Fuse curves

[junkhound]

OK, understand I squared t, melting, etc, and that at microsecond times fuses break vs melt.

Looking at some snad filled type fuse curves (e.g littelfuse L50S), there is a slope to the curve in the sub-millisecond region indicative of a cube function of current.

Why cube? -- or is that just a coincidence transitioning from melting to breaking?

 

[Ingenieur]

how did you determine this?
time = k x 1/i^3 ???

<0.001 sec???

 

[junkhound]

just drew a straight line along the logarithmic fuse curve at times < 1 ms. Slope is 10x times = 1000 X current on the fuse curve.

 

[Ingenieur]

not sure that is valid
10 A fuse

time(sec)...i(A)
100.......28
1..........40
0.006....60

slope = (100-0.006)/(28-60)=-3.12 sec/A
slope = (100-1)/(28-40)= -8.52 sec/A
slope = (1-0.006)/(40-60)=-0.05 sec/A

not linear

guessing of the form time = k x e^-(i - i100)
i100 = i at 100 sec
don't feel like messing with the math