RE: MK4 Wishlist
Hmmm, maybe not after all, this revolve contains a complete linux system as well as slicing software and more besides and is MUCH MUCH bigger in size of code.
It does look very interesting though.... Ah well, guess we will have to see what Prusa come out with.
Normal people believe that if it ain’t broke, don’t fix it. Engineers believe that if it ain’t broke, it doesn’t have enough features yet.
RE: MK4 Wishlist
Please don't do a touchscreen. It is gimmicky and will only add unnecessary cost to the printer.
Please do consider some of these common mods or their variants:
- OctoPi with x-axis mounted pi camera by default (i.e. give us timelapses and ability to control/monitor the printer from our phones)
- LED lighting for printbed (there are many examples)
- Fixed IR sensor that works with soft filaments: https://www.thingiverse.com/thing:3554066
- Magnet under FINDA ball and closed gap so filaments don't wander off track: https://www.thingiverse.com/thing:3608229
- Use PC4-M10 everywhere you can to reduce the number of jam points, ex: https://www.thingiverse.com/thing:3233579
- Auto-rewind spool holders for MMUv3 (ability to drybox them a bonus), ex: https://www.thingiverse.com/thing:3268855
- Power switch for printer and LED on LCD Cover, ex: https://www.thingiverse.com/thing:3213892
- Although it would be also adequate to simply ship a tplink plug and have the LED lights controlable from OctoPi such that they turn on before print and turn off x seconds after print for good timelapsing
- Physical bed leveling tuning knobs (dont get me wrong, I love the auto-balance that is there, but being able to tweak physical can help a lot too)
- Filament dust filter (many examples)
- Silicon socks for the E3D hotend
- Enclosure mod available for purchase as a separate upgrade (which would include moving the power supply out 😀)
RE: MK4 Wishlist
Maybe not one connector to rule them all, but I’d like connectors for the fans, thermistor, heater, etc. on the print head. That is, I want to remove bits for cleaning, replacement WITHOUT having to unwrap the umbilical back to the einsy.
The printers that do this well seem to use a printed circuit board on the print/extruder head and then just have a multi connector cable from the board so there is a solid plug connexion on the head and the whip can firmly attach to that without flexing at the connectors.
jup my old Wanhao i3 Plus had it like that, one flat band cable going to the extruder and plugging into a little PCB that has connectors for all components.
RE: MK4 Wishlist
Replace fastenings that involve trial-and-error settings with something more permanent. e.g: The MMU unit and the extruder door are in need of being opened semi-regularly, but involve "losing" the spring strength settings whenever you do this.
The MK3S "door" is no longer a door, it has to be removed. This shouldn't be necessary.
The MMU2 buffer is an absolute pain to load, especially without 360 degree access to the printer. Change it (rewinding spools seems to be the way to go).
The MMU2 is highly intolerant of bad "tips" and good tips have a very low tolerance of slicer settings, making it a hard machine to tune. In the event of 2 failed loads, cut the end of the tip off with a blade (or have another mechanical "shaping" tool) and try again.
The MMU2 blade needs to be better. I find the filament bending and jamming the selector as likely as getting "cut".
Make the wizard complete all calibrations. If support are going to tell people to run temperature calibration and mesh bed level is basically needed, the wizard should do these.
Resetting firmware should not need to be followed by reflashing firmware. Resetting should... well... reset.
Design for basic maintenance, please! Unjamming the extruder requires access from the left, and right of the printer (and now the rear on the Mk3S). Basic operation of the printer the front. Turning it off, the right. Tuning the belts, the rear and the left. Loading the MMU the rear. It doesn't need to, move it all to one or two "access" sides for most common maintenance work (can you tell, my printer is in a corner?).
Secure your nuts. I have several that I want to glue down, specifically the square nuts that fall out of slots (when I open the MMU2 cover) or jump out their holes (when I remove the blower fan).
Design the MMU to have a straighter filament path. The buffer deliberately runs through a "kink" (which will eventually grind through the PTFE on either end of the buffer) then runs through >90 degrees turn to the back of the MMU and the exit of the MMU points "up" for another >90 degree angle to the extruder. It shouldn't be designed with all this unnecessary friction.
Have cheaper shipping for things that fit in an envelope/jiffy bag. Buying PTFE for my extruder for £0.65 with a £18 delivery charge sucks.
Switch to e3D nozzle X for a better top layer
Slow down your print farm! The quality of your parts should be showing off what the printer can do, not displaying ringing and require work clearing filament paths, etc.
