First you need to understand that the upper bound of a Proctor curve is saturation or "zero air voids". As for a boundary of 5% for moisture that is not the case.
When you run a moisture-density relationship (Proctor) test, the moisture contents should be chosen so as to "bracket" the expected optimum moisture content. Let's assume your optimum moisture content would be near 6%. Your curve points should be set up at moisture contents of 4%, 5.5%, 7% and 8.5% or your could use 3.5%, 5%, 6.5% and 8%. This should produce a well-defined curve wherein the graphical peak of that curve is the maximum dry density. Occasionally, some soils exhibit "double peaks" with one peak being slightly lower than the maximum dry density peak and somewhat dryer than optimum. This usually happens in clean fine sands.
If you are dealing with clayey materials, it is important that you allow the moisture to fully distribute through the sample so that proper results will be achieved.
Post your Proctor curve and each of the values so that we can help you.
