Issue with my 3d Prints

Maybe not really related, but I’m getting a bit “antsy” for when I’m gonna print my Cutty Mc C… sorry, Primo I mean.

When I print something, for instance I printed a case for the Ramps 1.4 + Arduino that they are suposed to slide into, which they don’t, it looks like the track that the arduino slides into is to shallow either on height or not wide enough, it’s kind of hard to determine

Same thing with a HDD case for a Odroid USB3.0 to SATA board that is suposed to slide into the case, it’s also to tight in one dimension I can’t really determine.

Any Ideas what could cause this? Should I try PrusaSlicer instead of Cura? What settings should I look at? Am I “squeezing” my prints to much?

1 Like

Print a calibration cube and verify the XYZ dimensions with calipers. Also verify that you’re not over-extruding and causing the filament to push past the edges. Dialing in the settings to the printer is very crucial when printing parts that require dimensional accuracy. I had my last printer dialed in such that I was able to print an Impossible Dovetail that fix together perfectly.


What you are describing is from a printer that needs to be tuned up. There are a lot of great YouTube videos showing how to do this.
Here is a small list of things to check for when tuning.

*Flexibility of moving parts. Grab the nozzle (when cool) and try to wiggle it. It should not move much if at all. If it does check for loose bolts, loose belts, worn out bearings, loose grub screws etc…
*accurate movement X,Y,Z,E this is where you tell your printer to move exactly 100 mm and check to see if it did.
*over and under extrusion. Most of the time these problems are caused by cheep filament that isn’t consistent in its thickness but it can also be caused by the printer.

To test I like to print something called a “tolerance maker coin”. Find it on thingiverse. It’s demonstrates how your printed pieces will slide together after printed. It lets you test different gaps 0.35 mm, down to 0.1 mm. Many poorly tuned printers get the pieces stuck in everything from 0.3mm and smaller. Well tuned printers can pass all of them. My Prusa MK3s and Ender 3 can both print so cleanly that I can move the 0.1mm test piece around. But my TAZ 5 is sloppy enough that 0.25mm gets locked and stuck together forever.

Last time I checked Ryan designed the MPCNC Printed parts with a 0.2 mm tolerance.

1 Like

You also need to check for square.

Print a 150mm square. I just did a 2mm empty rectangle that was 5mm tall. Before removing it from the bed, take a ruler and measure corner to corner. If it’s not square, figure out how to adjust the bed’s movement to get it that way.

It could also be your layer height. Someone on another thread posted that you should pick layer heights that are even multiples of the nozzle (or something like that). For most .4mm nozzles, .2mm is a great choice. I usually print at .25mm and notice that it is sometimes overly tight on prints/holes that are horizontal.


0.25mm should be fine. A little faster and a tiny bit stronger than 0.2mm.

A 150x150x5mm, single wall cube is great, but can be tricky to adhere. If you can get away with it, it is quick and great for calibration. Then print a 20x20x150mm cube with a few perimeters and about 10% infill. That will help you with Z. Those are both for movement, not for extrusion.

Oh, dear sweet Jeffe… Bless your soul.

If the three dimensions aren’t equal, it’s not a cube. It’s a rectangular prism. And that’s still presupposing that all the angles are (supposed to be) right angles. :angel:

But… I start with a cube, and then I scale it. Does a rectangular prism calibrate your printer by any other name?

Hey, I’m stuck waiting for Operations to run mainframe jobs in test so I can validate code that has to go to production at the end of next week.

I spent 2 hours debugging logstash on the corporate servers yesterday because the DBA’s that own that stack swore the issue was with my server.

I’m flying blind with my hands tied behind my back. I can’t run the jobs (no biggie, that’s actually par for the course), but I also can’t access the datasets on the mainframe, so I have to rely on the developer to send me copies to compare to the input files. PLUS, the transfer jobs are borked because the process user doesn’t have permissions to the source drive (despite the request for access being included on the production access ticket which has been completed), and that guy is out of office today. So I have to have a manager ftp files up to the mainframe, then I can run the job, then I have to get the developer to email me the result files, now I have to run some more jobs, then I can query the database, then I get to beat on Excel until the numbers all match.

Did I mention that I’m using input files from yesterday, but the system has to run with today’s date? So I have to account for a date offset, too…

And this is all financial sector stuff, so if it screws up, it’s audit and possibly regulatory issues…

I’ve heard layer height should be no more than 80% of the nozzle diameter, and that it should align with a “round number” of steps for your Z axis motion. The Z moves in steps, not continuously, so you should pick a layer height that you can actually hit repeatably. The slicer will let you pick anything, but if the machine can’t match the settings then error creeps in and can add up.

Check out the Optimal layer height for your Z axis section at

"Optimal layer height for your Z axis
Helps you to select layer height in a way, that Z axis moves only by full step increments. Z axis isn’t usually enabled during inactivity. If the axis is disabled during micro-step, axis jumps to the closest full step and intorduce error. This effect is occuring to some extent even while leaving the Z axis motors enabled. This is most usefull to machines with imperial leadscrews but also for unusual layer heights with metric leadscrews. "

I don’t buy that at all. First off, my Z is enabled when I’m printing. Second, there’s no guarantee you are starting on a whole step. When you home the Z you essentially get a random Z offset, which can end up with any phase between whole steps. I’m not sure what kind of printer they are thinking of, but prusa printers are also going to be rising and falling to compensate for the bed mesh.

Any miniscule difference microstepping would have on the Z would be much less impactful than the other considerations of layer height, such as the number of layers, or the effective area between layers on top of each other. With a big nozzle, you will end up with fewer boundaries between layers, but they won’t be as strong in tension because they won’t have as much surface area between successive layers.

But I still think all of that is not worth the effort and optimizing print times is probably more important.


Ditto. I was thinking the same thing about the starting z offset. You put it onto better words though.