Constant time step in scikits.odes

Question

For various reasons I would like to use a constant, instead of an adaptive timestep when solving a dae using the scikits.odes package (using the ida solver). In the documentation, it implies that I can achieve this using

dt=1e-5
solver = dae('ida', sol_fs.residual,
             first_step_size=dt,
             atol=1e-6,
             rtol=1e-6,
             algebraic_vars_idx=[0, 2, 3],
             one_step_compute=True, #
             old_api=False)
solver.init_step(t0=0, y0=[0,0,0,0], yp0=[0,0,0,0])
times=np.arange(0, dt*2, dt)
for i in range(1, len(times)):
    solver.step(times[i])

But this still results in an adaptive timestep. Is it just not possible with idas? I'm also having difficult finding the full options list for the idas solver in the documentation.