MK3S... Was it worth the upgrade from MK3?  

RE: MK3S... Was it worth the upgrade from MK3?

@texy

Texy, the MK3 impressed me no end for PLA, PETG and ABS, it was only flexi-filaments that were finicky.  Smaller models with short run times (like zippo lighter sized) were fairly easy to print in TPU, but larger more complex models often failed for me (flex filament forced out the hole in the idler door or wrapped round bondtech gears). 

So far I'm not getting good and consistent TPU print results with the MK3S at all.  The filament feed is more direct so filament doesn't get out of shape so easily (although it can still bend and twist below the gears), but there's something wrong with my MK3S extruder setup that's causing flex prints to under extrude (I've got a thread on it in the 'how do I print this section).  Sorry, but in all good conscience I can't give an opinion on whether it's worth upgrading or not yet as I've still got to figure out exactly what's going wrong. 

I hadn't read many reports of issues with the MK3S extruder setup and the problems that did exist seemed to be mostly with the new filament feed.  I did read a thread about under extrusion due to the PTFE throat tube being too long and too close to the bond tech gears, but I knew about that when I upgraded and that issue's already been solved by Prusa anyway (new parts supplied in newer kits and instructions to cut down older longer tubing) so I'm currently at a loss as to what's causing the issues I'm experiencing.  I'm going to take the extruder assembly apart this weekend and check all parts carefully.

RE: MK3S... Was it worth the upgrade from MK3?

Posted by: @nathan-b20

[...] The MK3S extruder was designed with a more direct feed path compared to the MK3 so it was suggested that flexible filaments would be handled better.  But in my experience whilst the filament is not coming out the side of the MK3S extruder like it did on the MK3, the bond tech gears are cutting into the TPU and something so far unknown is stopping the filament from successfully exciting the nozzle intermittently during prints.

I just want to jump in and clarify here. There is not any unknown behind what you're describing. The E3D V6 hotend has a finite capacity to melt and move filament through the hot nozzle. This limit varies depending on temperature and the size of the nozzle opening (to a point). If print speeds are too high (meaning the extruder is pushing too much filament) you get the extruder skips with stiffer filaments, or filament squirting out the side with softer filaments. Slow the speeds down (best using the Maximum volumetric speed setting in the filament profile) and there's less chance of back-pressure.

The Mk3s improvement was further constraining the filament path so there'd be less chance of it squirting out the sides. The Mini has an even more refined path. There were mods for the Mk3 that essentially created a tighter fitting shroud from the Bondtech gears to the PTFE tube.

The issue with over-tightening the Bondtech gears is that it deforms the filament. Too tight on the original Mk3 design and you may have feed issues and soft stuff likely to curl around on the gears rather than feed into the PTFE tube into the hotend. Too soft and you get the clicking quickly. There is no sensor or logic behind this. It's all mechanical.

The Mk3s attempted to "fix" the filament sensor and hotend fan cooling airflow. The part cooling fan was meant to be more efficient by reducing sharp angles. I made a conscious decision not to upgrade based on the following:

1. The original optical sensor requires regular cleaning out. I use an air cannister. Fail to do that and it becomes highly annoying with failures. I get a failure per week if I'm swapping filaments. I do disable it for the odd filament it has a hard time with. The new design seems susceptible to adjustment needs of its own. Not sure it was a big improvement for the gain.
2. The heating "issue" was a debacle in my mind. Prusa issued a fun blog post about the Lack enclosure in May IIRC. Lots of people built them only to encounter heat issues in summer of the same year. The E3D V6 hotend is rated to 40C ambient. Get above that and it loses cooling efficiency. The Mk3s design is supposed to improve air flow but I never had the issue once I figured out airflow in general is essential. Put off plans for an enclosure since I don't print ABS or Nylons much that most benefit.
3. The cooling fan was angled down in front which doesn't impress me too much. I pulled the original fan into Fusion 360 and added a 45 degree slope at the sharp edge. I tested 6 different fan designs and found no improvement so have stuck with that design. I haven't seen any Mk3s results that make me want to change. So far as "too much" fan, cool print temps or reduce cooling.

None of these are a big deal either way. If I got a new Mk3s, I wouldn't worry about it. They're just not improvements, but they're certainly not defects.

So far as printing TPU, I've had good luck by:

1. Slowing down print speeds (via Max volumetric speed setting).
2. Using hotter end of printing range to improve hotend flow rate.
3. Using a larger nozzle to reduce back pressure.
4. Turning retractions down to minimum. Springy filaments simply don't handle retraction well.

I'm burning off a 2 year old spool of Inland TPE with a shore hardness of 85A without much issue other than the expected stringiness.

So the point of this long-winded post is that while the Mk3s was not an amazing upgrade from the Mk3, it's not a disaster. The latest part revs seem to have dealt with the melting PINDA mount.

RE: MK3S... Was it worth the upgrade from MK3?

