Z-Correction not working properly on delta

Fughazi

I followed the video guide on using G32 and G33 to level the bed on a delta. However, regardless of whether z-correction is applied, my height map looks like this:  

Here are the height map text results:

X: -65.00 -32.50 0.00 32.50 65.00

y:65.00 40.00 40.01 40.04 40.00 40.00

y:32.50 40.01 40.00 39.99 40.03 40.00

y:0.00 40.01 40.00 40.01 40.01 40.01

y:-32.50 39.96 40.00 39.99 39.96 39.98

y:-65.00 38.61 39.98 40.04 40.01 40.00

And here are the results (including the offset amount) from G33:

00:40:38.418 : Z-probe:38.61 X:-65.00 Y:-65.00 zCorr:0.04

00:40:39.947 : Z-probe:39.98 X:-32.50 Y:-65.00 zCorr:-0.20

00:40:41.539 : Z-probe:40.04 X:0.00 Y:-65.00 zCorr:-0.34

00:40:43.135 : Z-probe:40.01 X:32.50 Y:-65.00 zCorr:-0.41

00:40:43.711 : X:32.50 Y:-65.00 Z:40.000 E:0.0000

00:40:44.283 : Z-probe:40.00 X:32.50 Y:-65.00 zCorr:-0.41

00:40:46.775 : Z-probe:39.96 X:-65.00 Y:-32.50 zCorr:-0.26

00:40:48.360 : Z-probe:40.00 X:-32.50 Y:-32.50 zCorr:-0.24

00:40:49.950 : Z-probe:39.99 X:0.00 Y:-32.50 zCorr:-0.36

00:40:51.543 : Z-probe:39.96 X:32.50 Y:-32.50 zCorr:-0.41

00:40:53.138 : Z-probe:39.98 X:65.00 Y:-32.50 zCorr:-0.36

00:40:56.022 : Z-probe:40.01 X:-65.00 Y:0.00 zCorr:-0.33

00:40:57.613 : Z-probe:40.00 X:-32.50 Y:0.00 zCorr:-0.29

00:40:59.202 : Z-probe:40.01 X:0.00 Y:0.00 zCorr:-0.26

00:41:00.791 : Z-probe:40.01 X:32.50 Y:0.00 zCorr:-0.30

00:41:02.383 : Z-probe:40.01 X:65.00 Y:0.00 zCorr:-0.34

00:41:05.266 : Z-probe:40.01 X:-65.00 Y:32.50 zCorr:-0.27

00:41:06.852 : Z-probe:40.00 X:-32.50 Y:32.50 zCorr:-0.20

00:41:08.432 : Z-probe:39.99 X:0.00 Y:32.50 zCorr:-0.15

00:41:10.013 : Z-probe:40.03 X:32.50 Y:32.50 zCorr:-0.17

00:41:11.600 : Z-probe:40.00 X:65.00 Y:32.50 zCorr:-0.31

00:41:12.171 : X:65.00 Y:32.50 Z:40.000 E:0.0000

00:41:12.740 : Z-probe:40.00 X:65.00 Y:32.50 zCorr:-0.31

00:41:15.224 : Z-probe:40.01 X:-32.50 Y:65.00 zCorr:-0.11

00:41:16.802 : Z-probe:40.04 X:0.00 Y:65.00 zCorr:-0.10

00:41:18.382 : Z-probe:40.00 X:32.50 Y:65.00 zCorr:-0.19

00:41:18.949 : X:32.50 Y:65.00 Z:40.000 E:0.0000

00:41:19.523 : Z-probe:40.00 X:32.50 Y:65.00 zCorr:-0.19

Am I missing something where you have to apply the offsets? Thanks.

Repetier

Actually these are very good results. You have only one value at -65/-65 that looks wrong and makes the image ugly. All other values are so near to 40 that I would not change a thing. That wrong error might have a reason maybe some pla on surface ... can not say without having your printer here.

Fughazi

Repetier said:

Actually these are very good results. You have only one value at -65/-65 that looks wrong and makes the image ugly. All other values are so near to 40 that I would not change a thing. That wrong error might have a reason maybe some pla on surface ... can not say without having your printer here.

