ABS infill issue

I'm starting to think that the problem is underextrusion.

I got good prints with my none auto setting (0.4 width with 0.4 nozzle) and I calibrated my printer with this settings. Calibration object was a thin walled cube (without top and bottom) with 0.4 wall thickness. Maybe this calibration process was my mistake.

Now even PLA prints don't turn out good. Perimeters have gaps and infill looks underextruded (see picture attached).

Right now I'm printing with automode and I expermiantally set the extrusion multiplier to 1,15. It seems that it gets better now.

But my concern now is that parts that should fit toghether will not fit after this change. As it was before the calibration.

I finally managed to print in a better quality with ABS. It was indeed underextrusion (mixed with to small width values).

There is still a problem with details. In the picture attached you can see that there are small gaps between the detail perimeters and the infill. How can I avoid this?

Also I'm trying to print the sun dial again. I attached another picture showing the first layer. Is this overextrusion or should it be like that?

The gaps between the perimeters and the infill may be closed if you increase the infill overlap. You may try to print the same thing but slice it with Cura. For some things, especially holes and structures like this, Cura gives better results. Or try the new Prusa Sl1c3r.
At the bottom layer you can't tell if it is over extrusion or a bed distance issue. My first layer sometimes look similar, probably depending on the room temperature, the moon phase, the cat's mood and other strange things. Try to lift your head by maybe 0.02 or 0.04mm. I think it's hard to find a good compromise between the first layer beeing squashed and first layer adhesion. If you are 0.01mm too low, it gets squashed, and if you're 0.01mm too high, small structures loose bed contact while printing 1st layer.

I i tried cura and you are right the results are better. I also tries the prusa edition but it does not show better results.

I designed a bridgetest that should show if its possible to print the sun dial. (it is not an important part that i need, but now it's a challenge ;-) )

I attached this bridgetest file. Maybe you can test if you are able to print it.
I printed a lot of them with several differnt settings and even two different abs spools. Some results are better, but never "perfect".

I had better results with 0.2 layer height, about 210°C, 0,5mm nozzle (ext. perimeter 0,5, others automatic).

The sun dial should be printed with 0.1 layer height.
So I also tested 0.1 with no good results.

The bridges in the testfile are 0.2 mm high. So should be two layers for 0.1 layer height.

I can't print at the moment. I did yesterday, and I noticed hotend temp issues. I print ABS at 240, and my temp varies from 239 to 240.5. Yesterday it jumped up & down, about 20-30 degrees. And it jumped so fast, I assume a loose contact to the temperature sensor. Usually the temp doesn't fall or rise 20 degrees in less than a second. Anyway the printed parts came out quite well. I didn't find any obvious problem, so I'll gonna rewire the hole hotend. I'll probably won't find time for this within the next 2 weeks.
And to your bridging: from my experience, Cura is better for making holes and similar, but bridging is better when sliced with Sl1c3r. So I don't think you'll get a successful Cura sliced sundial print.

I don't get any bridges when slicing with Cura or Sl1c3r, probably because of my nozzle settings.

Hi the bridges are 0.5 mm wide.

I attached a new version with 0.6 mm width, maybe this one works with your settings.

You may try the bridging thest from here: [www.thingiverse.com]