From G28 to G33 and beyond !

Hergonoway

Hi, I'm french, my english is not the best ( :) )

I've got a delta and want to prevent futur full re-calibration, after reading the official page from repetier website, I still have few question :

1 - To to a full autocalibration I'll have to send the command (I don't write the manual operation, positioning Z or move carriage ) : G92 -> G28 -> G131 -> G132 S1 -> G32 S2 -> G33 ; seems good ?

2 - If I understand correctly G32 do some magic so the printer see the bed as something totally flat (horizontaly I mean) right ?

3 - I didn't find any advance description of G33 command, does G33 modify the rod length in the EPROM and correct also curvature of the hotend movement? From what I red, its seems to correct the curvature of the bed if this one is not totally flat.

4 - Is there any benefit to do a G32 then a G33, the opposite, or re-doing a G32 then a G33 after a first pass ?

Hergonoway

I forgot a question (my bad)

5 - Does I need to use M320 to M323 if I mainly use G32 and/or G33 commands ?

Repetier

Between G131 and G132 you need to position your 3 sliders to have same height using a measurement rod or something. For ths M99 X0 / M99 Y0 / M99 Z0 can diable single motors for 10 seconds as shown in the video. The G132 calibrates endstop offsets. I use same distance from bottom (not bed but where bed is mounted on) as reference.
2. Yes,No magic, just mathematically rotate the coordinate system.
3. G33 does not change any physical parameter. It just moves z along the measured bumps so you get better bonding on the first layer. Then it reduces correction over height as you defined. Or does not reduce if you choosed that.
4. G33 after G32
5. The commands enable autoleveling/distortion correction until you say otherwise, so not needed.

Hergonoway

Thanks a lot for your answers !

So basicaly, with G32/33 there's no EEPROM delta physical parameters modified, I'll have to do it the manual way to prevent curvature from the hotend ?

In the comments of the repetier firmware source files there's a schematic in ascii art with the different "things" to measure for a delta

Schematic :

CARRIAGE_HORIZONTAL_OFFSET (recommend set it to 0)

DELTA_DIAGONAL_ROD (length)

END_EFFECTOR_HORIZONTAL_OFFSET (recommend set it to 0)

ROD_RADIUS (Horizontal rod pivot to pivot measure)

PRINTER_RADIUS (recommend set it to ROD_RADIUS)

But in the parameters to set there's no ROD Radius :

#define DELTA_DIAGONAL_ROD 345

#define DELTA_MAX_RADIUS 200

#define END_EFFECTOR_HORIZONTAL_OFFSET 0

#define CARRIAGE_HORIZONTAL_OFFSET 0

#define PRINTER_RADIUS 124 // measured from the center of the print area to the vertical smooth tower

In the schematics it's recommend set PRINTER_RADIUS to ROD_RADIUS, in the parameters there's no ROD_RADIUS but instead the comment suggest to measure it from the center of the print area to the vertical smooth tower, which is not the ROD_RADIUS.

There's something I can't understand, if OFFSETs are set to 0 (because recommended, even if there are physically ), and I measure the PRINTER_RADIUS like explained (from middle to tower) and measure my DELTA_DIAGONAL_ROD, How can I estimate the horizontal or vertical height from pivot to pivot of the rod (at XO YO) which I suppose is mandatory to calculate the movement of the hotend ?

I'm not a geometry mathematician sorry if it's "obvious"

Repetier

The original config has CARRIAGE_HORIZONTAL_OFFSET and END_EFFECTOR_HORIZONTAL_OFFSET which compute

#define ROD_RADIUS (PRINTER_RADIUS-END_EFFECTOR_HORIZONTAL_OFFSET-CARRIAGE_HORIZONTAL_OFFSET)but they do not influence formula at all, all needed is the horizontal distance between pivots when carriage is at the center position.
You get the distance from your vendor or building plans. Measuring yourself is quite difficult. You could but a weight on a string and let it down from upper pivot while nozzle touches bed. Then it is the distance from weight touching bed and center.

Hergonoway

OK I didn't see this formula. So ROD_RADIUS is the horizontal distance between pivots when carriage is at the center positio, so I can directly try to measure it and set to printer radius (with offsets = 0) or compute it from real offsets and real printer radius, right? I don't have this info, so I can try to measure it with 2 pendulum attached to the screws which fixed the rod. But it kind of an adventure :) I think it's easier to measure offsets and printer radius rather than measuring the distance between pivots. Carriage and effector are in aluminum so it'll be easier I think. I'll give a shot, thanks for your help!