The standard VFA test that almost everyone seems to be using seems well enough designed for Cartesian machines with independent axis control but is just not enough for modern core XYs.
There are a few reasons but the primary reason is that it doesn’t catch the interaction between the two motors and belts. Just shifting a model 0.5mm to the left or forward could mean that the VFAs completely disappear or peak.
Two people could print the same test on two machines with the same VFA-ness but with slightly different homing positions and one person could get great looking results while the other gets awful results.
the standard test also doesn’t cover what happens when the tensions for the top and bottom belts are slightly or a lot different - there could be certain angles where there is suddenly a lot of VFAs.
I’m working on a new test to capture these effects and here is what I’m seeing so far.
I even managed to find VFAs on the H2D with the fine tooth pitch belts which hardly showed anything on 0, 90, 45 and 135 deg angles as you can see below (this is similar to the previous pic which is a sample from the P2S but this was printed on the H2D)
I’ve done a more comprehensive write-up on this page, with the test models, for those that want to really nerd out.
(This is not a P2S complaint thread! It’s about VFA testing on core XY machines in general)