New Delta - effector not tracking flat during print? - help with autolevel/correction

Background:
Custom Delta printer
RUMBA electronics utilizing DRV8825 drivers (set to no more than 0.8A for each driver)
55 oz/in NEMA17 motors - 1.8* step
355mm diameter build platform
480mm build height
Repetier 0.92.8 firmware
Repetier Host 1.6.1

I can't find a really good, detailed starter guide for custom setups.  I just tried a short sequence of things I will list:

Set in firmware:
483mm z height (x/y/z max all set same - also in EEPROM)
383mm diagonal rod length
204mm horizontal offset
z probe enabled

- Powered up the RUMBA, disabled steppers, centered each carriage on its tower so effector is at true 0,0.
- Ran the G131 (clear any offsets), then G132 S1 command to measure/save endstop offsets in EEPROM - completed successfully
- Ran 3 separate scripts that would bring the effector down right in front of each tower close to the max bed radius right until the hotend touches the bed (with a piece of paper in between).  I then ran a G100 X (and subsequent Y and Z for the other scripts that brought the effector down in front of each tower).  This sets the X/Y/Z mins.  By the way, each script would first home all axes before bringing down the effector.
- I would then run a G100 r - but I found this would calculate a diagonal rod radius that was too small.  It would give me 201mm.  This causes linear distortion when running across the bed.  I found 204mm to be close enough my eyes can't tell the difference.

Now, here is where I am confused.  I ran both G32 S2 and G33 on two separate occasions.  Neither seemed to make a difference with the issue I am having.  The G32 correctly measured all the points in the grid I defined.  The G33 also correctly measured all the points in its planar grid array (granted G33 doesn't automatically home when it is done).

Now, I can manually track the effector across the print surface and it appears to run perfectly flat across the bed.  When I draft run a print, the effector lifts the closer it gets to the Z axis tower, but lowers back down as it approaches the edge of the print bed.  This 'bump' is not centered on the print bed - that would indicate a horizontal offset error.

HELP!
Sign In or Register to comment.