Hi, well that was a quick reply! Thank you.
Justine, you're right, the pick and place automation is the hardest part, but it has been done before.
https://www.youtube.com/watch?v=viX_WLRqxYMThank you.
Greetings,
Alex
Nice being mentioned!
Curious to see your results.
Docking positions are fixed, the upper part of all tools features the same docking geometry. But the position of the tool tip may vary from tool to tool. The offsets are defined with G10. However the docking and undocking procedure is completely independent from theses offsets.
Regarding the tool change macros, only tpreX.g runs in machine coordinate system.
I'm starting to run 2 FFF toolheads and a paste toolhead with a tool changer.
Every time I change something on the paste thingy which causes a different tool offset I have to take care of readjusting tfree / tpost as well. Otherwise the printhead crashes into the dock during docking / undocking. OMG I watched this quite a few times now 
Would it be possible to have gcode to switch to machine coordinate system (this seems to be somehow related to the CNC coordinate topic)?
This one http://foehnsturm.com/3dp/P1070095.m4v was actually working reliable but threw the stuff all over the place
Hi Ian,
Nice!
What is your experience with reliability. Did a lot of different wiping setups for the magnetic tool changer and most times the single blade would fail like 1 out of 50 times just wiping the filament strand to the other side of the nozzle. Especially if it tends to curl upwards for some reason.
A little more complicated: multiple toolheads but only one stepper.
Great that I could help.
@dc42, could there be an issue with always completely executing the macro?
@topiolm@hotmail.com, no yet using a slicer Just finished the park and pickup sequences.
This is probably not closely related but I'm running 1.19.2 (2017-09-01) and noticed that the final line of my tfree macros is not reliably executed (when using the select tool command in the WebUI). It sometimes finishes the macro without performing the last move but will perfectly execute any following command (which ofc could cause a crash into the dock). I ended up with duplicating the final line to safeguard its execution.
Magnetic tool changer on a corexy.
[[language]]
; tfree0.g
; called when tool 0 is freed
;
M83
G0 E-5 F1200
M82
G0 X252 Y215 F16000
G91
G0 Y47
G0 X30 F4000
G0 X4 F2000
G90
G0 Y215 F6000
G0 Y215 F6000
But isn't this only for Z?
I'd like to probe for the X and Y offsets as well.
If I'm not totally wrong, it automatically detects the offsets of a toolhead (e.g. nozzle) by moving it while checking a sensor.
Would be great to have.
Wiki mentions "in development, probably coming in RRF 1.20"
For a different project with my corexy.
Is it possible to map n tools to n pairs of temp sensor and heater in/outputs (as usual) but just 1 stepper driver. Like using M584 3:E0 3:E1 ?
If so, all I need is an idea to remember the e axis values of each tool in order to correctly position the stepper before activating it again.
Hi guys,
Thanks for the fast reponse.
Videos about the Ultimaker toolchanger project can be found here: https://www.youtube.com/playlist?list=PLeNzDYMnqAqIeZHWU9iOWnj_2OaemupQG
Most likely this has already been explained here and I'm just not smart enough to find it …
Anyway, I created the Mark2 magnetic tool changer for the Ultimaker2. Now it's time to add a third head. For doing this we might replace Ultimaker electronics and our customized firmware by Duet3D. Like before offsets and tool changes should be handled within the firmware. We actually attach and detach head 2 or 3, with a z offset, while the standard head always remains in place. Extra heads are parked in two corners.
So, what coordinate space I'm in when e.g. switching from T1 to T0? In other words, do I have to adjust the docking position within tfree1.g using the T1 offsets?