@bobstro

Excellent information. Thanks very much. Unless i am missing something, it seems to boil down to the fact that "business requirements" driving the design/engineering requirements, were to "equalize the Ins and the Outs". Of course, this makes total sense  

Engineering history/detail is always interesting. Thanks again!

edit: sorry, i just reread your posr and realized i had gotten it in mu head that you had done the actual engineering. 700am, not enough coffee, reading too fast on a small device....excuse, excuse, excuse. Nevertheless.....

RE: MK3S... Was it worth the upgrade from MK3?

I have no regrets upgrading my MK3 to MK3S, the MK3S prints TPU (98A, 92A, 85A and NinjaFlex) fine with a slow filament profile. When switching from PETG (primary filament) to TPU I loosen the extruder drive two turns.
Before upgrading I had trouble with TPU 85A and NinjaFlex as they could curl up in the extruder gear house. The MK3S has fixed that problem.

Be sure that the filament path to and from the extruder gear is smooth and no contractions or obstickles in the PTFE tube. Auto load can be too fast and for the softer TPUs I prefer a manual load.

RE: MK3S... Was it worth the upgrade from MK3?

Posted by: @bobstro

 

So the point of this long-winded post is that while the Mk3s was not an amazing upgrade from the Mk3, it's not a disaster. The latest part revs seem to have dealt with the melting PINDA mount.

Excellent clarification, thanks bob.  You've obviously managed to explain it better than I was able, and I see the jump off point was my mention of the gears cutting into filament.  I think I may have confused matters by being less than fully transparent when describing the issues being experienced in order to try and explain the situation to those who showed interest (and clearly failing somewhat!)  Sorry about that everyone - I can see how the overly expressive descriptions of gear skipping could have been misleading.

In actuality the issues I have been experiencing whilst trying to print TPU with the MK3S have only led to minor gear skipping, and that was mostly due to the idler screw being too loose during my testing.

It's interesting to read about your experience with the MK3 filament sensor requiring regular weekly cleaning and I've heard the same being reported many times.  Thankfully I never had to clean the MK3 filament sensor as it remained suitably free of dust and never failed to operate correctly, so the only way I can currently compare it against the reliability of the MK3S sensor is that they are as effective as each other - neither have caused any bother.

I've not experienced exactly the same as you when it comes to the cooling fan situation though.  Of course, every printer is potentially slightly different and each owner has their machine set up in a different environment and is potentially in a different part of the world with varying climate/moisture etc.  I'm in the UK, inland, in an averagely insulated property, and I run my MK3/MK3S in a custom vented twin stack enclosure that I built for my earlier printers.  On first look I assumed that the original Prusa MK3 print cooling fan position was a less than perfect design, but during my first year with the printer I was pleasantly surprised as to the quality of PLA and PETG prints due to the effective cooling.  ABS was more difficult but still absolutely printable, and Flex was challenging but doable with the correct settings depending on the size and shape of the model to be printed.

I'll admit I was convinced that the MK3S redesign was going to be an obvious improvement, making prints easier to complete with less trial and error needed.  But that's where I was wrong.  I'm discovering that the MK3S fan angle, coupled with the PrusaSlicer settings for Prusa Filaments, is too effective and cools the nozzle too much and too quickly, negatively affecting prints.  Since manually adjusting the fan speed for standard cooling and bridging I seem to be achieving better results (more tests needed to be absolutely certain).  Perhaps my results are largely down to my environment, but I feel Prusa should have provided amended print settings for the MK3S rather than simply adding a new printer profile and leaving the settings the same as the MK3, which obviously had a very different extruder setup.

The increased cooling efficiency ties in with the mention of reducing the print speed to achieve an improved balance of filament feed to filament expelled through the nozzle (that I also mention I had applied within my TPU focused thread in the "how do I print this" area).  I actually experienced the reduced print speed caused increased levels of under extrusion, which I now think I can pin point to the over-effectiveness of the new position of the MK3S print fan that is still set in PrusaSlicer to run at the same fan speeds as the MK3, because the standard settings have not been altered by Prusa.

I totally agree with you that the upgrade is not a disaster, it's still a fine printer and is still the best I've owned.  But the issues I've experienced have also been experienced/highlighted by some other owners who upgraded, and I do feel that Prusa should have addressed the situation and supplied amended print profiles, if my findings are correct - which I hope they are as I've just completed 2 consecutive successful prints in PLA and then TPU.  😀  

RE: MK3S... Was it worth the upgrade from MK3?

Posted by: @area51

I have no regrets upgrading my MK3 to MK3S, the MK3S prints TPU (98A, 92A, 85A and NinjaFlex) fine with a slow filament profile. When switching from PETG (primary filament) to TPU I loosen the extruder drive two turns.
Before upgrading I had trouble with TPU 85A and NinjaFlex as they could curl up in the extruder gear house. The MK3S has fixed that problem.

Be sure that the filament path to and from the extruder gear is smooth and no contractions or obstickles in the PTFE tube. Auto load can be too fast and for the softer TPUs I prefer a manual load.

Thank you for the good advice area51.  Much appreciated.  I had thought the filament load speed was a bit zippy too.