The image on the left below shows backlash in the y plane. Notice the groves on the ends of the part. These are supposed to have 90 degree corners, but instead, have odd, uneven shapes. The y plane belt is loose and as the y motor stops, the bed continues to travel, until the belt is no longer slack, overshooting the intended position. The bed then bounces back and forth until the belt tension is equal on both sides of the motor pulley.
The image to the right below shows how the part is intended to look. I tensioned the y belt just enough so that the belt had no sag along its entire length. You have to be careful not to over tension the belt which can cause the bearings to bind on direction changes, excess friction on the y axis motor, and the possibility of missing step counts.
I've read about a 3d printing problem called backlash, but I really didn't know what it was or how to spot it until I started working on my dish rack project. I knew it was a problem related to belt tension, but couldn't find any examples or clear definitions of it. As I was printing the pieces for the dish rack, I noticed some squiggly lines in the parts as they were printing, especially when there was a rapid change in direction in the y plane.
I needed a small dish rack for my small apartment in Charlotte. I usually only have a plate, a glass, a few cooking utensils and such to wash so it doesn't make sense to use the dishwasher. I designed and printed what I thought would be a pretty decent setup, shown to the left. It worked OK for a salad plate, but a dinner plate would not stand up on it's own. I violated one of the basic tenets of prototyping in that I printed the entire model without first testing the design with a few pieces. I wasted plastic and time printing pieces that I would never use. Live and learn.