The cabling inside the einsy box is a mess. Too much spare cable, too many cables, the board is pressured by cables when the door is screwed shut
RE: MK4 Wishlist
Please don't do a touchscreen. It is gimmicky and will only add unnecessary cost to the printer.
Please do consider some of these common mods or their variants:
- OctoPi with x-axis mounted pi camera by default (i.e. give us timelapses and ability to control/monitor the printer from our phones)
- LED lighting for printbed (there are many examples)
- Auto-rewind spool holders for MMUv3 (ability to drybox them a bonus), ex: https://www.thingiverse.com/thing:3268855
- Power switch for printer and LED on LCD Cover, ex: https://www.thingiverse.com/thing:3213892
- Although it would be also adequate to simply ship a tplink plug and have the LED lights controlable from OctoPi such that they turn on before print and turn off x seconds after print for good timelapsing
- Filament dust filter (many examples)
- Silicon socks for the E3D hotend
- Enclosure mod available for purchase as a separate upgrade (which would include moving the power supply out 😀)
While I agree with and have done some of these, these are more user preference that needed or (to some people) even desirable. I, for example, have no interest in an OctoPi or lights on the printer. I'd rather them keep the cost and complexity down by not including such things.
Now keeping abreast of how users are using/moding the printers to make that process easier/better (e.g. that it has built in support for plugging in the PiZero) is a good goal I think. Maybe simply having some related options at purchase would be helpful too (e.g. I know the power supply is a bit under powered for some upgrades so people have to toss a perfectly good power supply for a new bigger one, so maybe offering a bigger power supply as an option would be good).
I look at the MK3 and it doesn't have the bells and whistles of many other printers out there, but at the end of the day when you read the various forums it is much more reliable and even easier to use. That's what I'd like to see Prusa continue to focus on.
RE: MK4 Wishlist
Replace fastenings that involve trial-and-error settings with something more permanent. e.g: The MMU unit and the extruder door are in need of being opened semi-regularly, but involve "losing" the spring strength settings whenever you do this.
Ugh. Even worse that the idler cover likes to flip down due to the angle and throw the bolts and springs out.
The MK3S "door" is no longer a door, it has to be removed. This shouldn't be necessary.
Isn't this only true for the MMU tower? I do agree that it is kind of annoying, but at least they did give us easy access to that bolt so getting the door on and off isn't really difficult.
The MMU2 buffer is an absolute pain to load, especially without 360 degree access to the printer. Change it (rewinding spools seems to be the way to go).
In general I think they need to consider MMU usage other than their ideal farm configuration. Most of us simply don't seem to have the space to set up the buffer and spools in their preferred configuration. The MMU, unfortunately, does not like much variation and I think this leads to much of the pain we go through with the unit.
I saw someone suggest that the MMU kit just include a roll of filament so that we could print auto-rewind spools. I like the core of that idea, but I would extend it to say that they should test different options of how to position the spools (directly behind with no buffer, above/below the printer with and without a buffer, etc...), get the MMU working nicely in all those configurations, and then give us the plans/gcode to build/print spool holders (and storage) that best suit our environment.
Design for basic maintenance, please! Unjamming the extruder requires access from the left, and right of the printer (and now the rear on the Mk3S). Basic operation of the printer the front. Turning it off, the right. Tuning the belts, the rear and the left. Loading the MMU the rear. It doesn't need to, move it all to one or two "access" sides for most common maintenance work (can you tell, my printer is in a corner?).
Yes! I was waiting for my MMU before building an enclosure and I'm glad I didn't as I am now loath to do so. I still would like to get it in an enclosure for various reasons, but I now realize that at least 3 of the 4 sides and probably the top need to be removable so I can easily get everywhere I need to while fiddling with the MMU. That much access complicates/prevents two of my goals in an enclosure (air quality and sound proofing).
Secure your nuts. I have several that I want to glue down, specifically the square nuts that fall out of slots (when I open the MMU2 cover) or jump out their holes (when I remove the blower fan).
This is one of the two reasons I am not a fan of the MK3S upgrade. Taking apart my MK3 extruder almost all the nuts stayed in place. Every time I looked at the nuts for the MK3S they'd fall out. Very annoying.
Design the MMU to have a straighter filament path. The buffer deliberately runs through a "kink" (which will eventually grind through the PTFE on either end of the buffer) then runs through >90 degrees turn to the back of the MMU and the exit of the MMU points "up" for another >90 degree angle to the extruder. It shouldn't be designed with all this unnecessary friction.
