You have to make use of the Rietveld method. If you are very well familiar with it you might also obtain quantitative results of the phyllosilicate assemblage.
Clay XRD is a world in itself! It requires a different approach to standard mineral analysis due to the nature of clay minerals. For a methodology to identify clays one of the best places to start is the Moore & Reynolds book - link below. This will give you all the practical information you need about oriented air dry, glycolated and heated preparations. Any software can be used at this stage, all you need to be able to do is obtain accurate d-spacings, the rest is done by inspection. There is a section on quantification in that book that can have a look at. For qualitative work you could also look at the USGS site - link below. Some of the information there is a little old fashioned, but the chart works.
For a software approach to quantification you can use the modelling program Newmod - link below. This will work for clay separates. If you are trying to quant the clays in a random powder mixture of crystalline minerals it starts becoming more challenging. SIROQUANT has a clays package, BGMN is good but I have found it to be slow. I generally use TOPAS for this task.
It is worth going and reading about the Reynolds Cup - link below, and seeing how previous entrants have approached solving this problem - yes there is an international prize for quantifying clays in mixtures, which should give you an indication of the difficulty!
Just a note about your range 4-30 2-th. Starting at 4 2-th you risk missing structures such as corrensite. I would start at 2 2-th. 30 2-th as a maximum is fine if you are looking at clay separates, but do run random powders up to 70 2-th so that you can examine the 060 reflections.