Solid Infill not meeting Perimeter (Page 2 of 2)

the delrin block at the rear of the x-carriage is supposed to be 'loose' - if its too tight it can cause binding issues, so there is a as-designed play of .1 to 1mm's  in it. The only guide/support required on the stock x-carriage is the front bushings.

They designed it this way so that if the rods aren't perfectly perpendicular and square, it doesn't cause grabbing and binding.

So you may actually want to adjust some play back into it - it shouldn't flop about, but it should definately 'rock' a little teeny bit...

I will carefully follow this thread because I have the same problem.
I performed all calibrations, made sure belts are evenly tightened, the pulleys stays parallel while moving, no belt play etc...
My circles went from horrible to acceptable, but they are still not perfectly round, they display the same shape as Gomisan printer, and this mechanical issue is also the same cause of infill not adhering on some perimeters.

By the way, the front left pulley block is getting bent. I have a new set of blocks, much stiffer than the stock ones, but I cannot unscrew the internal screw that keeps them in place. It doesn't seem to be neither 2.5 nor 2 mm hex screw, but a Torx T8 fits.
Are these screws glued? Any advice on how to remove them without destroying everything?

Thanks
Nick

lol  i looked at it also.. thought the same thing.. then thought, Gomisan is going to try it, i'll wait for him to figure it out then ask him how to use it! 

The "solid solution"  is calibration, lubrication and ensuring general smoothness of every moving part.  Past that adjusting VREF's on the stepper drivers may help a bit.

Thanks for the reply, it and additional infill threads pointed me to;

h t t p colon slash slash vimeo dot com slash 55031028   <-- Evidently I'm not allowed to post links.

This is for getting the Y belts right.  Now I didn't think too much of it, but the roller bar for the Y axis had fallen out during shipping to me and I had to put it back in.  So the above calibration helped a lot, in fact knock on wood, I think my infill to perimeter issue is now solved.  My circles are almost perfect... now I need to find an X calibration procedure.

I've seen the Vref pots, not entirely sure their impact on the output, yet another procedure to find...  I'm actually pretty happy with the current output, but will continue to tweak... damn ocd....

That block mechanism was a pain. This gadget to springload the block helped a lot for me:

http://www.thingiverse.com/thing:214112

Any time the print head moved near the back of the bed, I was getting filament wandering all over the place because the head would rock. Adding the spring to hold the head down with more force than gravity seemed to improve my prints a lot.

No, its the max feed rate - the actual print speed is based on an acceleration vector from the X/Y or Z Jerk (E is hardcoded to 5 in firmware) to the value in max feed. For example, any 5mm moves for any axis will never reach 'max print speed' based on 1000mm/s2 acceleration (default in official firmware) and 70mm/s feedrate.

And then we get into the complexities of combined x/y acceleration.. where its actually the max feedrate of the slowest moving axis.. so if its a 30° diagonal to square line for example, its going to not accelerate to the same velocity that a 45° line will or a 90° line will...

Your printer will basically spend only about 30% of a print at 'max feedrate'....
This is why the acceleration vectors are actually far more of an influence to quality often than outright max feedrates... assuming you have max feedrates that are appropriate for your printer that is !