E£D Cyclops No Stringing fix/tips

Hey Everyone,

Just installed the e3d cyclops on my custom built kossel and after battling stringing I think I finally sorted it. Thought I'd post this for anyone that is still battling oozing and stringing with this hotend.

First some specs:

Custom Kossel XL (250x300mm build area)
E3D Cyclops hotend
2off Bulldog XL extruders
approx 400mm bowden tubes
Slicer: Simplify3d 3.0.2
Controller: Duet 0.8.5 running latest dc42 delta fork

What made a huge difference for me was drastically increasing travel speeds to 30000 mm/min.

Some other settings in Simplify that also helped:

Retraction: 2.5mm
Vertical Lift: 0.1mm
Coasting distance: 1.2mm
Wipe distance: 5mm

Notes
This worked properly only after I ensured my extruder are properly calibrated
I experimented with retractions up to 3.6mm without any clogging but reverted to 2mm after I upped my travel speeds.


Hope this helps anyone and please let me know if you guys have any other suggestions to remove stringing/oozing on the Cyclops hotend.

Thanks for the useful info!

If coast to end helped, then using extruder pressure advance instead may help even more. It actually starts retracting filament towards the end of moves with zero extruder end speed. You need to be running version 1.09o or later of my fork for it to work properly. Enable it by sending M572 D0 S0.1 and M572 D1 S0.1. The D parameter is the extruder number and the S parameter is the pressure advance constant in seconds. If you use larger values of S then you may need to decrease your retraction distance in S3D.

---

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].

David, I wonder how a time fixed S-parameter would work with different speeds for infill ( say 150mm/s ) and ext. perimeter ( say 50mm/s )?
Wouldn't it be better to start the pressure advance at a certain number of steps? ( say 500 steps before speed goes zero )
Just my two cent, sorry for OT
-Olaf

It's not a fixed time parameter. It's (extrusion advance distance) / (extrusion speed). The units happen to come out as seconds.

---

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].

Hey David,

Sorry for late reply. Cheers, the bowden elasticity setting made a big difference! Have now removed coasting from my simplify3d settings.

had a question for you, any chance of implementing 626Pilot's depth mapping algorithm in duet in the near future ? think this would make a big difference in getting that first layer perfect across the whole bed surface.

A properly built delta is probably still the better option but still pretty difficult to eliminate all the variables, especially effector tilt (still fighting this) meaning printing with a dual extruder setup on a delta is still pretty difficult.

Regards,

Mihai.

I'm glad the pressure advance worked for you. What value did you use in the end, 0.1 second or something else?

Adding bed plane compensation to RRF on a delta printer is quite low on my list of priorities, because I have more important features to add, and IMO it's not too difficult to get good calibration. The effect of effector tilt on the height sensing can be compensated for using the H parameter on the G30 commands in bed.g. It's tedious measuring the probe trigger height at each probe point, but it works. Bed plane compensation wouldn't help with that. My 300mm printable diameter delta calibrates to 0.032mm RMS error, which is flat enough for the 0.2mm first layer height that I use.

As for dual extrusion, my experience is that even when both nozzles are at exactly the same height, the non-printing nozzle interferes with the printing nozzle. So either a deliberate tilt, or auto-lift nozzles, or something similar is required.

---

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].

Hi David,

I ended up using 0.14, varied the pressure advance while doing a print to see how it affects things.

Re dual extruders, I agree, something similar to the spiderbot dual extruder setup is probably the way to go.

Thanks again for your help!

Kind Regards,

Mihai.

Hi Mihai,

I´m experimenting with M572 too.
Do you enabled coast in Simplify3D as well? What about Wipe?

Can you please post your settings in S3d? I know that I can´t copy it to my settings because of different filament, speeds etc.

But what is your best experience so far?

regards
Dirk

hey Dirk,

Nope, after I started using M572, i have disabled coast. I'm still keeping wipe on @ 5mm with retract during wipe turned on.

