If from (
) we have
and integration in () can be done analytically to yield
Here 
is the modified Bessel function:
and 
is its derivative.
Asymptotic behavior of this function is well known and for times
much bigger than the characteristic time L/2r we have
Eq. () is a very disturbing result, showing that in the case
when inductance l of an individual current source r is much
smaller than the characteristic inductance L of the power
distribution line we have a non-exponential decay of the induced
perturbation in the biasing line. Noting also its independence of the
junction number n we conclude that small perturbations ()
of the biasing current, induced by different junctions switching at
different times can accumulate to give a large value rendering the
entire device inoperational. This effect is determined only by
the relation between L and l is independent of the value of the
biasing voltage. At very large times, of the order of 
, where N is the total number of junctions, asymptotic
() becomes invalid and decay becomes exponential. For any
sizeable design with 
these times typically are of the
order of 
.
