Incorrect data entry is often the cause of a bad first star slew but in this case, I think it may be something else.
Miquel said he made an error in the first alignment and then redid the alignment. I took that to mean that the first stars of that initial alignment were within tolerance as normal and that he did not power down but just back out of the menus and then go back into the alignment routine to redo it. That was the main clue that I used to suggest the corrupted CE/NPE values.
I encountered exactly that scenario when I had a similar problem years ago. I was doing a three star alignment on a partly cloudy night and just as I finished the third star, I realized I had centered the wrong star. (The correct one peaked out from behind a cloud as I finished.) So, I simply went back to the Star Alignment menu and began a new alignment. I did not power down between alignment attempts. The first slew for the realignment ended up pointing at the ground. I powered down and reset the mount to the Home Position. After entering all the initialization data again, the first slew went crazy once again. I tried a few more times of powering down and starting over with similar results. Finally, in frustration, I did a Factory Reset and everything was back to normal. Only later did I figure out that the bad three-star alignment had corrupted the CE/NPE calculations and stored some silly values. The same thing happened about six moths later under similar conditions. That time, I used the menus to manually reset just the CE and NPE stored values and all was well once again.
If you powered down between your alignment attempts, then what Dan suggested could have been the cause. But if you can no longer use any first star and get close during alignment after rechecking data entry, then try resetting the Hand Control or zeroing out any values in the CE and NPE stored in the HC.