Validity of setting constant to zero

#1
Can anyone tell me the validity of setting the constant term to zero to make the other coefficients pass significance?

For example, I have been sent a multiple regression model in Excel, the model has had the constant term set to zero and the other coefficients are significant. However, when I recalculate the same regression model without setting the constant to zero - all of the coefficients fail to reach significance. Is the setting of the constant to zero a valid action?

Thanks

Craig