- User Since
- Feb 4 2017, 5:27 PM (115 w, 4 d)
Tue, Apr 23
@DaniAO These things are already there, unless this is insufficient:
Step 4: https://ohai.lulzbot.com/project/x-end-left-assembly/quiver/
Step 3: https://ohai.lulzbot.com/project/x-end-right-assembly/quiver/
Step 24: https://ohai.lulzbot.com/project/final-assembly/quiver/
We have also had some success using a flat head screwdriver in the slot along side the nut while tightening, to keep it from spinning.
Using spaces for spacing after M117 messages rather than tabs alleviates this issue. I replaced the tabs following M117 messages in the linked gcode above and did not see any odd characters.
Mon, Apr 22
@marcio Yes, first thing in the morning.
@marcio So this could possibly be alleviated by using spaces following the message for M117 prior to the comments rather than tabs?
@Steven I tried with another USB stick, off the shelf in MER. Was preloaded for Mini, I added the test gcode linked above and repeated the test on one of the units. The same character is displayed with every message throughout the test.
@Steven Yessir, will do.
The unit will not be bootable, unless you have a quiver or quiver CB test stand to attach to.
Would it be worth a try running the CB Assembly test gcode on 184.108.40.206 on a completed printer over there?
If not I can work with inventory to get one of these units to you.
This has been present on 3 of 3 units I have observed through the testing process today. Assemblers are unsure if this has been present on all of them prior to today.
The next CB Assembly to test is displaying this character also, so it is not actually limited to one unit.
That strange character moves with the text displayed on the screen; if the text string is shorter it moves left, and if longer it moves to the right.
I have a CB Assembly that is displaying this strange character during the testing process.
The unit was running the CB assembly test gcode from quiver repository, on Marlin 220.127.116.11
CB Assemblies do not get serialized, if you would like this unit kept separate for further diagnosis, please let me know ASAP.
@DaniAO Looking at this a bit closer, it looks like that was put in packaging because one cannot reach both Idler Tube Clamp fasteners in order to remove the feed tubes if the Z axis is at home (shipping position).
Perhaps we'll need to supply packaging with a power cord at that station so that once the feed tubes are removed they can plug in, turn the unit on, and the turn it off once it reaches the top.
Alternatively, we could remove the feed tubes at calibration, have them placed on the bed with the cables/paperwork/test print, and have them home the z axis.
Which would you prefer?
My plan is to combine steps 4 & 8, 3 will move to calibration, and then clean up the rest and re-photo as described above.
@kent Would you like this OHAI updated to reflect the corrugate sleeve's seam positioned downwards as we had boxed the last unit?
@DaniAO Yes, correct quantity is 41 of HD-BT0148 for AS-PR0125
Fri, Apr 19
@DaniAO as far as I can tell this is complete for https://ohai.lulzbot.com/project/taz-6-extruder-full-assembly/hot-end-tool-head-assembly/
No luck on that machine, QB12
Going to place it back on the red rack for now, I cant get it to pass
working with @MichaelM to gather materials for this
putting this in general subs
So the test print I ran after changing E steps didn't come out any better on either the vernier or the circle. I'm thinking the extrusion on this machine may not be consistent. I am going to cold pull both sides and then reassemble the tool head to see if I can get it more consistent.
Thu, Apr 18
I have QB12 that was tagged for failing dimensional accuracy. While trying to alleviate that issue I noticed a good bit of under extrusion.
I was able to get repeatable extrusion calibration on this machine at 436 E steps/mm.
Filament was not through the filament sensor for this test and was not meeting extra resistance from the feed tubes.
I did confirm this setting to be correct for both nozzles on this machine.
At first it was under extruding by 4mm. I am going to load more filament and run another print to see if this makes the circle more accurate. It is circular visually, yet undersized.
Updated step 7 here: https://ohai.lulzbot.com/project/calibration-taz-pro/quiver/
And step 11 here: https://ohai.lulzbot.com/project/final-assembly/quiver/
Nobody ever confirmed if they were ready to publish, they are still residing in drafts
Added to step 8 here: https://ohai.lulzbot.com/project/electronics-case-mini2/mini-2/
Ok I will get that in, thank you!
