We will need a bushing burn-in and octopus calibration gcode for quiver. Also a dimensional accuracy test. Probably the quick circle since that is what we use for mini 2 and we want to be able to compare these two printers. We will want to also test both extruders, so maybe we need to print two quick circles.
- Mentioned In
- T6451: vernier print reslice
T5891: Dimensional Accuracy - Calibration Print
rTAZQUIVER0867940fd51f: fixed burnin gcode to eliminate slight bump in bottom right T4129
rTAZQUIVER747a424d65ff: Updated vernier calibration print T4129
T5617: X/Y Dimensional accuracy is short roughly .5mm
rTAZQUIVERafbf287519a3: Merge branch 'master' into T4129
rTAZQUIVER41d3b7767f51: Added vernier calibration print gcode and model files to…
rTAZQUIVERb64dc07aba76: Added burnin.gcode to sample_prints/calibration T4129
T5070: Quiver First Print materials
- Mentioned Here
- T5033: tool head rotation about Y-Axis
I have been thinking of this... Would it be better to have a dual color quick circle to test the accuracy of using the 2 hot ends at the same time? (The quick circle would still be the same dimensions as the mini)
Throwing my two cents in here, I don't think multiple would be required. Extrusion will need to be verified off of each nozzle, but the dimensional accuracy will only need 1 print since the entire tool-head is moving, there shouldn't be a difference in shape between the two prints. One print with a few layers of the other tool head in the middle will let us know if the offsets are in place and give general information on extrusion. This is what I believe @oliver is getting at with the original comment. Printing multiple is redundant.
I agree with @EricNugent I think we can come up with a dual color, one piece print that would show us what we want to know. I also agree with @kent that we want the closest comparison possible.
That having been said, one thing to keep in mind while creating calibration gcodes is time. There is a limited amount of space on the burn-in racks and naturally we want to keep calibration moving as quickly as possible, so we need to keep calibration gcodes to the minimum necessary. If it doesn't provide us with a useful metric it should be eliminated.
What do we benefit by tracking results of dimensional accuracy tests? Performing the test is great to make sure our printers are good when they leave, I just am not understanding how tracking that data is useful. As far as I can tell, all it tells us is how many of our calibration prints fail.
Here is a quicker vernier scale for dual calibration:
The quick circle is split into 2 colors which gives a bit of visualization for Z nozzle offset since the vernier wasn't working as well in that scenario.
Further profile mods: (ABS)
Fan: 12 reg 15 max
Connect infill lines: No
Infill offsets X/Y set to align center of infill in circle tower
extruder 1 no top layers
extruder 2 no bottom layers
Here is a PLA gcode for vernier with 2 small squares to assist with Z nozzle offset. I was able to measure the two small squares and apply the difference between the two to correct z nozzle offset.
Including cool down time this print takes 25 minutes.
I have made a burn in gcode
Cycles X/Y/Z, fan 100% and tool switch half way through, takes 18 minutes to complete.
Let me know what y'all think, if there's anything missing or could be improved upon.
Just be aware that the circle on the vernier print will fail due to T5033
This is merged in now.
$ git diff master --summary create mode 100644 sample_prints/calibration/burnin.gcode create mode 100644 sample_prints/calibration/quiver_vernier_PLA.gcode create mode 100644 sample_prints/calibration/vernier_noz1.stl create mode 100644 sample_prints/calibration/vernier_noz2.stl create mode 100644 sample_prints/calibration/vernier_v4.fcstd
I just noticed that the last vernier gcode I pushed didn't have the 2 squares set up right; selection was wrong when exported.
I'm working on drafting instructions for calibration OHAI now; will get a new one pushed once verified.
Dave Robertson is working on the first prints that the customer will do, but we also need to have a print that calibration. @logan has designed a print that MER has been using to test the calibration with. Logan is leaning toward the LulzBot Green base with black lettering as it's easier to ready. Are we okay with sending that to the customer, rather then the octopus?
The above PLA gcode results in good circles, but makes the vernier unreadable. I'm not sure what the settings were that resulted in that.
Here's what the gcode looks like when imported into CuraLE:
There are both burn-in and calibration print g-codes in the repo now, so the original purpose of this ticket is complete. If there are issues with either the burn-in gcode or the calibration print gcode, they can go into a new ticket.