G29 for Delta. Weirdness

kazolar

I must be doing something wrong. I did g29 r69 s23 and the result is not a circle but a square. And it doesn't probe the full bed.

dc42

That command makes no sense in RRF. See https://duet3d.dozuki.com/Wiki/Using_mesh_bed_compensation.

kazolar

right, I do M557 R69 S23 then G29

kazolar

The heightmap is generated -- top looks weird:
-46.00,46.10,-46.00,46.10,69.00,23.00,23.00,5,5

I never gave it square bounds, just a radius

kazolar

fixed it -- not sure what changed -- I updated firmware to 2.02RC6 -- anything above that gives me trouble

dc42

@kazolar said in G29 for Delta. Weirdness:

The heightmap is generated -- top looks weird:
-46.00,46.10,-46.00,46.10,69.00,23.00,23.00,5,5

I never gave it square bounds, just a radius

That's normal, it constructs a square grid that covers the radius you specified.

dc42

@kazolar said in G29 for Delta. Weirdness:

fixed it -- not sure what changed -- I updated firmware to 2.02RC6 -- anything above that gives me trouble

What sort of trouble does 2.02 release give you?

kazolar

2.02 the spikes produced are super pronounced. Also as I mentioned before 2.02 has issues with i2c end stops -- they stop responding after 4 hours or so of printing. I haven't bothered to troubleshoot that, so I just run 2.02RC5 or RC6

dc42

@kazolar said in G29 for Delta. Weirdness:

2.02 the spikes produced are super pronounced.

What spikes?

kazolar

when I did g29 there is a significant difference with RC6 and final. Nothing else changed -- this is a little 3 color mixing diamond hotened delta I've repaired for a friend with a maestro board + additional 2208 in uart mode. Doesn't make much sense for there to be spikes since least squares for 13 points says difference is 2 micron and g29 shows .8mm spikes. (basically flat with RC6)

dc42

I think you are referring to height map spikes, but you haven't posted your height map. Neither have you told us what type of Z probe and bed surface you are using.

kazolar

@dc42 yes mesh map spikes -- bed is aluminum. I'm using a bl-touch. After g32 it's telling me variance is 2 micron, so I'm going to try and print with no map, just least squares leveling. Will try 2.02 final with that again.