With the thermal gap filler kits this isn't an issue and the assembly instructions can be easily followed in reverse for removal and replacement.
OHAI does not mention testing the box or what to look for when testing.
We don't test these AFAIK, @wolffchadd could you elaborate?
No explanation of how to install fuses and no mention of fuse's part number.
This is in step 5. I have clarified "Install two fuses into the power entry module." to "Install two fuses into the power entry module by opening the fuse tray and inserting 2 fuses [EL-MS0414]"
We are now achieving passing calibration prints, see @kent's comment on T5891
However, we should test implementation of backlash compensation in X/Y across more materials/part geometries to make a more seamless user experience.
Updated, see here:
This ticket is for OHAI creation
In relation to the Z axis print direction of the part, maybe. But as can be seen in @west's photo above, and in addition your comment:
The chamfer of the retaining ring should match the internal chamfer of the printed part that is receiving it
This means that retaining rings are facing differently depending on where they are on the part. Some are with the chamfer facing the bearing, others are with the chamfer facing out, and it isn't easy to tell unless you look at the model.
Tue, Apr 16
Thank you for the clarification, now that I have that clarification I can retake a few photos.
@west so, its not uniform where they are all facing the same direction in relation to the bearing? Why is that?
Would it be better (or acceptable) if they were all installed as described in step 3 here?: https://ohai.lulzbot.com/project/linear-bearing-install-guide/quiver/
This has been implemented step 1 here: https://ohai.lulzbot.com/project/linear-bearing-install-guide/quiver/
It looks like from T5047 that we want the chamfer facing in?
@kent Could you elaborate on the correct orientation for the retaining rings? Is it with the chamfer facing in, or out?
Implemented step 3 here: https://ohai.lulzbot.com/project/y-axis-assembly/quiver/
Implemented step 3 here: https://ohai.lulzbot.com/project/y-axis-assembly/quiver/
Can resolve after BOM change is merged in
@anolen I'm in MER all day with the exception of lunchtime. Feel free to call me at my extension also
Could probably include T5666 in that too
@kent So do we want to make another OHAI section for "Igus Bearing Install" specific to quiver?
I have taken photos of this to add to Y-axis OHAI and requested BOM correction on T6423
Looks good, I think we should implement this change and see how the situation improves through the different scenarios that may arise.
@DaniAO looks like we need to add 2x- HD-BT0148 to AS-PR0125. I'm not sure if they are elsewhere that they need to be removed from.
I have photos to add this to Y Axis OHAI and resolve T6530 at the same time
waiting on stickers to photograph
Updated step 4 here: https://ohai.lulzbot.com/project/calibration-taz-pro/quiver/
I have added this as step 24 here: https://ohai.lulzbot.com/project/final-assembly/quiver/
Left it in calibration also to verify there as well
@tutley Do you have a spool that is closer to the end that you could mount to see how well the new length functions with filament that has a tighter curl to it?
Clarified actuator function on step 4 here: https://ohai.lulzbot.com/project/control-box-testing-process/quiver/
Now reads as follows:
The actuator test located on bottom right of the test stand is used for visualizing proper function of the actuator circuits. Both indicator should rest within the green zones in either position. To cycle between positions, select MENU > Change Filament and use the 1 & 2 buttons at the top of the display. The left actuator should be at the top when number 1 is selected. When number 2 is selected, the left actuator should lower and the right raise to the top.
Updated steps 17 & 18 here: https://ohai.lulzbot.com/project/control-box-testing-process/quiver/
Added to step 4 here: https://ohai.lulzbot.com/project/control-box-testing-process/quiver/
See step 19 here: https://ohai.lulzbot.com/project/control-box-testing-process/quiver/
See step 20 here: https://ohai.lulzbot.com/project/control-box-testing-process/quiver/
Mon, Apr 15
Any determination made on this?
We need to get this into the OHAI
Also we think you may be having an issue with cura (@anolen and i both had this problem). You can select a profile on extruder 1 and click on the tab for extruder2 but even though the dropdown bar says its the same profile, it is not. if you click the down arrow you will see that they are not the same. I think you may have accidentally been selecting 2 different profiles because of this, but expecting they are on the same one.
maybe you were aware of this weird caveat but i figured it needed mention.
