confusion about carriage offsets in IDEX mode

oliof

Now I am even more confused. Out of over abundance I did not dare move the print heads with the wrong X position displayed in DWC. But copy/mirror mode actually do move the print carriages correctly. So maybe one for @chrishamm? PS I love that DWC3.5 does only show copy/mirror mode once and switches to the (last) selected one automagically. Nice save of screen estate (-:

T3P3Tony

@oliof are you still having this issue?

oliof

@T3P3Tony I have not updated to 3.5-beta3 yet, but will do that in a hot minute. I will report back after that.

oliof

@T3P3Tony okay, updated Duet 3 mini 5+, DWC, and toolboards to 3.5-beta.3, and Wifi to 2.1b4. Thanks to @jay_s_uk for sorting the RPico firmware with two different addresses quicker than I could have set up a working build environment.

I am still confused about certain things, and there may be bugs about my tool changing macros (Sometimes active/standby temps are set to 1/0; I probably would need to set the temps for T2/T3 but I could swear at least one occurrence I did set the tool temps with G10 and they were reset to 1/0 ... but that's for another discussion)

Here is how things go when I activate tool 0, then switch to tool 1, then tool 2, then tool 3, then tool 0 again. Logs below, as you see X position changes to unexpected values when switching tools.

Also, when I switch from T0 to T1, T1 tries to go to the X position from T0 -- but that is beyond the axis minimum I did send for U axis. How would I make sure that does not happen? I am aware that usually yes, I want T1 to take the position of T0 for a multicolor print; but I would expect T1 to keep within the axis limits of the U axis. How would I do that?

4/16/2023, 9:40:17 PM 	T-1
4/16/2023, 9:40:23 PM 	M114
X:-139.800 Y:-105.000 Z:16.000 U:140.700 E:0.000 E0:0.0 E1:-1.5 Count -11184 -8400 25600 11256 Machine -139.800 -105.000 16.000 140.700 Bed comp 0.000
4/16/2023, 9:40:46 PM 	T0
4/16/2023, 9:41:05 PM 	M114
X:-139.800 Y:-105.000 Z:16.000 U:140.700 E:0.000 E0:2.0 E1:-1.5 Count -11184 -8400 25600 11256 Machine -139.800 -105.000 16.000 140.700 Bed comp 0.000
4/16/2023, 9:41:50 PM 	T1
4/16/2023, 9:42:05 PM 	G28 U
4/16/2023, 9:42:16 PM 	M114
X:140.700 Y:-105.000 Z:16.000 U:140.700 E:0.000 E0:0.0 E1:-1.0 Count -11184 -8400 25600 11256 Machine 9999.900 -105.000 16.000 140.700 Bed comp 0.000
4/16/2023, 9:42:47 PM 	T2
4/16/2023, 9:43:55 PM 	G10 P2 S200 R180
4/16/2023, 9:44:00 PM 	M114
X:55.450 Y:-105.000 Z:18.000 U:140.700 E:0.000 E0:2.0 E1:-1.0 Count -11184 -8400 28800 11256 Machine -54.550 -105.000 18.000 55.450 Bed comp 0.000
4/16/2023, 9:44:39 PM 	T3
4/16/2023, 9:45:34 PM 	M114
X:-85.250 Y:-105.000 Z:21.000 U:-140.700 E:0.000 E0:0.0 E1:-3.0 Count -11184 -8400 33600 11256 Machine -195.250 -105.000 21.000 85.250 Bed comp 0.000
4/16/2023, 9:46:07 PM 	T0
4/16/2023, 9:46:24 PM 	M114
X:-85.250 Y:-105.000 Z:21.000 U:140.700 E:0.000 E0:2.0 E1:-3.0 Count -6820 -8400 33600 11256 Machine -85.250 -105.000 21.000 140.700 Bed comp 0.000

T3P3Tony

@oliof

Interesting observations, its been a while since I have used an Idex machine so i may need to check back and forth with you a bit to get to the bottom of this.

change from T0 to T1:
T1 uses U as X, as expected to the X user position is showing the U position (the Machine Coordinates look like they are reported oddly by M114, X being 9999.9 is odd, one for @dc42 ). The reason why its trying to move to the X position could be because of a restore position (G1 R2) in your tool change macros somewhere?

Change from T1 to T2. X is now both X and U in copy mode. With 109.1mm between the two "X" axis machine positions of the heads (54.550 *2) that's also a bit strange unless you have updated your config.g to change the distance between the two tools from the 110 you have in the config.g posted a few days ago in your first post in this thread?

Also i think you may have missed this point from the documentation:

In this example we want to print one object using the X carriage and another identical object 100mm further along using the U carriage. So we specify X and U offsets of 50mm in opposite directions. However, tool offsets are given as the offset of the nozzle relative to the print head reference point, so the signs are opposite to what you might expect because tool offsets are subtracted from the required printing locations during printing. So we need to use an X offset of +50mm to force the X carriage to be 50mm lower along the X axis, and a U offset of -50mm to force the U carriage to print 50mm higher along the X axis.

in your config.g you have:
G10 P2 X110 Y0 U0 S0 R0
when i think you should have
G10 P2 X55 Y0 U-55 S0 R0

T2 to T3 mirror mode, lets look at that after sorting T2.

oliof

@T3P3Tony

T0 to T1) yes, second head tries to move to the restore point, but its going out of bounds for the axis second head actually sits on. It's not a problem during most normal operations except for very wide prints since the first head does not clear the printable area completely

T[01] to T2 -- the documentation note does not apply this way in my case since my original bed dimensions are zero centered.

Same with T3, only difference is the U axis is mirrored and I lose some Y to keep the print heads from crashing into each other.

The main issue here is that the displayed X position is wrong, the axes work as expected (I did a combined print and a print in copy mode to verify this).

Anyways, here are my config.g and my tool change macros for completeness sake.

oliof

tpost3.g

T3P3Tony

@oliof said in confusion about carriage offsets in IDEX mode:

T[01] to T2 -- the documentation note does not apply this way in my case since my original bed dimensions are zero centered.

ahh right, that may explain the odd reporting. i am not sure that zero centering the bed means that the two tool offsets should be different through. ultimately you want the heads to be evenly split around 0. no matter if 0 in in a corner or in the middle.

oliof

@T3P3Tony I thought about why my setup is working, and the missing piece of information is that the slicer setting for the copy/mirror mode is not using zero centered coordinates and simply splits Y length in half. So the X head needs to be offset from zero to the left hand side, and the U head runs unmodified, and by magic ... err ... two errors canceling each other out, things work.

If I were to use zero centered setup in the slicer for copy/mirror mode, your offset definition would be correct. I will adjust that for consistency's sake (and it should allow me to take prints to copy/mirror mode without needing an extra slicer profile).

T3P3Tony

@oliof thanks for confirming what was going on!