Geeetech

Hi all,

I have 2 issues with my G2S when trying to print dual extrusion prints.

Firstly, the Z level of the extruder 1 pass is always set at 0.19. Nothing I do seems to change this and auto levelling seems to have no effect. Same goes for altering the Z offset. This means the first run is down too tight and doesn't print evenly (the bed is slightly out).
Secondly, once the extruder 1 pass is done, the switch to extruder 2 causes the z height to go up about 3mm, so extrusion happens in the air. Once that pass finishes, extruder 1 prints at the same incorrect height. LCD is displaying z=0.4, but it's way off of that in reality.

Printer works fine on single extrusions. I've already had to change the Marlin firmware to get rid of one dual extrusion code bug (i.e. on switch to T1, spider starts climbing up frame!), so I'm guessing this is another one.
I'm guessing that the initial z level issue is the root cause and the second part is a side-effect.

Any bright people got ideas on this? I'm stumped!

Hi there,

I am having the same problem when trying to use firmware to set the extruder offsets on my G2S, switching to T1 will adjust my Z height and I'll either be printing in the air or crashing into the bed.

You mentioned a bug in Marlin, what were the characteristics of that bug and what version did you adopt to fix it?

Many Thanks!

I've the same problem. Surely it is a bug of the firmware.
When gcode switches from T0 to T1, spider climbs up/down 10 mm about!

Help needed, please.

With further tests, I've seen that just changing extruder from T0 to T1, the spider begins to
change height. Does not respect the set Z, while the display shows the set in gcode, and the
print plane seems to be slanted, not parallel to the bed, with the risk of crash into the bed itself.

Does it exist a further/different firmware, which solve this issue?

I have redone my firmware with RC7 now, it's a bit better but dual extrusion still seems to send an unregistered climb of .2mm whenever switching extruders. Obviously, if you are printing layer "x" from two heads, they should both be set to the right z level. What happens is that t0 prints, then t1 prints, at right height. For the next layer, t1 is left active and prints at right z level. Then switching to t0, it should finish that layer at same height - but it climbs another .2mm before printing. The end result is a too tall print with vertical gaps.
What I did find interesting was that I had to reconfigure the RC 7 installation with new values in many places, copying my old settings didn't work. Haven't been back to double head testing, will post if I make any progress.

Mine makes a 10 mm climb and my gaps are "big holes in the air"
The result? A hole without donut!
I'm very disappointed with Geeetech for this issue.
After the first extruder switch, the print plane is shifted up until the end of the job.
Wait for firmware upgrade. Meanwhile are interesting your next progress

Thanks for the updates. I think the issue is really within the Marlin firmware and I believe it's related to using the combination of auto bed leveling and firmware based extruder offsets. From what I read on GitHub, the extruder offset setting are (were) stored in a two dimensional array (X, Y) and of course the bed leveling code uses three dimensional arrays (X, Y, Z). Some bugs referenced the fact that the leveling code would contain a random value for Z when applying the offsets. This might account for everything from slight Z movements to total carriage crashes when switching. I'm disappointed that it doesn't seemed to fixed in RC7 though. And hearing that reconfiguring the new version was a hassle, maybe it's time just to let my slicer apply the offsets in the resulting G-Code.

Additional informations ...
The problem occurs regardless whether or not I use ABL.
I also tried to apply gcode M420 (mesh smoothing) unsuccessfully.
In this video http://www.youtube.com/watch?v=yqBLLo4SWQs
published last June 11, 2016 the G2s seems to work properly (sob!).
Since that date how many firmware updates have taken place?
Should we have downgrade to Marlin older version?

I've spent another full day trying to correct this, can't get any consistent results from it. I thought I had a lead regarding the slant on extruder 2, apparently it's caused by firmware extruder offsets - if you zero them and use Cura/Slic3r soft offsets it should work. Having made that change, I still can't get the staring height right; it does seem that ABL is not helping. With or without ABL, the starting z position is generally too high once printing starts. I can get it perfectly positioned with gcode that places it at z0.2, once the firmware planner kicks it does all kinds of crap.
This printer is a farce - it should have been provided with working firmware, but what's provided is not fit for purpose.
I'm giving up on open platform deltas as a dual extrusion solution, I'm going to convert this one into a dual head Prusa type model, which should be a bit beefier than the Prisa i3 I already have.

Looks like @aleXall has fixed it:
http://www.geeetech.com/forum/viewtopic ... &start=120