I was not aware of that, however when selecting the desired profile on extruder 2, it changes all of my extruder 1 settings anyway.
wall line count to 3 - requested to provide more surface area to measure on the 2 small squares
nozzle 1 zero top layers 4 bottom - this is so that there aren't extra unneeded skin layers in the middle of the tower and the squares remain solid
nozzle 2 zero bottom layers 4 top layers - same as above
infill to grid @ 20% - to make the infill in the tower similar to hibiscus
infill offset nozzle 1: x2.5 y0.5 - same as above
infill offset nozzle 2: x-2.35 y0.5 - same as above
z seam alignment to user specified - so that the end of vernier lines is away from eachother; the blobs where it finishes the line were said to make it more difficult to read
seam corner preference to none - so that user specified actually works without taking into account model geometry and placing the z seam on its own
nozzle one z seam x 0, z seam y 500 - see above
nozzle two z seam x 500, z seam y 0 - see above
hot end 1 for skirt - this may be default
skirt line count 2 - requested by calibration
skirt distance 50mm - same as above
nozzle 2 skirt min length 0 - so that cura doesn't print a skirt with nozzle 2 even though nozzle 1 is selected
max retract count increase both nozzles 99 - allows for retraction between vernier lines
min extrusion distance window 0.5mm both nozzles - same as above
They are completely different machines, different electronics, almost double the belt span (increases backlash), different carriage drags (more than hibiscus from what i can tell, and this increases backlash), different filament feed (more drag on toolhead), different toolhead (more opportunity for play in the system which adds to backlash), etc.
@tutley You worked out the math for the circle spec on hibiscus, given the same dimensions and the average backlash measured from quiver, what is the new spec? Does that provide an explanation for what we're seeing?
i dont think all of your changes are necessary.
All of the changes were based on feedback from production and @kent. It was all changes I was specifically instructed to implement.
why are you making 2 different profiles when we are printing with the same material. They should both be the same profile, and should be from @anolen 's standard print profile (what the customers will be using) or very close to it.
See changes listed above, I can go into greater detail on each of the changes I made if need be, but base profiles for both nozzles does not work for calibration prints.
Text is there, just have to get a hold of a unit to photograph
@west Please advise on a path toward resolving this issue
No updates since 3/7
@TyTh closing this one out, please reopen if this issue reoccurs.
Anyone have a path forward for this?
Leaving the motors off until after the frame is squared is not a viable option at this time as it would require major rework of the OHAI and BOM.
@anolen What are we trying here?
I don't have a Quiver to test with.
Looking at the photos, this gcode doesn't seem suitable for production calibration. We would need at least all of the modifications I listed above, in addition to whatever you're trying here.
Thu, Apr 11
How do you square a frame sitting on rubber feet?
Tried a few prints with some aborts on Mini2 and Mini and was not able to reproduce on 3.6.6 (local)
Will cherry-pick to master through T6321 after completion of review phase.
@karrad None that are linked between extruders when I was trying the above.
This is all based on feedback from multiple sources, and trying to get the circle tower portion as similar to hibiscus as possible.
wall line count to 3
nozzle 1 zero top layers 4 bottom
nozzle 2 zero bottom layers 4 top layers
infill to grid @ 20%
infill offset nozzle 1: x2.5 y0.5
infill offset nozzle 2: x-2.35 y0.5
z seam alignment to user specified
seam corner preference to none
nozzle one z seam x 0, z seam y 500
nozzle two z seam x 500, z seam y 0
hot end 1 for skirt
skirt line count 2
skirt distance 50mm
nozzle 2 skirt min length 0
max retract count increase both nozzles 99
min extrusion distance window 0.5mm both nozzles
Is it anywhere in retail parts repo?
This is fixed, see changes here: https://ohai.lulzbot.com/project/control-box-testing-process/quiver/
Cosmetics much improved due to less warp and a better outer wall. The SD card slot support comes out much cleaner as well.