What went wrong with this?

The picture is of the same orientation as it was printed. The bottom cone where it is getting bigger is fine but the top cone where it is getting smaller came out knackered. If Madonna's bra had done that she would never have made so much money.

I did wonder if the other extruder had been hitting it but there is no green stuck to it so maybe not. It just looks as though maybe it was too hot when it got there or maybe it isn't waiting long enough between layers when it gets to the small bit at the top.

It would be helpful to see another perspective of the print. But it kind of looks like there is some 'slop' in the left to right movement. Printing a 25mm calibration cube would give better clues. Your bridging looks like it is working well however! How about you print this and post a few pictures?

I believe that is residual heat from the hot head causing the newly deposited and already deposited material on those tiny layers to stay liquidy longer, because on tiny layers the hot head doesn't ever move far enough from the already deposited material to allow it to cool very much.

There is a setting in slic3r where you can slow down the print speed if a layer takes less than (enter your threshold) amount of time. I set this to 10mm/sec if the layer takes under 30 seconds. This allows the cooling fan to have more cooling time, so the already deposited material gets cooled down.

Attachment 1860

I am quite sure a 25mm cube would print fine. I think that it is, as the second reply states, getting too hot generally under the print head. What it needs is a "go home" for 5 seconds at the end of each layer particularly when the layer is only a 5mm circle.

It is on a Wanhao Duplicator 3 using ReplicatorG. I don't know what slicer that uses but I will investigate.

Thanks for the replies. I will investigate further and see how the cube prints.

What did you find, Mjolinor?

I didn't find anything yet, I haven't turned it on.

I have been trying to work out how to make it do what my Stratasys does, that is to go home and wait for a few seconds then wipe it head at the end of each layer. It does seem to me that a head wiping brush and that short pause are two of the little things that the Stratasys does to gain small improvements in quality.

I can add the Gcode manually but figure it can't be that hard to change the source code for Marlin to make it insert the C code on print.