using a VERY large of tstab in PSS analysis

I wonder if it is possible to improve convergence in PSS analysis by using a VERY large of tstab (stable time)?
Is there anything bad about that approach?

I read that in order to increase the possibility of convergence we need to increase tstab. Is that right?

Yes and No..
Depends on your circuits "Steady State Settling Time"..
Convergence is also depended on some other factors.