Any ideas?

So 2nd print, see 1st in background.

I posed an update in another post but for Ref. Here are the new prints. I still need to PID tune the temperature and improve the bed level and tweak my phase settings but it is night and day. I think now it is tweaking the settings and I will be on to .1-.1 settings I hope. then I can print something I designed

and a v8 block

Banding is inconsistent over the prints.

I posed an update in another post but for Ref. Here are the new prints. I still need to PID tune the temperature and improve the bed level and tweak my phase settings but it is night and day. I think now it is tweaking the settings and I will be on to .1-.1 settings I hope. then I can print something I designed

and a v8 block

well just installed and ran zwobble and it seems promising
I also remeasured the first box from above (on the left) Where I had a0.71 wall thickness I split it up and tried to measure just the top 1 strand thickness and I was closer to 0.4MM so almost on target

After a few more trys I am around .73-.76 flow I would say even .8 works but also depends on speed.

after z wobble and the flow test I will try to reprint the iphone dock as a test and see if:

1: zwobble improves
2: delamination is minimized

Show us.

After a few more trys I am around .73-.76 flow I would say even .8 works but also depends on speed.

after z wobble and the flow test I will try to reprint the iphone dock as a test and see if:

1: zwobble improves
2: delamination is minimized

I am wondering when I measure these cubes I am measuring the width and the errors of the banding. so it is not the tru width of the wall but the stacked error. So to get a stacked wall of .42mm and if they are off by say 30% +/- my all could be 30% too thin? hence my flow issues and why .778 flows well?

I tried again with and without the 2 z braces and the round one definitely improved things so I left on. Then I tried to tackle my flowrate and something seems way off.

left to right a .5" square cube at .42mmm wall 2 solid layers on bottom of cube no infill 1 perimeter .3 print .3 first layer 35mm/s all 1.7 diameter (after several averages) and .778 multiplier (original)

I did the first print on left. Came out solid (which I like) but the I could tell the first layer stuck well and flat. But he solid infill drags lines into the plastic and they turn bright white as the extruded pulls thru them in some cases pulling up or dragging a small blob.

I figure too much plastic makes sense and the wall thickness was .7mm vs .42! so that meant a .455 wall thickness so I tried that That was the print on the far right (not middle) It flowed even and very thin but as you can see the bottom solid layer had gaps of missing plastic, the strands didnt stick together almost cooled as they came out, and the model collapesed on itself as the extruder pushed down.

the wall measures at .405 so right on target but still seemed so low.

So the middle model I ran at .5 flow multiplier and it came out much better. the wall thickness around .5mm
but still banding was very bad, no rigidity and still gaps and wholes in the wall like not enough flow

Then I printed up at .7 just to see what happened.

I ran a 2nd backlash nut

see the 1st one i printed the bed was not level on that edges (in background and that was at .6 multiplier and still had gaps in fill so I said lets do original .778 multiplier)

I think it looks good? at edges you see the slight banding from the x-y combined error or too much flowrate?

.3-.3infill all 30mm/s and 18mm/s 1st layer

Happy with that one and I was playing with various fill patterns and speeds I determined this:

for solid walls 3 perimeters is good 2 is too thin and .2 fill feels too light and hollow for rigid parts (if needed 0.4 was good fill rate)

And the 1st layer I think I have a too slow of a speed from 12-18mm/s for the .3 and flowrate it globed up doing the solid layer passes and dragged over perimeters. so with the next model I went up to 22mm/s

This was my first real model and fairly large
I stuck with .3-.4 infill-honeycomb fill every 1 layer. I kept speeds at 35mm/s except for small infill (at 30mm/s) and external perimeters at 40mm/s 1.7mm diameter-.7 multiplier (backed down from znut to see if made a difference on 1st layer) it did less globing but wonder if upper layers delaminated because of this?

The first layers went down ok some very slight curling on these bigger models seem hard to dial out?
the first layer you can see the variation in layer colors? what is that too much flow? speed?
and the final product seems solid but you can see several layers look pulled apart and didnt stick?

I had perimeters to 3 (and generate more was unchecked) and had 3 solid layers top/bottom
was this too thin for this slight slope?
is this a heat issue?

again at .3 the banding layers is there but the model doesnt seem zig zag all things considered it looks fairly even edges and holes are square and tight no flashing just sticking together seems issue?

http://wiki.solidoodle.com/z-axis-anti-backlash

What does the brace really do that a spring and washer and nut wouldnt? I mean the tolerances variations are on order of fractions of MM I do not understand what the guide brace between the rods will do (since it appears to be a loose fit) compared to fractional MM?

The problem is that the loose drive nut sits lowest on the threaded rod only when it is centered.  When there is any lateral force (such as due to an imperfectly straight or centered rod) it causes the nut to rise up slightly.  When printing with .1mm layers, the amount is significant since each layer is only 1/14th the width of a single thread.  Here's a post I made with a diagram:

http://www.soliforum.com/post/10340/#p10340

The brace only functions to keep the top nut from rotating when the rod rotates so that the spring stays compressed.  It doesn't have to be precise because the exact force of the spring isn't important; only that it is sufficient to keep the primary bottom nut from rising up.

what size spring did you get? diameter,  force?

It was just one I had in my parts bin, so I don't know the exact specifics.  It was about 1/2" diameter overall and about the same in height after I cut it down.  The wire diameter about the size of a large paper clip, so pretty beefy, and stronger than the one I used initially.  The force has to be enough to seat the nut and overcome any lateral force from wobble in the z-axis rod (which has a tendency to make the nut rise up), but you also don't want to put any more load on the motor than necessary, which is why I recommend using some way (like a dial gauge or my laser pointer rocker mount) to measure how much backlash you still have.

what size spring did you get? diameter,  force?

Thanks that makes sense. Is my banding extreme ?

Also I think my flow is still a little too high. the multiplier seemed off by a lot but I wanted to take small adjustment steps at first.

what wall width should I use for a calibration measurement at .3,.2,.1? just use .42mm to measure flow rat at .3?

any good links to spring znut?

The banding looks pretty bad; you should be able to do much better.

This is the preload I use.  I use a fairly strong spring but cranked down the nut just enough that I can move the bed up and down and the laser comes back to the same spot:

http://www.thingiverse.com/thing:34397

For the calibration, you don't have to iterate, and should be able to get the correct extrusion multipier in one shot.  Mine was way off at first. I leave the slic3r profile set to the default width (.42mm) and print a solid cube with one perimeter and 0 infill, stopping it before it finishes (as wiki suggests).  Then I measure the wall width with calipers and divide .42 by this number.  You can just apply this number as a multiplier to your "extrusion multiplier" value in slic3r.  Don't just replace the extrusion mulitipier with this number (as wiki says) unless it was originally "1" when you sliced the cube.

For instance, if you, say, measured .63mm (instead of .42) then your flow rate would be high by 50% and your adjustment factor would be .42/.63 = .67.  You'd then multiply your current extrusion multiplier by .67 and should be done.  You can slice and print another cube to verify.