Would you please send me the numerical laplace inversion code of "de Hoog" 1982?

when I use s:=evalf(...), maple give me reasonable results but when I use s1:=evalf(...) for the same problem the results are different from previous case!

Can you help me, please?

maple fsolve function fails to give correct result for the following problem

aa:=evalf(fsolve(sin(xx*a)=0,xx,0..pi));

if i ignore a coefficient, the results will be correct.

Does someone know why?

Does maple have any function without such limitation?