Nice

With 4 screws I'd get an duex 5 and have full automatic bed levelling.

I don't see any guides for the bed travelling along Z. No linear rails or rods. What's keeping the bed wobbling with the screws? Or do you have perfect screws?

Just about everything roughly dialed in. Just realized I forgot the ferrules on VIN and bed out.
I wired the case fan to fan 3 and is controlled by drivers and MCU temp. The noise is just a fraction now.
The PSU fan is wired to E1 heater output. I have to order a couple of thermistors or PT1000s I can glue to the PSU to control its fan.

Set up my old Titan I had laying around. Just have to remember to reverse E0 as I just did an basic E-steps calibration in reverse. I think the filament sensor should work ok with its original bracket until I can print something better.

Still have the improvised magnetic mount. Need to find something permanent.

A PanelDue would do nicely in that hole.

Now the fiddly bits start. Calibrate and fine tune everything.

Just want to share a few changes I made in the python script.

• Prints a box instead of two walls
• Tidied up the skirt generation as id did some strange moves
• has a text block for starting gcode and end gcode

advance_cal_box.py.txt

@engikeneer

Thanks! That sounds sensible.

@o_lampe

Thanks. Since you often need less retraction when using PA, I thought to do an initial PA calibration without any retraction. Then calibrate retraction and eventually a new PA calibration to see if it changes in any way.

I usually have retraction set between 0.2 - 0.4, so it's not a lot anyways.

In reality I have a constant speed of 40mm/s in all moves in the test which coincidently gives me a constant flow with only a single wall. At least according to speed and volumetric view in Prusaslicer.

Hi,

I could use some suggestions on where to go from here.

Decided to try calibrate a new 0.8 Revo nozzle with PETG. I have been printing a 100x100x100mm cube with a single wall and no infill or top layers. I set the seam in the middle of the wall along the x-axis (towards the front of the printer).

PA = 0, I get a bulge in the seam. The corners look pretty ok, maybe a slight bulge.
PA = 0.06 the seam is the prettiest, but now I got strangely rounded corners. Like they are being cut of too early.

In between 0 and 0.06 I have to choose between rounded corners or bulge in seam. The inside corners look ok.

I did two tests with different acceleration and jerk increasing. Tested acceleration between 100 and 1300 and jerk between 150 and 2000 but nothing changed much, just slightly. 40mm/s is probably to slow to show much of a difference.

Could it be low quality PETG?
I'm levitating towards having crisp corners and having to live with a bulging seam.

• Ender 5 Plus
• Hemera direct drive with Revo 0.8mm nozzle
• Volumetric flow for is constant ~9mm3/s for whole print
• Speed 40mm/s
• 0.3mm layer height
• 0.87 mm width
• All retraction has been disabled for tests

M203 X15000 Y15000 Z1200 E3600              ; Speed - set maximum speeds (mm/min)
M201 X500.00 Y500.00 Z100.00 E1000.00       ; Accelerations (mm/s^2)
M566 X900 Y900 Z60 E3000                    ; Jerk -  gammel. brukt en stund nå

Any suggestions?

Anyone still has this issue? Was there a resolution? I might have something similar but I'm still testing.

How about you @jatmn?

Thank you very much @mwolter. I have a lot better prints now.

These prints where printed with the new cooling fan shroud. The temp tower in the photos are all printed at 235C but the fan increases 10% for every block. Bottom level is no fan and the upper is 100%. 10% - 20% seems a good value.

Seems like slight under extrusion but overall this part is almost perfect. Printed at 235C and 30% fan

I hope to be able to print a fan shroud now to enable me to print a better fan shroud

Made some technical parts and some are very good but some are bad. So next issue is cooling.

I did not believe ABS wanted cooling? Is there a way to only start fan when there are overhangs in S3D? There is a bridging option to use fan.

@mwolter

I made two more runs with single wall boxes and expecting 0.5mm wall I get an average of 0.495. Almost 50/50 measurements of 0.49 and 0.5mm. The last was at 220degC and it's too easy to delaminate the walls but it looks pretty perfect!

I made an temp-tower to show the effects of changing temp. It's surprisingly forgiving in a large range. Or maybe it sucks at just about any temp Need to buy another brand to test.

In my opinion it's the best temp-tower I have made so far. Bridging and steep overhangs are not very good. I have no enclosure but a very small room that gets about 40degC. I have no functioning cooling fan at the moment as well.

My goal is to make solid parts with good geometry for the printer it self. They do not have to look perfect but of course I want them too look as good as possible with the limitations. I want to use ABS to better withstand an future enclosure.

Trying to break the tower 215C broke very easily. Then 230C snapped with a bit more resistance but still easily delaminated. 240 required more force and 245 was much more solid. Not a very scientific test but confirms earlier observations that below 230 the layers do not bond properly. To avoid delamination for thin high parts I need to keep it above 230, preferably 240-245. I am not able to see much improvement below 240 anyways.

Lowest block is printed at 250 and lowering 5C every block above.

Looks a lot better but the Y-side overhang looks a lot worse than the X. Z top surface looks pretty good. Again, it looks a lot better in real life as I try to show off any imperfections with the lighting.

Changes:

• Extrusion 87%
• Temp to 220degC
• 0.1mm coasting

To check my extrusion I printed two 30x30x30 boxes. One with 4 layer outline and one with a single layer outline. Both with 0.5mm width.

I sampled the walls 4 times on each side. The single wall is slightly wider but the 4 walls is barely measurable.

Any suggestions which measurement I should weight the most?

I think I have put too much pressure on the calliper on earlier tests. I get slightly longer measurements when I only press the part of the calliper where it actually measures. Maybe spend a bit more on a sturdier one.

Strange stuff.

I have tried lots of combinations of high and very low jerk and acceleration. I have tested most combinations with PA from 0 to 1. I have tested Prusaslicer in addition to Simplify3D and I still get the strange line/shadow before a corner.

I went through the mechanics and re-"trammed" the printer. Adjusted all the v-slot wheels and retightening belts. Still no help.

So I'm beginning to think it's the filament that just get pulled like that in a corner. I would believe that this could be improved by temperature and speed but no. Tested the ABS between 215 - 250. Below 230 I get worse layer adhesion so I tend to print 230-240.

So I ditched the single wall tests. I have printed some XYZ-cubes in default size 20x20x20mm that actually doesn't show the line/shadow very much. Looks like some over extrusion but else not bad.

0.4mm nozzle
Layer height: 0.28
width 0.5
extrusion Multiplier 0.9
Retraction 0.35
topSolidLayers 3
perimeterOutlines 4
infillPercentage 90
temperature 240
bed 110
fanSpeeds 0

Pressure advance 0.01, 0.025, 0.035 and 0.05

Please disregard the bottom layers. Seems I need to recalibrate the z probe offset