Did you get this sorted? I'm having a similar problem on a cartesian and I'm wondering if they are related, perhaps.

https://www.duet3d.com/forum/thread.php?id=4026

Richard

How is compensation done? The height of the first layer or bed up/down?

If your sensor is offset in x and/or y then use for calibration doesn't directly require this offset data as it only needs to measure the height of the bed under the probe, but indirectly it needs to know where on the bed it is probing and since the coordinate system refers to nozzle position the offset data is useful. (it will also use the offset to avoid probing outside the bed assuming you've defined your travel limits based on nozzle position correctly).

For compensation/levelling the offset is translated to nozzle position for each probed point, so when the nozzle next passes over that point when printing, it does so at the correct height. Assuming your probe is accurate.

As a maker of probes which have no x or y offsets as they use the nozzle to touch the bed, there are great advantages; no offsets to work out, you can probe anywhere that you can print, they're lighter and less of an obstacle and much more accurate than most offset probing methods. See my signature if your offset probe eventually begins to annoy you, mine did that's why I got into the business of making probes better.

With large parts you can get away with excessively high temperatures or insufficient cooling because when it prints perimeters, by the time it finishes the first one and starts the second, the first perimeter has cooled sufficiently to become solid. That's why the base of the pyramid is ok. What's happening at the sides is that when the second perimeter is printed, the first one is still molten, so it is being pushed out of place by the second one.

Solutions:

1. If using slic3r, set a low speed for "small perimeters". Slic3r has had this setting for at least 4 years, but sadly the other slicers still don't provide it.

2. Reduce extrusion temperature.

3. Increase print cooling.

4. Reduce print speed.

I was wrong. When there is no P parameter, G30 probes from the current position, and the only other parameters that are significant are S and H.

Thx alot…have to do it 😉
Even thats its so much easier than the mega config.h from marlin..sometimes its still too much for a G-Code Noob like me

Those couplers are only held together by friction between the aluminum and rubber parts. They are only good for compression forces. When the nozzle touches the bed, it pulls those type of coupler apart. You can push it back together, but it will just keep happening when you probe.

I used them for a bit, and they do work well in normal operation, they don't have the same springy effects of the typical couplers, but they do have this unfortunate downside. I ended up switching back to spring couplers on my i3.

I just had another thought…

HomeAll.g stores the configuration for the inductive probe (as does config.g, just for safety sake). Homez.g stores the configuration for the nozzle offset switch.

Then, using bed.g

[[language]] M561 ; clear any bed transform ; Probe the bed at 4 points G28 G30 P0 X10 Y10 H0.33 Z-99999 G30 P1 X10 Y290 H0.33 Z-99999 G30 P2 X290 Y290 H0.33 Z-99999 G30 P3 X290 Y10 H0.33 Z-99999 S4 ; Nozzle offset home G28 Z

After digging to validate, it does:

M81 S1 sets private variable Platform::deferredPowerDown to true… (and Platform::Spin() acts on it)
M80 calls platform.AtxPowerOn() which is defined as:

[[language]] void Platform::AtxPowerOn() { deferredPowerDown = false; IoPort::WriteDigital(ATX_POWER_PIN, true); }

I should have checked the source before asking. (Of course, now that it's posted in the forum, the next person might find it with a search.)

Take care
Gary

Id look at extrusion width. So one way to do this is extrude some filament into the air. Let it cool then measure it with calipers. For a 0.4mm nozzle typically it will be 0.48mm diameter due to die swell. For a 0.3mm expect 0.36mm. If so great as the default extrusion width in most slicers is 1.2x nozzle diameter. If your filament in free air is bigger set a wider extrusion width.

Also make sure set your slicer to print inside perimeters first or anything overhanging gets printed in mid air.

I'm glad you solved it. I'm guessing that there is an incompatibility between DWC 1.20 and RRF 1.19.

Short answer yes.
Longer answer I can't be sure as I've no idea what sort of printer you have… But these are not extreme values, so if anything prints badly try changing them in a systematic way, if everything prints good push them up a bit and see what it can do.

Can someone please send me a test print for a cube?
I am trying to print a cube I did in craftware but as it is the first time
i have something wrong as it is over extruding,

If someone can give me a file to try and see if its the printer or the file i made.
Thanks

You are right, you can use the manual bed levelling assistant (which uses M671 and G32) followed by G29 mesh bed compensation.

To use mesh bed compensation, you must either run G29 S0 (or just G29) to probe the bed, or G29 S1 to load an existing height map. You can run G29 S1 in your slicer start GCode, or even near the end of config.g.

Hmm…. thats right. In S3D I have Z-hop on retract and 'force retract on layer change' is checked by default. Silly me... 🙂

Yes that is a good motor to use. The stepper drivers are constant current, so it is normal for the phase voltage to be much lower than the supply voltage.

If you have the trigger value set to 500 in G31 then the probe will briefly produce a reading of 535 but then it will stop and return to the dive height. So the 535 reading doesn't last !omg enough to see.

The small difference in nozzle height at the two corners after bed compensation could be caused by any of several factors. For example, any slight variation in the tilt of the print head between the two corners will change the relative heights of the sensor and the nozzle.