I would like to create this battery pack, consisting of 13 Samsung ICR18650-26F connected in series. There are balancing leads for balanced charging, and resettable polyfuses connecting all of the batteries. My current issue is determining the proper fuse rating.

Fuses are rated in amps, but how does the voltage rating play a factor? Would I be able to use a resettable polyfuse rated at 16v 4a, or 30v 4a? Overall, how would I calculate the proper fuse necessary for this application?


simulate this circuit – Schematic created using CircuitLab

*This is my first post, so I hope that I was able to be as direct as possible.

  • \$\begingroup\$ Why are you putting resettable fuses between each series battery? \$\endgroup\$
    – Filek
    Apr 5 '16 at 5:06
  • \$\begingroup\$ If any fuse opens, it will have the full pack voltage across it. So all of the fuses must be rated for the full pack voltage. So that is 4.2*13. \$\endgroup\$
    – mkeith
    Apr 5 '16 at 5:09
  • \$\begingroup\$ @Filek From my understanding, this is the method Tesla, and others employ to prevent an internal short or over-discharge of a single cell from effecting the other cells in the series. Although Tesla's are non-resettable from what I have read. \$\endgroup\$
    – Alex Sky
    Apr 5 '16 at 5:24
  • \$\begingroup\$ I suspected that was your reason, but they use those fuses for parallel batteries, not series. I cannot really see the benefit for series, but for parallel it makes sense for the fuses when you have a lot of high power batteries. \$\endgroup\$
    – Filek
    Apr 5 '16 at 5:28

Fuse ratings in a design should be set so they interrupt the current for an over-current condition. You determine how much current your design needs. If your load cannot tolerate large currents or there are extreme safety requirements, you may want many fuses. They should be rated so they won't trip during normal use, but trip if something shorts out. This really really depends on the load so follow these steps:

1) Find your loading current (If its an active load like a motor then experimentally find the nominal current with a power supply)

2) Give yourself some margin, if your load is drawing 2A normally, then multiply that by a saftey-factor so that if there is a voltage spike it won't blow the fuse out. A 50% saftey-factor would give you 3A (of not tripping). If the load shorts (lets say 1Ohm worst case) Then for a 50V pack you would get 50V/1Ohm = 50Amps). So you would need a fuse that blows with less than 50Amps of current, but give yourself some margin in that direction too, and maybe get a fuse that trips at 10Amps.

3) Find a fuse with the rated voltage and current, keep in mind every fuse takes time to blow or trip, will the load or source have a problem with that? If the fuse takes 1 second to blow, its probably not going to protect the batteries, if it takes 1us it probably will because not much heat will be dissipated anywhere.

Edit: 4) If you have a captivate load, you will want to watch the inrush current. (Li-pos are "kind of" capacative in the fact that they are a very low ohmic source and their chemsitry can deliver lots of current fast.) There are slow blow fuses that are more suited to this sort of thing. Double check the inrush rating and design for it.

  • \$\begingroup\$ Thank you, this is very useful. I completely understand points 1 and 3 (I will have to consider options for point 3 in another topic), but I lost you halfway through point 2. You mention the 'load', which in my case is indeed a motor, I would need a 10amp fuse. Does that imply using 13, 10A fuses rated at 54.6v to connect each of the batteries in the series? Would this also apply to the case of an internal short of one of the cells? \$\endgroup\$
    – Alex Sky
    Apr 5 '16 at 6:38
  • \$\begingroup\$ If the motor is rated for 10A that is under maximum conditions. Connect it to a power supply (with a simmilar voltage, or measure it with the batteries) and measure the current under normal torque conditions and then under worst case torque conditions, then size the fuses around that. If your worst case torque is 10A, and you have a 10A fuse, you could potentially blow the fuses. Also keep in mind you only need one fuse to protect the circuit, but if you are really worried about the cells shorting to each other then also fuse each cell. \$\endgroup\$
    – Voltage Spike
    Apr 5 '16 at 16:39
  • \$\begingroup\$ This is great information. Pertaining specifically to protecting each individual cell from each other, would those fuses need to be rated based on the individual cells, or the circuit as a whole? For example, your suggesting one fuse for the entire circuit to protect it from the motor, but If I have 13 fuses between the batteries, as displayed in the diagram, would they each need to be rated to the same specs as one for the whole circuit? I'm not replicating his battery, but it has me stumbled (16v 4a fuses across his serial connections). youtube.com/watch?v=5EE4TsuzD_Q \$\endgroup\$
    – Alex Sky
    Apr 5 '16 at 23:48
  • \$\begingroup\$ I would watch some more youtube videos. That one isn't the best! That particular guy doesn't seem to be worried about the amount of heat he is applying to the batteries via a hot glue gun or the massive amount of solder. He even overheated his soldering gun - very dangerous with 18650 lithium! Also, would you consider regular non resettable fuses? If one of your batteries has an internal short, it is dead forever, and your fuse will NEVER reset, so why spend the money on resettable fuse? Also resettable fuses usually work by getting rather hot as they near their 'blow' amperage... \$\endgroup\$
    – Filek
    Apr 6 '16 at 6:14
  • \$\begingroup\$ A hot fuse is not good for lithium battery. \$\endgroup\$
    – Filek
    Apr 6 '16 at 6:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.