You need to be sanding and shaping the tips into a rounded dull point. Every time you see a spark you should inspect the tips to see if that point took damage and needs reshapen.
If you’re getting sparks on 50% then you’ve got up to 50% compromised welds. That’s not a spark, it’s an arc and what you see is the spark. Just as much as it can blow holes in the nickel it can create similar damage to the probes of the welder
It’s most certainly not steel…it is very easy to vaporize and doesn’t spark. The sparks are usually pretty small and only from one side, they always pass the pull test still so the welds are mechanically secure
I’m very happy with how this turned out… everything is well insulated (plastic holders space the cells out, fish paper rings on the positive terminals, all nickel has been rounded, all balance wires are layed as flat as possible with an ass load of tape and fish paper, there is a card board plate on the nickel and fishpaper wrapping the cells, it’s been double heat shrinked with the wires coming through little slots), it looks pretty, and it’s idiot proofed (XT90s termination and 5A fuse wired in). I feel this is good enough to give to someone I care about