Could also be motor pulley slipping on the shaft.
@Ryan_Carlyle So I ripped out the core XY gantry and installed a H bot style gantry. I got precision Linear guide rails from Hiwin for the Y axis to avoid racking.
But the racking still continues. It is a very tiny amount but it happens when there is a direction change in the x axis. Can i do anything else to reduce the racking ??
@Ryan_Carlyle I have seen many H bot printers and they seem to work fine. Am not sure how they have solved the racking problem though.
Doesnt precision linear bearings sort this out ?
@Hussain_Bhavnagarwal Linear bearings alone, no. You need Y carriages that resist racking more than just a single linear block. Using a “long” linear block or two linear blocks in a row helps a lot. Check this image: http://www.etotheipiplusone.net/pics/kewlmoter/mkf4.JPG
@Ryan_Carlyle Thanks for your reply !
Is there any rough calculation to know the length of the linear bearings compared to the length of entire Y carriage ?
That would be helpful in procuring the right parts and laser cutting the Y carriage again.
@Hussain_Bhavnagarwal ehh. I think if the Y carriage length is ~1/4th as long as the X bridge, that should work well. A smaller value may work just fine, but I would be confident that 1/4th would be pretty good if the X bridge is mounted stiffly to it. The downside is lost Y travel. You could try cutting a couple different lengths and seeing what works.
Or just go back to CoreXY which doesn’t have racking problems 
@Ryan_Carlyle so I tried my best with the H bot system and though it prints well it does leave some artifacts due to racking.
Am back to the Corexy system again, but having similar issues as before. I have ensured that all pulleys are levelled out and are sitting at the same height. The tension on the belts are good and do not cause any slip. I tried recalliberating the X and Y axis,
The x axis seems to fall short by less than 1 mm, will that cause a major issue ?
For the z axis I have used an 8mm lead screw, the lead screw sits into a 5mm to 8mm coupler and the problem is when I tighten the coupler around the 8mm shaft it tends to get off centered a bit. This can be noticed when the z axis is being moved down by 100 mm and the slight wobble due to an off centered lead screw is visible.
Can this be an issue ??
Below is a picture I have taken after printing at 0.2mm layer height.
This is pretty close to some good quality but still feel like am missing something.
missing/deleted image from Google+
What do you mean the X axis is short?
For the Z screw, you can either get a perfectly straight and centered screw, or you can make sure that when the screw wobbles it doesn’t transfer forces into the Z stage. What I like to do is mount the Z motor on an astrosyn style rubber damper, which lets it gimbal a small amount (and makes it pretty quiet). And then just make sure the far end of the screw is free to move around if it needs to.
I’m seeing a little bit of raggedness on the left corner of the print in your picture… which could mean the Z stage is moving around a bit, or you need a little more print cooling, or some slop in the XY stage. Hard to tell from a single picture.
The verticle ripple type lines in your print are something that a lot of printers have when the quality is good. Try changing print speeds to a few different values (like 10, 25, 50 mm/s) and see if the ripple spacing is consistent (which could be a belt/bearing issue) or if it changes with speed (which could indicate a firmware or stepper driver issue).
@Ryan_Carlyle By falling short I mean that , when x axis is moved by 10 mm it actually moves only 9.8 mm. I recalibrated but did not make much of a difference.
The raggendness of the edges decreases when I decrease the layer height from 0.2mm to 0.1 mm
The vertical ripples are actually the points where the infill touches the perimeter. I experienced this when I changed the settings of the infill and the vertical pattern changes according to that.
I have also noticed that the first 5mm of the print is perfect but after that it gets a bit wavy. I suspect the z axis vibrations, I think I will pick up a 10 mm lead screw, machine the bottom to 8mm and try fitting it perfectly. I also print the cube as close to the lead screw as possible to avoid any artifacts due to the vibration of the cantilever bed.
Is there any way of checking the repeatablity of the x y gantry ? I am wondering if I should stick a pen to the nozzle and make it draw a line over and over again to see if there is any slop in the xy gantry.
Do you think this will work ?
Try a backlash calibration print.
You should be able to use the calculated steps/mm value for your motor, microstepping, and pulley size and get accurate motion. Try big motions when you measure travel (like 100-200mm) since that will minimize the effect of small errors like backlash, extrusion volume, etc.
@Ryan_Carlyle whats a backlash calibration print ?
I was also looking at the verticle ripple problem and it is not due to the infill. I found a detailed forum post on this but no one has been able to solve this. Looks like a lot of corexy machines are running into problems.
I am planning to stiffen the z axis even further with thcker rods and and bigger bearings and also need to ensure that the z axis does not wobble. Planning to machine the z axis rod to 5mm and use a 5mm to 5mm coupler.
Those vertical ripples can appear on any type of printer with sufficiently high-quality motion to eliminate other flaws. There are multiple possible causes. I see them on my R2x Cartesian and a couple CoreXYs. Deltas also get them in many cases, although that has been definitively narrowed down to microstepping ripple when the layer heights are certain multiples of the step size.
This thread is getting pretty long and no one else is contributing; you want to make a new post so more people see it?
@Ryan_Carlyle Will start a new thread from the next post, am currently redoing the Z axis with thicker rods and longer bearings, once am happy with its stability I will shift to adding bearings on the Y axis.
Thanks