I agree here (see my comments about testing/supporting other spool configs). Based on my experience my working theory is that friction/drag in the filament path is a big source of headaches for MMU users. It really doesn't seem to take all that much to make the MMU unhappy.
Switch to e3D nozzle X for a better top layer
That's an easy thing to change when/if you decide you want one. Not sure there is enough value there for it to be the standard nozzle.
Slow down your print farm! The quality of your parts should be showing off what the printer can do, not displaying ringing and require work clearing filament paths, etc.
Good lord YES! I had to reprint my idler door for the MMU tower and even though I ran it off for speed rather than quality it looks so much better than the original (and works unlike the original). In addition to poor visual quality, assembly was difficult in some areas due to obvious issues (e.g. over extrusion).
The cabling inside the einsy box is a mess. Too much spare cable, too many cables, the board is pressured by cables when the door is screwed shut
RE: MK4 Wishlist
Free upgrade.
Nigel
Life is keeping interested and excited by knowledge and new things.
RE: MK4 Wishlist
What do you think about e3d,s toolchanger? I know its going to get the price a bit up but maybe just two heads instead of four and can be as an optional addon. I have the mmu2s but is not really practical for being an actual multimaterial (instead of multicolour).
RE: MK4 Wishlist
What do you think about e3d,s toolchanger? I know its going to get the price a bit up but maybe just two heads instead of four and can be as an optional addon. I have the mmu2s but is not really practical for being an actual multimaterial (instead of multicolour).
The biggest issue with multiple physical tools is that you drop X distance to accommodate them. Either you get a smaller bed so that the "inner" nozzle can get to the edge or the nozzles simply can't reach the opposite edge (e.g. the left nozzle can't go to the right edge).
As far as the MMU and "true" multi-material, have a look at: https://forum.prusa3d.com/forum/original-prusa-i3-mmu2s-mmu2-general-discussion-announcements-and-releases/mmu2-real-multi-material/
RE: MK4 Wishlist
no true with the e3d tool changer, it actually changes docked tools not multiple tools on a single X axis so the work space is not impacted
https://e3d-online.com/blog/2019/04/18/toolchanger-the-update-youve-all-been-waiting-for/
RE: MK4 Wishlist
Ahh. I hadn't seen that yet and thought you were talking about their multi-extrusion heads. That looks pretty cool, but I like Prusa's design better.
The biggest advantages I see for e3d's design are:
- Less/no material waste without a need for a wipe tower.
- Less time waste due to not needing to go through the retraction and wiping.
- Less time waste waiting for temp changes when using different material types.
The downsides I see:
- Parts cost (e.g. 4 extruders rather than 1)
- Power consumption of having 4 extruders heated for the whole print. I expect you could tune gcode to trim that some for cases where a tool is only needed infrequently, but for common tools you are now paying for the electricity to have them heated the whole time.
- More things to break.
- Space. Each tool takes up a fair amount of space.
- Expansion. For all intents and purposes, once you build the printer you are going to be limited to that number of tools.
While the MMU certainly has some room for improvement:
- It's relatively cheap.
- Works on existing printers.
- Someone that is so inclined could make it support more tools relatively cheaply and easily (needs a stepper with a longer worm gear, modelling work for a new housing, and firmware tweaks for additional tools) without having to replace or rebuild the entire printer.
Both clearly have their pros and cons. If I had to divide it up cleanly I'd say that the MMU is better for the hobbyist with something like e3d's Tool Changer is better for the professional.
Definitely an interesting solution to the multi-extruder limitations though.
RE: MK4 Wishlist
The biggest advantages I see for e3d's design are:
- Less/no material waste without a need for a wipe tower.
- Less time waste due to not needing to go through the retraction and wiping.
- Less time waste waiting for temp changes when using different material types.
I believe that you would still need a "Priming Tower" to ensure that the nozzle is primed after a tool change. I agree that this is not as time-consuming as a ram/purge, but it does take a little time. And there's the heating of the tool after a change (inactive tools are usually cooled to prevent ooze).
Peter
Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage…
RE: MK4 Wishlist
Multimaterial related hardware:
- Increase one axis to be slightly larger then the printbed. X-Axis would be suited best for that. Why:
To get serious about multicolor, time and filament-waste are still a thing. Both are greatly reduced with a purge-bucket instead of a purge-tower. However - for this - you need to be able to move the print-head outside of the printer-bed area - which is not possible with the MK3. For further explanation, maybe see also: (not my video)
Just the time-saved would be worth it - not mentioning the reduction of filament waste. - Turn the latch on the MMU2-plugs around, so that cables can actually be removed without tools.
