First Print-off — MK3 Wins Hands down!!!

Last night I set out to work with Polycarbonite (PC) and it didn’t go well, at least at first.  It ended in a print-off with my old printer.

Initially it didn’t look good for the MK3 with the A8 appearing to print flawlessly.  Despite it’s shaky start (due to me getting used to the printer I think) the clear winner was the MK3.  In the end the A8 printed pretty much as it always does, adequately but not stellar.

The Competition

I printed a submarine, a featured item on Thingiverse, and it required water proof bearings.  Turns out you can print them too.  Here is a 3D printable bearing that uses air-soft 6mm pellets.  I picked up a jar of the air-soft pellets for $6 USD.  That will provide me with enough bearings to make hundreds of these units, and other bearing based items. What a great discovery, bearing in all shapes and that can get wet.  Just design the rails.  I saw a turn-table design based on the air-soft 6mm pellets and I’m sure there are others. Designing bearing rails is fairly easy in FreeCAD so I was really excited to discover these.  However, I didn’t get to really see them work until I got my new printer.

I’ve been trying to print these with the A8 but sizing has been a little off.  I also was printing in PETG which isn’t as hard of surface as PC.  For a bearing race harder is better.  Sizing on the A8 is something I’ve had trouble with all along and while I’ve come across a settings that helps with holes and such I’ve found it also introduces issues so I won’t go into details here.  The point is, the prints from the MK3 worked without additional settings which is what you want.  The A8 prints all had errors or zits and the completed bearing never actually spun. The PolyCarbonite bearings are really smooth.  I’ll have to try different resolutions to see if there is any noticeable differences not that I’ve experienced success.

MK3 and Calibration

I’ve been having some issue with Y Axis crashes on very slow movements.  It is interesting to see how they are handled.  Unfortunately there are several posts in the forum and most people just turn off crash detection due to Y axis crashes.  With crash detection off the prints are perfect so I don’t know what it is detecting.

Turning off crash detection was my last option. I want to fully understand what Prusa offers.  There was one person that commented that crashes indicated there is a problem and it shouldn’t be ignored.  I didn’t ignore it!  There is no evidence of binding and the belts status indicates the belt tension is within range.  I don’t know what else to do. Since several others having Y axis crash issues and nothing shows up in the prints with it off it could easily be a firmware related issue because crash detection is a new feature.

I continued to look into it but for now turning it off seems to be acceptable.

Thought I had calibrated…

My first attempt to print PC on my MK3 was a disaster.  It just would not stick and I couldn’t see any good reason why it was happening.  At one point I put painter’s tape on the steel plate because I’ve had good adhesion on painter’s tape on the A8 but that didn’t change anything. I guess the nozzle height was way off and it wasn’t sticking at all.  I didn’t want to Live Z-Adjust the nozzle into the PEI surface and the number range I was working with was tried and proven with PETG and PLA.

I moved the spool over to the A8 and printed.  It printed just fine so the filament was not the problem.

Help me Mr. Wizard…

Just prior to changing filaments to PC I had gone all the way back to running the configuration Wizard but stopped short of the First Level adjustment. I already knew the correct Z adjust value so I didn’t see the point.  Out of desperation I loaded up some PLA and went through the adjustment after the PC printing failed.  The end result was the same number I was already using.

Now it works?

I loaded up another roll of PC filament and started a print.  It worked flawlessly.  I don’t know what was up but running the First Layer Calibration seemed to have fixed it. As far as I can tell the only thing that changed was that I ran First Layer Calibration.

I had dozens of Y axis crashes during the very slow and fine printing of the top layer of the bearings so I had to disabled crash detection.

The Judgement

With the A8 printing already I had an unofficial print-off in the works.  The bearing races are something I’ve had a lot of trouble printing with the A8 and this run was no exception.  On the A8 there was a Z axis crash that created zits  on the surface and messed up the smooth railing. The bearing works but is not as smooth as it should be.

As I said, the MK3 had a shaky start.  The part printed well but the inside rail broke apart really easily (that setting I mentioned earlier).  The outside rail was viable but over-sized (that same setting again).  Resetting the setting to zero I also raised the printing temperature and produced a perfect rail.  Then I printed several more to prove it was repeatable.


I keep upgrading the A8 and will have other print-off sessions.  For now the ultimate goal is to build a printer that has similar print quality to the Original Prusa MK3 .  I’ve seen that when it really matters there is a noticeable difference in the quality of the print and I want to identify the components that make the difference.  I have ordered a new mainboard with TMC2130 drivers and Titan fed E3D hotend.  I’ve also ordered a 3 in 1 out mixing head so there will be many posts about what I learn from those upgrades.

Joseph Prusa stated goal was to ‘print smarter’. I really like having some tools to help me tune the equipment and report on it’s health.  I would like to say to Joseph and his team, job well done.  Keep up the good work!

With that said, I’ve turned off Crash Detection, and I’ve turned off Filament Detection.  I found that when I printed with a translucent orange it stopped mid print time and time again demanding I change the filament.  I finally thought about the filament sensor and realized semi-transparent filament might not register properly.  I was right!

So, Crash detection may be buggy and filament detection doesn’t work with some filaments.  Progress is rarely straight forward.  The MK3 is young and I expect these are minor problems.  Overall I really like my MK3.


Author: Marc

Leave a Reply