The capping of forces to prevent bad overlaps at the start of simulations is often quite tricky. I find it is helpful to monitor both the minimum distance between particles and the energy:
puts "# Equilibrating charges step $i"
set act_min_dist [analyze mindist]
puts "# step $act_min_dist $i"
set myenergy [analyze energy]
set totE [lindex [lindex $myenergy 0] 1]
set kinE [lindex [lindex $myenergy 1] 1]
puts "# Energy $totE $kinE"
The minimum distance shouldn't be less than about 0.85 if I recall correctly, and the kinetic energy should be around 1.5 kT per particle (assuming you have set the temperature to one). If both of these quantities are reasonable when you've stopped the force capping then there is probably a problem with the variables you chose for your simulation, such as the potentials, the thermostat or the particle density.
I hope that helps,