- redesign the "chimney" on the mk3s/mmu2s setup. It works eventually, but its really annoying to set up. Also, loosening/tightening the PTFE-tube holder on the extruder is likely to move your chimney and destroy the setup.
General MK3
- Linear bearing get loud pretty soon 🙁
RE: MK4 Wishlist
The biggest advantages I see for e3d's design are:
- Less/no material waste without a need for a wipe tower.
- Less time waste due to not needing to go through the retraction and wiping.
- Less time waste waiting for temp changes when using different material types.
I believe that you would still need a "Priming Tower" to ensure that the nozzle is primed after a tool change. I agree that this is not as time-consuming as a ram/purge, but it does take a little time. And there's the heating of the tool after a change (inactive tools are usually cooled to prevent ooze).
Yeah I expect there to be some waste in that regard, but far less than a wipe tower that weighs more than the object you printed.
From the linked page and what else I could easily find on their site I don't fully understand how the system works, but if you aren't keeping the tools hot and primed that could still eat a lot of time itself. I would expect that your best option (for speed) would be to keep them hot and just retract the filament back into the cold zone.
RE: MK4 Wishlist
not sure of all the details but you could drop the temp to a holding temp ?100C? between uses so the reheat is faster. so psuedo code:
- end print section
- retract
- drop temp on current head
- raise temp on next head { smart programming could even move this earlier during last seconds of current print}
- do tool change
- quick short purge
- back to printing
one thing e3d is looking at is non-FDM heads ( camera, measurement, ??? ) . they are hoping that others will design compatible heads for interesting new functionality. the mechanical interface design is open-sourced
there are a couple of interviews on YouTube over the last year that go into some of the ideas.
RE: MK4 Wishlist
Ahh. I hadn't seen that yet and thought you were talking about their multi-extrusion heads. That looks pretty cool, but I like Prusa's design better.
The biggest advantages I see for e3d's design are:
- Less/no material waste without a need for a wipe tower.
- Less time waste due to not needing to go through the retraction and wiping.
- Less time waste waiting for temp changes when using different material types.
The downsides I see:
- Parts cost (e.g. 4 extruders rather than 1)
- Power consumption of having 4 extruders heated for the whole print. I expect you could tune gcode to trim that some for cases where a tool is only needed infrequently, but for common tools you are now paying for the electricity to have them heated the whole time.
- More things to break.
- Space. Each tool takes up a fair amount of space.
- Expansion. For all intents and purposes, once you build the printer you are going to be limited to that number of tools.
While the MMU certainly has some room for improvement:
- It's relatively cheap.
- Works on existing printers.
- Someone that is so inclined could make it support more tools relatively cheaply and easily (needs a stepper with a longer worm gear, modelling work for a new housing, and firmware tweaks for additional tools) without having to replace or rebuild the entire printer.
Both clearly have their pros and cons. If I had to divide it up cleanly I'd say that the MMU is better for the hobbyist with something like e3d's Tool Changer is better for the professional.
Definitely an interesting solution to the multi-extruder limitations though.
i agree it will up the cost and power consumption. But i dont see why it should go with all 4 extruders heads. It can be an option to go from two to four. And since in my opinion it will cut the time sufficiently due to less wiping and loading unloading it will , in that aspect at least save some power. Another plus is having two different nozzle sizes. I mean if it can be incorporated in the slicer you could print with a 0.6 nozzle the bulk of the project and use maybe a 0,4 for the top and bottom layers and even the perimeters. The tools may go to a "cooling" temp to avoid oozing but that doesnt mean that they will cool all the way down. And (i may entering the unrealistic wishful thinking realm now ) it may sync reheating to the correct temperature a few seconds before the actual tool change so it saves time. Another plus that i can see is that you will have less problems if you want to experiment printing multiple materials with .25 and .15 nozzles that the mmu and its loading unloading issues.
I know that mmu is a good idea for its price range but sometimes it becomes a headache with all the things that can go wrong in a print. And that is more frequent with soluble materials or particle filled. Not having to worry about how the filament will come out and if the nozzle is contaminated or things like that for me is a bit plus. I've seen the link you send me and it looks extremely interesting and may help with my mmu migranes but another big issue is time. With more complex projects time increases significantly. And i dont know when i'll feel comfortable to live the mmu2 to go on for a 12h turn to 24h print unattended as i used to do with the mk3s.
