How to generate the standard normal random vaiables from LCG method

Second time to post here, and I hope if there could be anybody give me some help, I'd appreciate that very much!

Because of the RAND() in excel is unable to replicate the results, so my assignment is to use the LCG(Linear Congruential Generator) with parameters c = 5100, d = 1251, m = 9999, and the seed is u0=500 to generate 1000 numbers as u(1) to u(1000).
so ,in the LCG function, it u(k) = (c*u(k-1) + d) MOD m , so 0<=u(k)<=m-1=9998. Then devide the u(k) by m, assigned as t(1) to t(1000),
so 0<= t(i) <= 9998/9999. And in my sequence, there is a t(i) = 0 which is possible. But, how could we use these t(i) to generate the associated stadard normal random variables, which is Norm.s.inv(t(i)) if there's a t(i) = 0 ?? How should I modify these t(i) generated from LCG function to generate standard normal random variables ???

Plz give me some help, or any comments.I appreciate for your kindness.

Thanks, but when t(i) = 0, 1-t(i) =1, Norm.s.inv(1) would result a invalid value ,too . Right?

Hi, Mr.shg, have you ever used @Risk, could you help me out with this problem I posted? Thanks very much.

http://www.excelforum.com/excel-gene...obability.html