After some tests I found that Ex Px was not able to solve my problem, partly because of the command itself and partly because the problem is more complex than I expected.
Sample code.
G00 X100 Y100
G01 X110 Y110 <------ From the G00 to the G01 movement there is a stop (accelerations-deccelerations)
G00 X200 Y200
E3 P1 <------ Output4 is not activated
G04 P1000 <------ Wait one second
G01 X300 Y300 <------ Output4 is activated together with the movement
E3 P0 <------ Output4 is deactivated with the start of the next G00
G00 X400 Y400 <----- Bacause of the stop (G01->G00) and the fact that Ex Px is activated with the next move command I still get a dot.
I also get a few liquid outside of my contour because of the valve response time and the liquid that is "flying".
M18 <----- M seems to work better with my semifluid liquid. The time delay of the M gives the liquid time to touch the surface
G01 X495 Y495
E3 P0
G01 X500 Y500 <----- This was my last trick. I close the valve some mm before the end of the movement. There is no acc/decc from
G01->G01 and the "flying" liquid (from the valve to the surface) is placed well on the contour.
I am using Sheetcam and I don't know if I will be able to implement this trick to the postproseccor. (freehand contours!)
Is it possible to know when I am getting to the deceleration phase and deactivate my output from inside Mach