RE: MK4 Wishlist
i agree it will up the cost and power consumption. But i dont see why it should go with all 4 extruders heads. It can be an option to go from two to four.
Certainly. If you change your mind later and want more, however, that's not going to be a simple thing (unless you "wasted" money and over built the printer to start with).
And since in my opinion it will cut the time sufficiently due to less wiping and loading unloading it will , in that aspect at least save some power.
To some degree, possibly. A lot depends on the size of the print. The longer it has to keep the extruders powered the worse it will be. Cost of filament is also fairly fixed and similar for large regions (e.g. all of the US can get filament via Amazon for the same price) while local regions have different (and sometimes very significant) costs in power. You can also have power cost fluctuations at different times of day too.
Another plus is having two different nozzle sizes. I mean if it can be incorporated in the slicer you could print with a 0.6 nozzle the bulk of the project and use maybe a 0,4 for the top and bottom layers and even the perimeters.
I had not considered that option. Yes that could be a pretty big benefit if you had a cooperating Slicer. Being able to lay down a few .05 layers with a .25 nozzle and then whipping out a .2 layer of infill from a .4 nozzle could be a significant time saver.
The tools may go to a "cooling" temp to avoid oozing but that doesnt mean that they will cool all the way down. And (i may entering the unrealistic wishful thinking realm now ) it may sync reheating to the correct temperature a few seconds before the actual tool change so it saves time.
Yeah that's what I was hinting at with playing games in the gcode. If you knew how long it should take to get a nozzle up to temp you could post process the gcode to raise the temp so there was no delay when it was needed. Similarly the post processing could determine when it was OK to let an extruder cool or to keep it hot.
If you have tools being infrequently used that could save time (waiting for it to heat) and money (by not having to heat it for that 5 minutes of a 90 hour print), but the tools that are used frequently you either waste power keeping them hot or time waiting for them to reheat.
Another plus that i can see is that you will have less problems if you want to experiment printing multiple materials with .25 and .15 nozzles that the mmu and its loading unloading issues.
Yeah I'm still bummed that the MMU is (from the factory) so limited for nozzle sizes and layer heights. Not having to deal with the retraction process/issues would definitely be a boon for using such nozzles.
And i dont know when i'll feel comfortable to live the mmu2 to go on for a 12h turn to 24h print unattended as i used to do with the mk3s.
Yeah the time extension is a killer. I haven't run a 24 hour multi-color print yet (mainly cause I apparently suck with MeshMixer and breaking up what I want to print), but I've run 4 in the 10-12h range that I've let go over night since I got it working reliably. Only one needed my attention the next morning because the spring on one of my spools got bound up and messed up a tool change.
RE: MK4 Wishlist
Multimaterial related hardware:
- Increase one axis to be slightly larger then the printbed. X-Axis would be suited best for that. Why:
To get serious about multicolor, time and filament-waste are still a thing. Both are greatly reduced with a purge-bucket instead of a purge-tower. However - for this - you need to be able to move the print-head outside of the printer-bed area - which is not possible with the MK3. For further explanation, maybe see also: (not my video)
Just the time-saved would be worth it - not mentioning the reduction of filament waste.- Turn the latch on the MMU2-plugs around, so that cables can actually be removed without tools.
- redesign the "chimney" on the mk3s/mmu2s setup. It works eventually, but its really annoying to set up. Also, loosening/tightening the PTFE-tube holder on the extruder is likely to move your chimney and destroy the setup.
General MK3
- Linear bearing get loud pretty soon 🙁
i dont think a purge bucket will actually decrease the waste filament as you still need to purge enough to get a good and clean color change the amount would be the same no matter if you got a purge bucket or tower.
they real key to reduce waste is to use waste objects and print with print beds as full as possible.
for example im printing handles for beer cans with names written on the side, printing one pair with names takes about 10h the handles weight 62g and the purge tower would be 70-80g depending on the purge volumes.
instead of doing that i print another pair of handles without any name on them that i use as purge objects, in this scenario the purge block will weight less than 2g and i get two fully functional beer can handles with only 2h of added printing time.
i could also print another set of handles with names that would leave me with the 70-80g purge block but Used filament to waste filament ratio would also already be much better.
RE: MK4 Wishlist
Your method is obviously valid and best practice today - but not my point.
For a purge tower, you built up a purge-tower every layer - even if there is no need to purge that layer. So often, the first few centimeters are all one color and a complete waste of material and time. This is what you can skip and safe with the purge bucket. Additionally extraction in the bucket can be at maximum filament speed and don't need to be as slow as they are on the purge-block.