moving problem

Those basically look OK, so I'm not really sure whats going on
I do think I would remove the parallel port driver and read/write commands

Did you run a latency test on the PC?
Have you had any other errors from LinuxCNC?
What version of LinuxCNC are you running?

remove the parallel port driver?

PCW wrote:

Did you run a latency test on the PC?

Yes i did

PCW wrote:

Have you had any other errors from LinuxCNC?

No, nothing else at the moment

PCW wrote:

What version of LinuxCNC are you running?

2.7 wheezy

I dont see anything obvious wrong in the hal/ini files

Could you plot the following error with halscope? something very strange
is going on here to get such huge following errors at such slow speeds
(maybe something wrong with the PC clock)

Are you using a 5I25 or 6I25? (wrong jumpering on a 6I25 might cause
this type of an error by selecting an unexpected clock frequency)

I am very new to this Linuxcnc, how do I follow the error? My card is a 5i25 + 7i76. if I leave the speed at 10 there is no following error, only that while machining I lose the Y for no apparent reason, the engine ceases to exist, restart and I have it again, but it happens again

Ok so are you saying that the hal/ini files you posted
(with 8.33 mm/s maximum per axis velocities)
Does not cause following errors?

The files are those, the following errors comes when I increase max_vel over 10, for example with max_velocity=12 and stepgen_maxvel=15

That's probably expected unless you also raise the PID output limit
(probably best to just set it to 0 so its not used)

P=0, I=0 and D=0?

Tried and nothing, the same, if I increase velocity over 10 the following errors appears.

No, the PID parameters are OK but
set all MAX_OUTPUT = 0 in the ini file

It works!!! You are a crack PCW, thank you very much, now I'm moving at 30 without problema, tried 50 too but the motor stalls after a while, doesn't matters, 30 is enough for me

On the other hand, losing the Y axis after a while of machining can be x driver heating? coincidentally in both tests is the axis that works the most, maybe is silly what I am saying