PCB Flexing with Nimble extruder

  • Hi there
    I've been trying to use a nimble extruder on the smart effector but noticed with the offset drive mechanism the smart effector flexes near the mag mounts, as a result I get a wavy finish at best. Is anyone else running a nimble on a smart effector or suffering problems with flexing?

  • Hi I run a nimble on a smarteffector and don't see any problems. How long is your drive cable? Does it go through any tight bends?

  • Hi, its the 95cm drive cable, machine is 1m tall. No tight bends.

    It takes very little to flex the effector, hold at the magballs and gently try and move to of nimble, i mean gently, i immediately see movement at the gap between the hotend and fan.

    Im sure if I print slowly I'd not see any issues, but a little quicker and the sway of the drive gives wavy lines on the walls. Quicker i print worse it gets.

  • Okay can you share a model youve seen this effect with? I'll try printing it too. What speed did you have to get to, to see the issue?

  • I have attached 2 pics, one shows a steady print speed and even waves, the other I experimented changing speeds.
    I don't remember print speed settings for these but typically i use a 40mm/s to start higher and i start to get issues on the corners. Thinking the flexing is contributing to all this maybe.


    cant work out image insert, does wierd things!

  • You might have a point with the second image that looks like the nozzle is deviating from the intended path by a small amount. The first image though is quite a regular pattern, it could be the stiffness in the drive cable. I've just tried moving my drive cable around and there is a perceptible movement in the effector, interestingly when I focus on the nozzle and do the same there is less movement. Let me try a couple of cubes at different speeds.

  • So I did a series of cubes 30x30x15, 1 perimeter, 2 bottom layers, no infill no top layers, 0.2mm layers, white ABS (always use grey for this type of thing, but I had white loaded).




    100mm/s Front left corner/seam

    100mm/s right side

    So not much happening until we get to 100mm/s (actual speed 120mm travelled in 1.2 seconds) and then the seam is failing, Im not using pressure advance although it would be useful as you can see the extruder fails to reduce pressure quick enough to prevent a nasty seam. Also I cannot explain why I get a surface a bit like yours on the right side of the 100mm/s cube only. I did get some warp as I left the side off my enclosure and its cold in here. I like how we see the deceleration lines/ringing approaching the corners.

    I am using these movement settings:

    M350 X16 Y16 Z16 E8 I1 ; Set 16x microstepping with interpolation
    M92 X200 Y200 Z200 ; Set axis steps/mm
    M92 E1200 ;extruder steps/mm nimble
    M906 X1100 Y1100 Z1100 E500 I60 ; Set motor currents (mA) and increase idle current to 60% (Nimble)
    M203 X30000 Y30000 Z30000 E3600 ; Maximum speeds (mm/min)
    M566 X1000 Y1000 Z1000 E40 ; Maximum instant speed changes mm/minute (Nimble)

  • Hi, I have re-assembled effector and hot-end assembly after filing down the worm gear in the Nimble which was slightly oversized.

    So this is embarassing,
    I am now:
    Unable to replicate the print finishes previously posted.
    Unable to print at different speeds based on specified speeds in Cura, the wall speeds never seem to change. Only way I can change wall speed is during print on lcd console. I have no way of knowing what the true speed really is.
    Tried slic3r, prints first layers at specified speed but still prints walls at low speed, even if i change every speed value to 100mm/s.
    I dont know what i'm doing wrong here but its off topic, but makes troubleshooting impossible.

    I have restrained wiring to nimble body, bound wiring to drive cable for about 250mm above Nimble and now have very good quality prints. Filing the worm gear I am sure has also helped tremendously.
    Now the swaying of the drive cable is minimal, print quality is as good as with Bowden if not better.

    I am still concerned that should the drive cable/wiring impose any horizontal force on the Nimble this will almost certainly affect the intended path of the nozzle. Only the slightest horizontal force at the top of the Nimble causes movement at nozzle in relation to cooling fan outlet, something is moving somewhere.

    I really appreciate your assistance DjDemonD, I am pleased you could replicate the issue, your tests are certainly more accurate than mine so far!

  • I'm also noticing a distinct loss of quality when I switched to the nimble. When I get home I'm going to take apart my entire effector and carriages, check everything for tightness and tilt and give it another shot. Could just be something came loose during the switch-over.

  • If it is indeed a flexing issue, I may design a modified mount that grips at more points, ie between the ball screws and the effector or something. I'm definitely getting a salmon skin effect that I didn't before though!

  • I found that I had a similar effect if the 3mm cap screw that holds the flex cable cap down to the Nimble was too tight. The Zesty guys have now updated their guidance and recommend 'finger tight plus 1/4 turn' so you may want to try that if you haven't already.


  • Sorry to hear you are having teething issues with your Nimble, we've had a small number of users experiencing similar issues dialling things in initially.
    You know that we provide live support via our chat, yes?

  • @clearlynotstefan:

    If it is indeed a flexing issue, I may design a modified mount that grips at more points, ie between the ball screws and the effector or something. I'm definitely getting a salmon skin effect that I didn't before though!

    You may be interested in a recent experiment I ran to test a similar concept. It failed but was interesting nonetheless! Design and description is at https://www.thingiverse.com/thing:2719039


  • @ssrpp

    I'm also having this issue but with less severity.
    Do you have an example print under light at an angle?

Log in to reply