Reversal Zits: Temperature Variations

Just for completeness, it turns out that extrusion temperature doesn’t have any effect on Reversal zits. A while back I dropped the standard temperature from 210 °C to 190 °C in one fell swoop and it didn’t change anything worth mentioning, let alone the zittage.

The white one was hotter, the orange one is cooler:

Octopodes - Temperature variation
Octopodes - Temperature variation

The zits are pretty much due to Reversal followed by in-plane motion, it seems to me.

A dramatically lower extrusion temperature works fine for smaller objects, but I’d expect very large objects to delaminate like crazy. The Barbie Pistol was, IIRC, printed at 220 °C, and it had some troubles.

It’s also worth noting that the indicated temperature has only a casual relationship to the actual extrusion temperature. I’ve put considerable effort into electrically insulating and thermally bonding the thermocouple to the Thermal Core, so I think it’s a good indicator, but your results will certainly differ.

8 thoughts on “Reversal Zits: Temperature Variations

  1. Ed,

    Just a thought. Dampen your X-Y. A little STP (? ;-) )or something to slow the yank/pull bang-bang of the axes direction changes.

    Just a thought. I was thinking. Your whizzing unconstrained axes have momentum/inertia which is being fully controlled by the belts which do have a little rubber band stretch. So, whiz to an extreme pulling plastic then a snap back. Ooops, a zit. Then off to another extreme or direction change point then rubber belt snap back – zit.

    Again – Just a thought.

    1. whiz to an extreme pulling plastic then a snap back

      Yes, but … if that were so, then the thing would make a whole lot more of a mess going around sharp corners at full speed, which doesn’t seem to be the case. For example, those helmet mirror mount parts have the usual zittage, but the corners look just fine. Even the acute angles (in the slot of the middle piece) don’t glob out.

      However, the extruder doesn’t stop while the nozzle goes around the corner, so the zits arise only at the combination of end-of-thread and extruder-stopping just before a move elsewhere. The nozzle stops moving at the end of the thread, during the Reversal action, which pretty much eliminates the following high-speed move from consideration: it’s really at rest before the Reversal starts.

      And, really, those timing belts don’t have that much stretch in a 3D printer; the transmitted power rounds off to zero in terms of their usual industrial applications. I’d be more inclined to believe the (as yet) unsupported Y axis pulley shaft would be tilting, but that doesn’t seem to be a problem. In the X direction, the setup is about as rigid as one could expect from bolt-in-plywood tolerances.

      FWIW, my bottle of STP may turn into one of those cherished family heirlooms handed down through the generations. I decanted some into a squeeze bottle and when I flip it the STP in the bottom of the bottle doesn’t even notice it’s upside-down before I poot out a dollop: that stuff is gooey.

  2. “Yes, but … if that were so, then the thing would make a whole lot more of a mess going around sharp corners at full speed…”

    Under power it is controlled. So, going around is directed movement. I was referring to the throw out – yank back cycles. This seems to be where your zits are. The drastic change of direction where you have the short cycle lag/stop of the head before the change of direction and speed. Just as you state here:

    “However, the extruder doesn’t stop while the nozzle goes around the corner, so the zits arise only at the combination of end-of-thread and extruder-stopping just before a move elsewhere. ”

    Then as your observation is: “The nozzle stops moving at the end of the thread, during the Reversal action, ” Here you can imagine the axes going a little bit farther with the stretch in belt from momentum. The thread gets stretched a little bit more and then you come back in the opposite direction with a little bunch of thread (zit). Like pulling stained window or a piece of taffy. Or even take some gum. Pull. It stretches then try to make it and 1″. But, make it an inch and a little more and you have more gum to some how compress back to make an inch.

    As for stretch in the belts. Well here we have engineering/mechanical tolerances of materials giving a little to be more than we think. The would gives a little. The belts give a little. The motor mount gives a little (cork and mounting materials – acrylic or wood) Then we have the entire frame distortion of the Y and X assemblies with the give in their joints and connectors.

    Ed – I am not – always – wrong.

    STP. —- Story! 40 years ago – My Dad used STP in transmissions and engines. Each year he would have all fluids changed in our Truck before our two week multi-thousand mile camper vacations. Well – long story short. One year the mechanics drained the tranny oil but, did not put any back in. 1,600 miles later at 10pm we pull into an Uncle’s place, sort of. We just got off the highway and the tranny went out. Next day. They open the tranny and find a carrier rod bearing worn down drastically in the middle. The gear slipped out of place. There was a little pile of metal shavings. The bearing rod was shot with only a little bluing. All gears – OK. All else mic’d OK. Just had to replace the one gear and rod. And my Dad, a machinist by trade, said they only changed the gear because it seemed like the right thing to do. It had shown no wear and mic’d perfect. All thanks to our Father in Heaven and STP. Got a couple of other – Thank our Father in Heaven – and STP stories but, then I don’t need to bore you more. Cheers.

    1. Under power it is controlled.

      The steppers are always powered. The drivers never turn off (except for Z?) while building; you can’t push the stage around with the motors on. So being stopped doesn’t allow any more stage motion than while moving: if there’s slop, then there’s always slop. Right?

      you can imagine the axes going a little bit farther with the stretch in belt from momentum.

      If that were true, then it would happen on all the corners aligned with the axes: the motor slams to a stop along an axis and the axis continues. If there were that much slop, you’d see distinct ridges along the corner where the stage overtravels the endpoint and then recoils while the extruder continues to run and the nozzle moves perpendicular to the recoil.

      I have documented oscillations after high-speed direction changes, so it does happen… but not to the degree required to produce a zit.

      In other words, if stopping either axis doesn’t produce a visible defect, then stopping both of them won’t, either.

      Take a close look at those outie zits. They’re basically featureless circular blobs, which wouldn’t happen with the nozzle wobbulating all over the place. They’re all in the proper place for the end of the thread, right up against the starting point (modulo a bit of Clip gap).

      Indeed, I just measured the OD of each blob: almost exactly 1 mm, equal to the nozzle OD. I think what we’re seeing is the nozzle parked at the end of the thread with a puddle of goo filling the space below it while the extruder inhales enough filament to relieve the pressure. When the nozzle moves away, the puddle remains.

      I am not – always – wrong.

      Aye, but the right story will explain all the evidence and contradict none: that’s the tricky part. Nobody’s gotten it right yet, as nearly as I can tell from the pix I’ve seen here & there.

      More science!

      1. Science, you say? As we have no more hypotheses to offer, perhaps we need more information. Time to bring the high-speed camera up and film this as it occurs?

        1. As it turns out, my new Canon SX320HS pocket camera has a 240 frame/s slow-motion movie mode that might do the trick.

          However, the images are a rather gritty 320×240 pixels, which means I must gimmick up a holder to snuggle the camera’s snout right up against the nozzle in the Hot Zone. I think macro mode will work in slow-mo mode, but it’ll take some puzzling through the manual to get everything set up.

          That’s on the back burner right now, as I must get a Circuit Cellar column done before I get the 3D printing presentation done before we haul my Shop Assistant off to college at the end of the month. Kvetch, kvetch!

Comments are closed.