Correct me if I'm wrong but having the filament change sides, would show up as backlash during the retact, which ought to be irrelevant on a retract because you don't need accurate positioning and it cancels itself out. Technically you might be off by the measured backlash after the first retract, but it wouldn't accumulate.
Now if the shapes was changing during the print I could see it changing the nozzle pressure slightly, and likely at consistent points in a layer, but I would have thought the amount of change would not make a visible difference, but I'm speculating.

Hi Brad,
Great work. I like the compact design. It may well serve the needs of many better than a larger print area. I agree that sometimes it might be better to print many objects consecutively and only suffer one failure, than attempt to print a tray of objects on a larger printer and suffer a larger loss. It looks robust and the print quality and performance looks excellent. The 'cables' are hard to see in the video clips and I know anyone that has a mechanical curiosity will be drawn at first to understand how the motion is transmitted. There's a magic about it!

The Bowden approach has taken a back seat of late but it's nice to see it revisited and apparently very successfully.

Best of luck with your project.
NumberSix

---

[numbersixreprap.blogspot.com]

destroyer2012 Wrote:
-------------------------------------------------------
> So it's 3mm filament, right?
>
> I am somewhat surprised you think that having the
> filament on one side of the bowden tube matters so
> much because when you retract, doesn't the
> filament pull away from the tube wall anyway?

It is important that the filament stay pressed against one side of the bowden tube _during_compression_, as this means that the filament doesn't ever have a moment of slack as the bowden moves due to the movement of the extruder. That means that any given extrusion doesn't have inconsistent speeds or pressure drops in the extruder head. Any slack during retraction however can be taken up by larger retraction, although that obviously could suck in air if you went overboard.


> You know I'm starting to think that the reason you
> are doing so well is the J-head nozzle. The
> ultimaker nozzle doesn't use a PTFE tube liner, so
> the melt zone ends up being about 2 cm long or so,
> whereas with the J-head the melt zone is small. In
> retrospect, the fact that anyone can get
> retraction to work with ultimaker is amazing
> considering that nozzle. Maybe I'll switch to a
> PTFE-liner nozzle and report back, but considering
> my machine works ok now and "if it aint broke dont
> fix it" that might not happen for a while .

The J-head definitely doesn't hurt, although I still have my own concerns about resistor heating. Seems like there aren't any failures though, so what do I know It would be interesting to compare surface finishes vs. heated volumes of the various heads.

rantenki Wrote:
-------------------------------------------------------
> The J-head definitely doesn't hurt, although I
> still have my own concerns about resistor heating.
> Seems like there aren't any failures though, so
> what do I know

The resistors that are used are essentially nichrome-wound. As long as the resistor is sufficiently thermally coupled to the heater block, the resistor will remain within its maximum temperature rating.