Can you upload the Maple file?

J. Tarr is right when he points that it should work with variable omega without problem. I checked it out, but another paradox emerged: although, fsolve gave results with both omega and alternative variable, when used with omega the calculations delayed a little. Again with time function: < restart: st:=time(): Digits:=10: eq[1]:=add((-1)^(i+1)*cos(omega*t[i]),i=1..5): eq[2]:=add((-1)^(i+1)*sin(omega*t[i]),i=1..5): eq[3]:=add((-1)^(i+1)*cos(r*omega*t[i]),i=1..5): eq[4]:=add((-1)^(i+1)*sin(r*omega*t[i]),i=1..5): eq[5]:=t[1]: eq[6]:=t[4]-2*t[3]+t[2]: eq[7]:=t[5]-2*t[3]: s1:=fsolve({seq(eq[i],i=1..7)},{seq(t[i],i=1..5),r,omega}): Solutions:=subs(s1,[seq(t[i],i=1..5), r, omega]): t:=[seq(op(i,Solutions),i=1..5)]; r:=op(6,Solutions); omega:=op(7,Solutions); seq(eq[i],i=1..7); time()-st; > If omega is substituted for om for instance, calculations do not delay (maybe this happens at random). Still, since Maple output generates the greek omega (something like a w) for omega variable, then I guess that one should be sceptical for using this variable. But, maybe he shouldn't. pantole

First of all I think you should change omega variable to sth. else, e.g. om, since Maple translates omega as a built-in function [except, of course, you need to use omega]. You may change the code writing: restart: eq[1]:=add((-1)^(i+1)*cos(om*t[i]),i=1..5): eq[2]:=add((-1)^(i+1)*sin(om*t[i]),i=1..5): eq[3]:=add((-1)^(i+1)*cos(r*om*t[i]),i=1..5): eq[4]:=add((-1)^(i+1)*sin(r*om*t[i]),i=1..5): eq[5]:=t[1]: eq[6]:=t[4]-2*t[3]+t[2]: eq[7]:=t[5]-2*t[3]: s1:=fsolve({seq(eq[i],i=1..7)},{seq(t[i],i=1..5),r,om}); Solutions:=subs(s1,[seq(t[i],i=1..5), r, om]): t:=[seq(op(i,Solutions),i=1..5)]; r:=op(6,Solutions); om:=op(7,Solutions);

You can check results by writing: seq(eq[i],i=1..7); pantole