Dimensions problem

[DeltaCon]

Hi,

I received my FSR's this week and am trying to print mhackney's mounts for them. But U cannot (by far) fit the plunger into the mount. I did some measuring and found that the inner diameter of the mount is too small (17,56mm instead of the 18,59 it measures in the STL), The outer diameter of this circle is too big (22,49mm instead of 22,28 in the STL). Also the total width of the mount is too big (40,62mm instead of the 40,26mm in the STL). I am confused as to how this can be: some too small and some to large discrepancies measured in the same direction. Only thing that comes to my mind is that the width of the walls are wider than they should be. Can that be a result of over extrusion?

I do recognize a little over extrusion in the top surface of the part. But at the FSR bed level the surface looks just fine. Looking at the back of the mount, in the first few layers there seems to be under extrusion. Ofcourse all without touching the flow dial.

Spoiler:

[img]http://hennekens.info/rostock/FSR%20mount%20top.jpg[/img]

Spoiler:

[img]http://hennekens.info/rostock/FSR%20mount%20back.jpg[/img]

This part is printed in ABS at a speed of 20mm/s, 3 perimeters inside out with 50% infill, just like mhackney suggests in his build thread (except he used PLA).

Any ideas?

[Jimustanguitar]

Since you drill your own holes for Michael's FSR mounts, the parts only need to fit each other. Just play with the scaling in your slicer until they fit and it works.

After you get the FSRs installed, put your gemoetries back to stock and let the auto calibrate do its work. You can check your dimensions with a test cube after the upgrade and worry about the root cause (if it still exists) then. I'm going through the same sort of upgrade shenanigans right now, and that's my plan.

[DeltaCon]

That's true Jim, I'll follow your plan
But I am seriously doubting that it is a geometry or calibration problem. If it was, i believe the discrepancy would grow along with the distance to the center of the bed, and it would not be possible to switch between positive and negative.

Looking at the pictures (I just edited the links) do you agree that at the first layers there is under extrusion, while at the top layers there is over extrusion? Any ideas about that? In Slic3r I use a first layer height of 120%. I am a bit concerned that the dragging of the nozzle that occurs sometimes moves the buildsurface the amount of the little play that seems inevitable.

[lbarger]

With OD's being too large and ID's being too small, it sounds like you could be over extruding. (A problem I just learned that I had.)

First, determine if the e-steps on the extruder are correct. I removed the boden tube and positioned the filament where it extended a few mm out the bottom of the extruder. I then extruded a specific length (50mm) via Repetier Host. Other programs should have similar feature. I measured the length of the filament (holding it a straight as possible) and took the difference. If not 50mm, would need to adjust the e-steps in the printer config. Formula would be new e-step = old e-step * (50mm/measured difference).

After doing this I found I still printed .4mm wall at just over .5mm thick. I found Sli3er was set to 'auto' for all my extrusion widths. I changed this to match the nozzle diameter of 0.4. The next thin wall I printed was nearly perfect.

I now find that my part was slightly small. I am currently printing a couple of objects of different sizes to determine the error scales with size or is a fixed offset.

Hope this helps!