Help with retraction

[e_hutch]

I've had a few successful prints, found .97 a good extrusion multiplier @ 235C for my E3Dv6 with my black ABS and the standard quality setting in matter control. I cut up a B-1 stl to make it printable and wanted to use fine settings as a test run, but I am having issues with retraction on travels... Getting more than a few clinging peices. Any suggestions for improved results? Thanks!

[Xenocrates]

Counterintuitively, Try slowing your retract down some, and see if it helps.

[e_hutch]

what would be a good starting point to reduce it to, 100 mm/s, or do you think I should reduce it more?

[Xenocrates]

around 50MM/s, maybe slower.

[Polygonhell]

There is little advantage over about 30mm/s. Really high speeds tend to result in lost steps either on the retract or the return, which leads to uneven flow for a short time after the retraction.
If you want to do the experiment, make up a model with 2 towers, can be cylinders or squares, doesn't really matter and print it with different retract settings.
I spent a day doing it once, on my E3D v5 there was little difference with anything over 1.5mm @ 20mm/s.
KissSlicer now supports assymetric retraction, and the prime should ideally be slower than the retraction.
E3D hotends really don't like a lot of retraction, with much more than 2mm they tend to consistently jam, it's probably a function of the short transition zone.

[e_hutch]

Hmm, 50mm/s gave me a wiry mess this morning, I'll have to try the towers idea.

[e_hutch]

looking around here I'm seeing that 300mm/s as the default travel speed for non print moves is too fast, so I halved it to 150 with 80 mm/s retraction speed for now and am printing this calibration pyramid https://www.youmagine.com/designs/hollo ... on-pyramid I'll see how that turns out and adjust from there.

[e_hutch]

Btw, the first one turned out ok after a bit of cleaning, but I will keep trying to make it better.

[e_hutch]

I printed this pyramid three times at different retraction settings before I realised that the travel distance wasn't generating retractions. What distance do y'all typically use for min travel for retract?

[e_hutch]

I was wrong again, the travel distance was not the only factor, the min extrusion prior to retraction was also too high. Right now I am trying it again at 25mm/s retraction with 3mm min travel and .1mm min extrusion requiring retraction.

[e_hutch]

There is little advantage over about 30mm/s. Really high speeds tend to result in lost steps either on the retract or the return, which leads to uneven flow for a short time after the retraction.
If you want to do the experiment, make up a model with 2 towers, can be cylinders or squares, doesn't really matter and print it with different retract settings.
I spent a day doing it once, on my E3D v5 there was little difference with anything over 1.5mm @ 20mm/s.
KissSlicer now supports assymetric retraction, and the prime should ideally be slower than the retraction.
E3D hotends really don't like a lot of retraction, with much more than 2mm they tend to consistently jam, it's probably a function of the short transition zone.

when you say 1.5mm @ 20 mm/s are you referring to the "length on move" setting in mattercontrol of 1.5 mm? the default is 6.8mm. I tried reducing this number to 2mm and it is not pretty. I have yet to find a sweet spot after half a dozen calibration pyramids. can you list the settings that typically work for you with ABS as a starting point?

[Polygonhell]

No I mean set the retraction to 1.5mm, the recommendation for the SeeMe hotend can be as much as 10mm, but E3D hotends don't like a lot of retration.

[e_hutch]

So my next test:
1.5mm retraction
at 25mm/s
with default 0.3 Z lift
3mm minimum travel requiring retraction
??mm minimum extrusion requiring retraction (I put .1 to ensure a retract on every move)
do you use "wipe before retract"?

[Polygonhell]

I use wipe, usually between 5 and 10mm, but I usually use KissSlicer, rather than S3D (though I do own it).

[tcat007]

I'm having pretty good luck with ABS with 3mm @ 180mm/s, .2 Z lift, 0 and 0 on next 2 items, retract checked, no wipe. I've read that the higher retraction rate tends to minimize strings. I tried a 40mm/s rate and got stringing between parts, 180 seemed to fix it. Using Hatchbox white if that matters. I'm only on my 3rd week, so it's all new to me. Only problem I'm having is OD of parts is about .24 mm large in X and Y, I "think" I'm calibrated.

[Polygonhell]

E3d's have a tendency to jam with much more than 2mm of retract.
The problem with very high retract rates is you tend to miss steps usually on the prime portion of the move.
There usually isn't a big win at much over 30mm/s unless you have a very long retract, the E3D in particular stops stringing pretty much as soon as you offset the hysteresis in the Bowden tube.
I strongly suggest people run test prints and tune in a deliberate manner.
Make a model with a couple of towers, doesn't need to be big, print your outlines out to in, so you can see the extrusion immediately after the restart, start at say 20mm/s, run with 0 retract as a control, then 1, then 1.5 etc. once you stop seeing improvements tweak the speeds by 5mm/s.
You can do binary searches between extremes if you don't know what numbers your aiming at, hotends vary a LOT on what's optimal, the secret (and I use the word loosely) is to be methodical and actually run the tests.

[derzaubererer]

E3d's have a tendency to jam with much more than 2mm of retract.
The problem with very high retract rates is you tend to miss steps usually on the prime portion of the move.
There usually isn't a big win at much over 30mm/s unless you have a very long retract, the E3D in particular stops stringing pretty much as soon as you offset the hysteresis in the Bowden tube.
I strongly suggest people run test prints and tune in a deliberate manner.
Make a model with a couple of towers, doesn't need to be big, print your outlines out to in, so you can see the extrusion immediately after the restart, start at say 20mm/s, run with 0 retract as a control, then 1, then 1.5 etc. once you stop seeing improvements tweak the speeds by 5mm/s.
You can do binary searches between extremes if you don't know what numbers your aiming at, hotends vary a LOT on what's optimal, the secret (and I use the word loosely) is to be methodical and actually run the tests.

would you mind sharing your e3d settings with me?

[DeltaCon]

would you mind sharing your e3d settings with me?

actually run the tests.