Odd Calibration Results- Read a Ton on This Already

I was getting extremely poor results out of my printer when I bought it used (barely) a few weeks ago. I’ve read countless hours of thread on here and other forums to try and figure this out. It came down to me thinking I need to adjust my steps for, essentially, everything.

I followed the eSteps advice from another thread and did the 100mm extrusion and adjustment process, with the feed slowed down to 40% of the firmware default. Made some changes (789.3 steps per mm) and got it to be exactly 100m.

I then started printing more calibration cubes… and they still weren’t right. Most notably the wall thickness was 1.15mm instead of 1mm like it should be!? But the eSteps calculation was perfect?! Just for giggles (cause I know threads say that flow rate should just be for quick and dirty filament difference reasons) I knowed down to 95%, then 90%. 90% flow rate gives me the correct wall thickness for the eSteps calculation that made for perfect extrusion.

I’ve now been adjusting the XY steps since my wall thicknesses are correct. I’m finding myself somewhere around X100.1 Y100.9 for a good square cube. However, the flow at 90% thing bothers me… so I feel I’ll have to go back now and adjust eSteps…

I feel like I’m going in a circle here, just to try and get things to print so that parts fit together???
Frustratedly your’s,

PS: M92 X100.1 Y100.9 Z1600 E789.3 but with a flow of 90% gets me exactly the following on a 1mm wall thickness hollow 20mm cube:
X= 19.99
Y= 19.98
Wall= .99-1.02

Sounds like you probably did this, but just checking - have you measured your filament diameter and updated your slicer with that diameter? Specifically for the spool you’re using to print right now?

Absolutely! It’s Inland black from MicroCenter and has a 3-points average of 2.86mm.

With the settings listed above I printed out some interlocking parts. A circle within a circle, that goes into a square, that goes into yet another square.
The smallest circle I was able to get in (but not back out), the larger circle I can get in and out of the smaller square with my fingers, the smaller square goes in and out of the larger square pretty easily… almost too easily for my liking…


Ok, one more update here. I printed out the S-plug calibration test here: http://www.thingiverse.com/thing:52946
This went fantastically and worked exactly how it should as per his videos!
However, here are my surfaces. Are these normal-looking to ya’ll? I’m still concerned that I have to reduce flow 10% to get my machine to work.


Have you measured your skirt thickness coming off the bed? Grab a pair of calipers and measure it first. That will tell us a bit about the initial layer and what it should look like.

That’s when I’m asking it to print at minimal resolution (aka: .25mm layers). That said, I’ve told it to z-offset .4mm as I was having issues with a really wide first layer that looked sort of squished.
Not sure how thickness of my first layer affects parts fitting in the XY dimensions though?

A mini is set to deliberately overextrude on the first layer. This promotes better bed adhesion for new printer users, but leads to override first layers and “elephants foot” on part edges. You can reduce that number if your adhesion levels are good.

Yes, I have figured out how to reduce that elephant foot as my adhesion is fine. I’m more curious as to why Jim2386 thought this might have something to do with parts, at 100% flow, not fitting together even if they fit together at a location that wasn’t associated with the first layer (see S-plug example).
Maybe I misunderstood?

Bump. Any additional info on how to get parts to fit and not have to under-extrude (leading to space between my bottom layer lines as seen above) would help me out a ton!
I had a friend print something for me on his Ultimaker and the parts both fit and had a better finish than I’m getting (same parts I tried).

In some slicers, there are controls for the initial layer width, could try 200% to compensate or lower the initial nozzle height (via gcod Z-offset) to increase the “ovalness” of the initial layer extrusion.