I decided to make a jump from a ramps to a duet 0.8.5 with DC42s latest firmware for my self built mini kossel and it was definitely worth it. Everything is running great except for some extruder weirdness and the duet web control isn't reading part height or first layer height from the gcode. I have scoured the inter webs and couldn't find the answers to the following questions.
Questions:
1. Do you recommend relative extrusion or absolute extrusion? If I use relative what should be in my start gcode script and what should be in my config.g on the duet? I am using simplify3d for slicing and I have the following gcode settings turned on: use relative extrusion distances, allow zeroing of extrusion with G92 E0. Should that last option be turned off? FYI I used the sd card files made for the mini kossel by DC42, the only change I made to the config.g is a command to invert fans for pwm. My problem is the extrusion rate is a little low on the first layer but is too high on the following layers.

2. What should be in my start gcode to get the duet web control print interface to know the layer heights as well as the part height?

3. The extruder steps in config.g are by default set in this format ###:### (ex 645:645). Do I need to follow that format for settings steps (with the : symbol between)?

Thank you!

1a. Relative extrusion is recommended when using RepRapFirmware because it avoids introducing rounding errors. Make sure you have M83 in your slicer start gcode to select relative extrusion. Zeroing of extruders using G92 E0 is not needed when using relative extrusion, so you can turn it off.

1b. If extrusion is too little on the first layer but too much on the subsequent layers, that suggests you are starting with the nozzle a little too high above the bed. So check your Z=0 calibration. I presume you configure the slicer to generate a skirt, to get the filament flowing well by the start of the print.

2. The web interface should be able to pick up the layer height. If it doesn't, please provide a gcode file that demonstrates the problem.

3. The format M92 E640:650 would mean use 640 steps/mm for the first extruder and 650 for the second. If you just say M92 E640 then it will use 640 steps/mm for all extruders.

Edited 1 time(s). Last edit at 10/29/2015 02:07PM by dc42.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Thank you for the reply David. Im a fan of your work and appreciate what you're doing for the delta community!

1b. I was thinking this could be the problem too and probably is. What process do you use for setting the z-height before each print? What I did was adjust my z probe z offset in config.g until G32 w/ 6 factor calibration would consistently pass the paper test at 0,0,0. I am using a mechanical switch z probe with some custom modifications to increase repeatability (ground the Allen key tip to a point and I use a 3 mm Allen key as opposed to the skinnier mini kossel standard Allen key), but maybe the mechanical switch isn't repeatable enough to properly set z height each time. Depending on part size I will always generate 1 or more skirts to prime the nozzle.

2. I will post my gcode when I get home from work!

I calibrate the probe height as follows:

1. Use the paper test to establish where Z=0 is. I have found that the results given by the paper test vary somewhat depending on the bed surface. For example, if your bed surface has high friction (e.g. PVA glue) then the nozzle will start gripping the paper at a higher nozzle height than when the surface is plain glass. One check you can do is to remove the paper, then put a light behind the nozzle, and check that you can's see a chink of light between the nozzle and the bed.

2. Send G92 Z0 to define that point as Z=0; or you might want to use e.g. G92 Z0.1 if you establish that the paper test leaves the nozzle a little higher than Z=0.

3. Raise the head 5mm, then send G30 S-1. Then read off the Z height on the web interface, and that is the trigger height.

Bear in mind that the first time you probe after deployment may give a different result from subsequent probes. So it is a good idea do always to a dummy probe after deployment and discard the result.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].