Thanks for catching that! I checked my build surface and there wasn't anything on it, but I flipped the glass over and rotated it just to make sure it wasn't causing it. I ran the bed map a few more times, and it behaves as if it's caused by software rather than hardware. Regardless of how far out I start it or how many probe points I do, it's always the first one that is off. Here are the results from a few different bed maps:

X: -65.00 -32.50 0.00 32.50 65.00

y:65.00 40.04 40.03 40.04 40.03 40.02

y:32.50 40.04 40.00 40.02 40.04 40.04

y:0.00 40.03 40.00 40.02 40.00 40.02

y:-32.50 39.99 39.99 40.00 39.99 40.04

y:-65.00 38.45 40.02 40.03 40.02 40.00

X: -60.00 -30.00 0.00 30.00 60.00

y:60.00 40.08 40.00 40.02 40.00 40.04

y:30.00 40.00 39.99 40.00 39.99 40.00

y:0.00 40.02 39.96 39.99 39.99 39.98

y:-30.00 40.03 40.00 39.98 39.99 39.98

y:-60.00 38.53 39.99 39.99 40.00 40.00

X: -60.00 -30.00 0.00 30.00 60.00

y:60.00 40.08 39.99 40.00 39.98 40.02

y:30.00 39.98 40.00 39.99 39.99 39.98

y:0.00 39.99 40.00 39.96 39.99 39.98

y:-30.00 40.02 39.97 39.98 39.98 39.98

y:-60.00 38.50 40.00 39.99 39.99 39.98

X: -55.00 -27.50 0.00 27.50 55.00

y:55.00 40.06 39.98 40.02 39.99 39.99

y:27.50 39.98 39.99 39.99 39.99 39.98

y:0.00 40.03 39.97 39.98 40.02 39.99

y:-27.50 39.97 40.02 39.98 39.95 39.99

y:-55.00 38.48 40.00 40.00 39.98 39.95

(Increased to 10x10 points)

X: -60.00 -46.67 -33.33 -20.00 -6.67 6.67 20.00 33.33 46.67 60.00

y:60.00 40.03 40.02 39.98 39.98 40.00 39.99 39.98 40.02 39.98 40.02

y:46.67 40.08 39.98 39.98 39.98 40.00 39.99 39.98 39.94 39.93 39.99

y:33.33 39.99 39.99 39.99 39.98 39.99 40.02 39.98 39.97 39.99 39.97

y:20.00 40.02 39.98 39.98 39.98 39.94 39.97 39.98 39.97 39.98 39.98

y:6.67 40.05 39.98 39.95 39.95 39.93 39.98 39.98 40.00 39.97 39.98

y:-6.67 40.03 39.97 39.96 39.95 39.94 39.98 39.98 39.99 39.98 39.97

y:-20.00 40.00 39.98 39.99 39.99 39.97 39.95 39.95 39.98 39.97 39.99

y:-33.33 40.05 39.94 39.98 39.96 39.95 39.98 39.99 39.98 39.98 39.99

y:-46.67 39.98 39.89 39.96 39.95 39.99 40.00 40.00 39.99 39.97 39.97

y:-60.00 38.52 39.90 39.94 39.97 39.99 40.00 40.02 39.98 39.95 39.99

Even if the maximum is set to 30, the first point it probes is ~38.5

X: -30.00 -15.00 0.00 15.00 30.00

y:30.00 40.00 40.00 40.02 39.97 40.00

y:15.00 39.99 39.98 39.98 39.95 40.00

y:0.00 39.99 39.97 39.99 39.99 39.99

y:-15.00 40.02 39.97 39.96 39.97 39.97

y:-30.00 38.56 39.97 39.97 39.95 39.94

It seems like only an issue with the bed map, but actual prints show an error in that general area as well. Regardless of how many times I run correction, the bed map continues to show that error along with actual prints: 

Thanks for the help, this is the closest I've gotten to a working printer in weeks!

Fughazi

Never mind, it seems to have been the result of a poor quality z_probe/mount. It has now improved dramatically:

X: -55.00 -27.50 0.00 27.50 55.00

y:55.00 30.00 30.00 29.99 29.99 29.97

y:27.50 29.99 29.99 29.99 30.01 30.00

y:0.00 29.99 29.97 30.00 30.02 30.01

y:-27.50 30.02 29.97 29.97 29.99 29.97

y:-55.00 29.92 29.97 30.02 30.04 30.02

I see it isn't quite even on the map yet, but the scan results are pretty good. Still, .12mm may be enough to throw off my first layer. Are the red and blue still within the margin of error or should I look for more accuracy?

Also, is there any way to probe the whole bed as shown below, because I currently have the largest square I can within the circular build plate and I think it may be ignoring some pretty crucial areas.

 Thanks again!

Repetier

I really would not bother. Normally you print first layer 0.2 - 0.3mm (I always use 0.3) just to get good bonding and no problems. But even with 0.1mm height it might print well. I would be happy about such accuracy myself. Since you have so much nonlinearies and error sources this is all you can expect.