I see a series stack of caps and MOS gates that has no
DC path to ground in several spots (unless your MOSFET
model has gate tunneling leakage, which is not that likely).
Now I have no idea about ADS harmonic balance simulation
but you might check whether it's using a good DC (OP)
solution, or has tried to make its own and perhaps ignores
the settings which might be helping a DC solution (gmin
etc.). Maybe you can get a DC convergence manually
and force it to be used explicitly. But the error message
is not really enlightening, and maybe the "no convergence"
is some other thing than the DC solution (for example PSS
fails when cycle end mismatches cycle start, such that
"period" is not "steady state").