Another Duet delta with 4th column



  • My machine is largely based on David Crockers blog. I have had significant problems with blobbing and stringing which I have been unable to reduce to an acceptable level. Hence the decision to reduce the length of the Bowden by implementing the 4th column.

    I have adopted a slightly different approach using a counter balanced linear actuator mounted in the centre of the machine with the extruder and motor attached to the bottom of the actuator.
    The arrangement is working well and I have all but eliminated the stringing and blobbing thus far on quite challenging prints.

    The downside: when the actuator is at its home position it projects well above the top of the machine which is already 1m high.

    Next iteration will be to replace the actuator with a telescopic arrangement again counterbalanced and mounted in the centre but driven with a stepper and capstan wound with 3mm dyneema cord. The stepper will pull the column up but will not drive it down. When stepping down the weight of the counterbalance will lower the column but will be controlled by the stepper. This will significantly reduce projection of 4th column above machine.



  • For anyone interested I have added a link showing the extruder in operation. I had read that a long bowden compromised performance of printers but am amazed at how much better the printing is with this arrangement.
    https://youtu.be/Y3GZLjHfz48



  • @PeterW Thats looking great, thanks for sharing

    @PeterW said in Another Duet delta with 4th column:

    The downside: when the actuator is at its home position it projects well above the top of the machine which is already 1m high.

    This is what kept me away from going the 'single rod in the center' route.

    @PeterW said in Another Duet delta with 4th column:

    Next iteration will be to replace the actuator with a telescopic arrangement again counterbalanced and mounted in the centre but driven with a stepper and capstan wound with 3mm dyneema cord. The stepper will pull the column up but will not drive it down. When stepping down the weight of the counterbalance will lower the column but will be controlled by the stepper. This will significantly reduce projection of 4th column above machine.

    A nice addition to this would be to mount the top connection point of the telescopic arm via ball and cup style joint. I have implemented this (no telescopic action though) in my delta and it allows for a considerably shorter bowden tube if done right: as can be seen. In its current state, the bowden tube is 123 mm long (including the part inside the hotend and extruder) for a 400 mm diameter delta.

    EDIT: video of what I mean: Youtube



  • Looks very good and I'll need to thing about the ball joint connection . My extruder is gimbal mounted but I think the bowden I'm using is currently too long to allow it to move freely. Can' t adjust the length until the print is complete!



  • @PeterW If you scroll back in the thread I linked (or the mega build log in my signature) you can see examples of my old 4th axis versions. Previously, I had my extruder gimballed as well, and that is certainly an improvement. It worked quite well for a good time.

    Though the shortest possible bowden with that style of setup is 1/2 the bed diameter (assuming a bed radius of 0 etc so in practice it must be longer but just for easy comparison). If your entire arm can gimbal, the shorted bowden you need to bridge the gap is 0, you only need enough length to accommodate for the bend radius of the tube and tolerances.



  • Thanks for that. Impressive build log.


Log in to reply