ABL with longer probe arm issue

[Schnupp]

Hello Ive been using ABL for months now and it has been working flawlessly. I just changed my hot end out for a different one that happens to be much longer. Since the new hot end is longer than my old one I had to print a new Z servo probe arm so that I would not have a hard crash! A picture below shows my new setup. When I go to print the probe homes itself like normal by extending the probe, touching down, and then retracting. Then when it moves over to probe the first point of the ABL routine the z axis moves down before the Z probe extends... This would cause a crash if not stopped . Any ideas what my problem is?

Things Ive tried in marlin are:
#define Z_RAISE_BEFORE_HOMING 4
#define Z_RAISE_BEFORE_PROBING 15
#define Z_RAISE_BETWEEN_PROBINGS 5
Ive changed these to 25 but it did not work. I also changed the z offset accordingly for my new set up.

I also think the problem could lie in my G-code. Here is a copy of my starting code:
G28 X0 Y0 ;home X & Y
G1 X95 Y95 F5000 ;move nozzle to center of bed
G28 Z0 ;home Z
G29 ;probe bed
G90; set absolute coordinates
G92 E0; reset extruder distance
G1 Z5 F300 ;move nozzle up 5mm for safe homing
G1 X0 Y0 Z0 F5000; move nozzle to home


new set up.jpg

[Schnupp]

It looks like my problem is with the G29 command. When I do a G28 command everything works as should be. Then when I do a G29 right after the z axis moves DOWN about 5mm and then move over to the first probing point. This causes the probe to hit the bed when extended. Any ideas why my Z axis is moving down at the start of my G29 command?

EDIT: Just changed #define Z_PROBE_OFFSET_FROM_EXTRUDER -27.47 back to what I had before (-0.89) and the G29 command works again. I guess there is a limit to how negative you go?

[Mjolinor]

Have you also set the safe height between probes?

#define Z_RAISE_BEFORE_PROBING 1 //How much the extruder will be raised before traveling to the first probing point.
#define Z_RAISE_BETWEEN_PROBINGS 1 //How much the extruder will be raised when traveling from between next probing points

I don't know if that number is absolute or relative, it's just a suggestion.