Problems while tuning the scale parameter

Hi people
I find a problem happening not always, when estimating a simple NK model. The problem arises when dynare tunes the scale parameter, sticking to 1.000 invariably, and then spending just a few seconds “climbing the hill”. Then, when the MH step goes on the acceptance rate is 1 the whole process, producing an useless output.
I don’t know exactly why this is happening, given that nothing wrong occurred in my first estimations with the same files.
I hope someone passed through the same trouble and could be able to lend me a hand!!
thanks