Claudio,
It's not clear from your first question whether you are correlating betas across subjects between the two tasks, or are correlating results in 1 subject across different events in the 2 tasks.
In either case, you have to create 2 datasets (via 3dbucket or 3dTcat), that contain the beta "time series" (or "subject series") -- one dataset for each task. Then 3dTcorrelate will calculate the correlation coefficient -- or similar measure -- of association between the 2 data series, voxel-by-voxel. Of course, the 2 datasets have to have the same number of volumes (=sub-bricks) for this to work.
3dttest++ isn't really appropriate, it seems to me -- the voxel-wise covariate is not the right thing to use, since you have a large number of "time" or "subject" points in each dataset.
If you want to look at the correlation maps between a fixed seed voxel in one task dataset vs. all locations in the second task dataset, you can do this interactively with InstaCorr, using the ExtraSet option. That is, you would pick a seed location (say the insula) in the task#1 dataset, and see how correlated all the voxels in the task#2 dataset with that one time series from the first dataset. You could use this method to interactively and dynamically explore the spatial inter-dependence of the 2 task activations. If you are especially adventurous, you open 2 AFNI controllers and do the task#1->task#2 InstaCorr in one window and the task#2->task#1 at the same time in a second window!