Also, I have been using 3.6mm retractions @ 5000mm/min without any issues. I know E3D recomment between 1.5 -2mm but did not see any issues. Looking at the technical drawings for the cyclops hotend block, you have 3.3mm between the centerline of the horizontal melting chamber channel and another 4mm where the heatbreaks thread.

http://wiki.e3d-online.com/wiki/File:CY-MIX-175.jpg for reference.

So far it's been printing pretty good, I have a slight issue with my layer width (top layers look under extruded even though I have calibrated my extruder and corrected extrusion multiplier) - Top layers have interval gaps. Here;s a test print:

[goo.gl]

Print took around 4hrs testing the 3.6mm retraction setting to death without an issue.

Hope this helps

Quote
thundead
So far it's been printing pretty good, I have a slight issue with my layer width (top layers look under extruded even though I have calibrated my extruder and corrected extrusion multiplier) - Top layers have interval gaps.

I find that different filaments need different extrusion multipliers, no matter how carefully you have calibrated the extruder, and even if the filament diameter is spot on. So I leave the extruder calibration and extrusion multiplier in the slicer alone, and adjust the extrusion factor in the web interface or on the PanelDue until solid infill looks right. Then I make a note of what extrusion factor to use for that roll of filament.

---

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].

Yes, unfortunately each filament roll will require different settings. That's pretty much my workflow as well. Got a pre-set extrusion multiplier in simplify, then I adjust it via the web interface to get a good 1st layer and finally, monitor the second solid layer just to confirm settings (as, sometimes I make the 1st layer smaller to get better bed adhesion)

Regards,

Mihai.

Reason why I´m asking is:

Very well calibrated Kossel (complete Bed in 0.02mm, 0,3mm Nozzle, 0,15mm Layerhight, M572 D0 S0.14)
using S3D:

1. First Layer perimeters are printed very good (0,15mm Layerhight). But when Solid infill of the first Layer begins, it starts the same extrusion like the perimeters. But the more its comming to the end the more filament is extruded. Can fight it with M572 little bit, but I think thats not normal. Or is the extrusion rate for first layers perimeter different to the infill extrusion rate set by S3D internally? Did you experienced the same? S3D is not answering my questions.

2. Second Layer and on are spot on. Perimeters looking same as the infill and printing very good.

3. since I use M572 I can hear the extruder sometimes retracting while printing infill. That happens near to the perimeters and when printing honeycomb infill its all the time. Normal?

I´m quite happy with my prints and the new added M572. But sometimes I do not understand the behavior of S3D!

thanx
Dirk

Quote
critical_limit
3. since I use M572 I can hear the extruder sometimes retracting while printing infill. That happens near to the perimeters and when printing honeycomb infill its all the time. Normal?

With large values of the M572 S parameter, it's normal to see a small amount of retraction towards the end of any extruding move that is followed by either a travel move or a sharp change in direction. It's not normal to see a large retraction other than before a travel move. However, I too heard something that sounded like a large retraction a few times on a recent print when I wasn't expecting it. So there could be a bug either in the firmware or in S3D. If you think you have a print that exhibits this behaviour, please provide the files.

---

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].

Yes, i'm pretty sure 1st layer extrusion is different if you set your 1st layer height greater than 100%.

I've also noticed that simplify does not do any retract whatsoever when it's doing the 1st layer infill which tends to increase the pressure in the bowden tube significantly by the end of the layer.

Thnx David.

I´m near to the printer the last hour and observed following.

M572 is working like you said. Either before a travel move or sharp corners.
But I observed also when printing the honeycomb infill it is not retracting when it is printed in 0° direction (means, no retracts when printing in the X-Direction). But retracts when it is printing the 120° direction which is set in S3D at the honeycomb settings.

So, why is it retracting when its not going in X or Y direction, but it is retracting when there is a angle to X or Y? All Angles in the honeycomb pattern are the same. Seems to me that your calculation of a "sharp angle" is related to X or Y direction of the bed but not to the last printed segment. Is that possible?

Not a big deal, I love the M572. But understanding how it works would be the best option. What is the definition of a "sharp corner"?