logical to follow the same route with the new build. This case the application is to feed an 1,2kW Electric outboard (Haswing Protruar 2.0HP) on different small boats for river and lake operation.
As the motor is designed for 24V, I decided on 7S configuration. I set the capacity target to 500Wh, with the possibility to increase it to 1kWh.
I was checking the forum for a while, what would be the best, but still affordable cells for this application and came to the conclusion, for me there is still no better value for the money than the PFs. Finally I decided on 7S8P configuration, means 584Wh capacity.
During my first build, I made tests to find out the suitable size for fuse wire. I checked different diameter conductors that time (0,2-0,5mm) how fast burning out when an aged, almost empty, cold cell from a laptop (worst case scenario) shorted. That time I found, 0,3mm is a good size. In that application I did not care much about losses due the short bursts of high current consumption and the existence of cooling airflow. In watercraft propulsion, continuous full throttle operation without much airflow is a realistic scenario, extra heating has to be kept as low as possible. Test revealed, at the designed maximum continuous current of 6,25A/cell I get 22mV drop on a fuse link. That is 0,14W heat/cell, 7,7W for the entire pack.
 
So dia 0,3mm fuse wire will be fine. This case I decided to use BMS. I need a carefree battery. As the cells don't have willingness to drift away much, I ordered a basic 60A 2,5-4,2V BMS with 50mA ballancing capability. I ordered a battery switch and a circuit breaker, but undecided yet, to use the breaker or trust in BMS overcurrent + fuse wire protection. To connect load and charger, I will use XT90 antispark connectors.
The cells are placed in custom milled plastic trays. From one side of the plate there are the battery cavities with openings for connections, on the other side pockets for wiring. The parallel connections I make on positive terminals with solder wick and fuse wires will be connected to here from the edge of negative terminal. The two trays will be screwed together to keep the cells tight.
 

Attachments

  • bms.png
    bms.png
    88.2 KB · Views: 1,736
  • circuitBraker.jpg
    circuitBraker.jpg
    31 KB · Views: 1,736
  • switc.jpg
    switc.jpg
    32 KB · Views: 1,736
  • batt_holder1.png
    batt_holder1.png
    14.3 KB · Views: 1,736
  • pack.png
    pack.png
    30.2 KB · Views: 1,736
I machined the holder plates from some leftover pieces of plastic.
akkufedlap.jpg
Batteries are placed to "bottom" half:
View attachment 1
Here things were happening so fast, I forgot to take more pictures during building.
akku2.jpg
As you can see, thick wire "busbar" soldered to the + terminals, where any damage to the cell is unlikely due the long thermal path. And just tiny amount of solder to fix fuse wire to the edge of the can on - side. I were using low melting point leaded solder & high temperature thick soldering iron. I went row by row and every time a paralel group were ready I immediately added kapton tape to protect the cells from accidental sorting.
In less than two hours I was able to finish with soldering.
 
takyka said:
As you can see, thick wire "busbar" soldered to the + terminals, where any damage to the cell is unlikely due the long thermal path. And just tiny amount of solder to fix fuse wire to the edge of the can on - side. I were using low melting point leaded solder & high temperature thick soldering iron. I went row by row and every time a paralel group were ready I immediately added kapton tape to protect the cells from accidental sorting.
In less than two hours I was able to finish with soldering.

Looks like a superb build.

What diameter and type of wire are you using for your fuses?
 
After assembling the pack, I made preparations for a real life test run of the battery. I added temporary main wires and simple car fuses. I placed the pack in a 16" plastic toolbox. To monitor the cells, I used a temp meter and a Lipo alarm.
IMG_2568(1).JPG
The load is a Haswing Protruar 2HP 1,2kW electric outboard motor, installed on my inflatable boat.
https://youtu.be/YMmEKE_66uQ
I did a ~10minutes full throttle run to see how the cells behaving. At around 50A load (just estimated from power rating) 6,25A/cell, I have got 20C temperature rise without any air ventillation arround the tightly packed cells in the box. The fuse links were cold to touch. At the end of charge I made another cca. 5 minutes full throttle operation, until voltage dropped to 2,7V under load. As one could expect, the first and last cells in the series got to the lowest voltage as those were the coldest due the ability to radiate more heat. By recharging the battery at home, I was able to push in 22Ah meaning I have left only 5% of the nominal capacity by the end of my run.
 
Nice job on that build - so
it looks like each cell is fused but only on the positive terminal is that correct.
Or
is the fusing just on the parallel connections not the series - schematic if you have the time
Much appreciated
Dennis
 
Back
Top