|
Re: Trying to figure this all out... January 19, 2014 01:10PM |
Go with Mendel90, my experiences with Prusa v1 was so so so bad, and the reputation of Mendel90 is so good (made by nophead) i wouldn't hesitate if shelling out that kind of money.
That being said - i have a makerfarm version Prusa i3 coming soon - mainly because i got it cheap, and hoping it's a usable machine.
With Prusa V1, which is *very* similar to i2, basicly i2 is an update to v1, i've had nothing but trouble, constant trouble. It was rare to have it working for more than a day without issues. 90%+ of prints are failed in one way or another, not always because of the printer, but most of them.
Important calibration such as Z-endstop level would walk away at times after every and each use - and this could be even upto millimeters at times! Seriously: I've had the bed just right the 10 prints before, and then all of sudden 3mm gap between bed and nozzle, next attempt it's where it should again.
And that's the least of the worries, on the v1 many parts are simply too weak of a construction, they've cracked under usage, some of the parts has been strengthened for i2 tho.
There is fundamental flaws in the design as well, from ground up, to the ways they have been implemented, for example, the Y axis belt has barely more than 90 degree hold on the pulley -> it will slip rather easily unless your belts are really tight. Thing is, you can't get them really tight with Prusa, there is *no* tensioners what so ever in the design. And one should have these belts tight to avoid backlash and missed steps ...
Thing is - the v1 worked *so* poorly, that i never got a full set of spare parts printed without one kind of a failure or another, some of that was due to the experimentaiton, some was because i needed to get actual work done with the machine too, rather than constantly trying to get spare parts done before it breaks again.
I'm now at the point i will probably throw the frame away in trash, just take off all the precious parts. Shame tho really, i wanted take a baseball bat on it, but it wouldn't be very entertaining at all after having to disassemble it to take precious parts away
Yea, it works *that* badly.
EDIT:
You will start blaming the workmanship on my machine in any case: Mechanics and electronics is something we know very well with whom we built it, we used professional tools, and built it in a professional setting, everything was true, properly torqued, properly aligned. In the end the assembly accuracy didn't seem to matter, because after a while pattern emerged: It always breaks before it gets any work done, and we started to "just slam it together", until every single issue that arose was solved, rinse and repeat style, and eventually the frame wasn't square probably neither because the vibrations would get the nuts loose, being broken so often couldn't care less anymore, and just put them back in without even checking if it measured still up as long as the X axis was true with bed (Y), and bed was traveling straight on the frame, and the frame was sitting straight on the desk.
Broken printed parts list is something like:
Wade's small + big gear
Several bar clamps, each type
Z-motor mount
X-carriage
Several belt clamps, both types
Several endstop holders OOHH which reminds me, your X endstop will never stay, since rotation of those smooth rods is not stopped by any means, you need to get it pinched against something for it to stay in that place for any period of time. Which reminds me that your Z axis can jump off if there is pull upwards, or your Z-endstop has failed again.
Good luck trying to get the bed stay leveled & tensioned with the default vitamins ... Makes one feel like, what, no one heard of nylocs before? Would have removed possibility of so many issues from get go
and oh yeah, to get the full Y travel you need to mount bed so high that you loose significant amount of Z *and* your Y belt will have kinda high angle (might not matter, might matter in your setup). That's because you need to still mount it somehow *and* tension the leveling springs (which should be double springed to begin with!) you need quite long screws for that, and they will hit your vertex threaded rods unless you battle with sweat, blood and tears to get that high tension with short screw *or* make yourself studs and raise it all so high that the bed can move "beyond" the frame partially, and the bolts wont hit the threaded rods, that's the only way to get even close to full Y travel. With proper Y carriage for the bed, you will have hard time to find good place for end stop holder to get full travel - with the aluminium carriage there is no way to find such a place.
You will also need to upgrade several portions of it, and i2 won't be stable on X direction because there is 0 support on X direction.
Tbh, the Prusa V1, and therefore i2, is one of the crappiest examples of mechanical engineering i've ever seen. It's clear Prusa is not a mechanical engineer, he has done many great things for the reprap community, but he would have saved so much pain if he would have even consulted a real mechanical engineer during the design. And what's with all that trying to save plastic crap, making things way tooo thin for them to be strong enough for the use. Plastic is cheap, all it *Really* costs to add that extra millimeter to the bar clamps, or x-carriage belt clamp area is just a tiny bit of printing time, which could be reduced significantly by driving high speeds on a well designed structured which can support those speeds.
Without Prusa however, the reprap community as a whole would not have grown so fast probably - but mechanical engineer he is not.
Edited 3 time(s). Last edit at 01/19/2014 01:31PM by PulsedMedia.
That being said - i have a makerfarm version Prusa i3 coming soon - mainly because i got it cheap, and hoping it's a usable machine.
With Prusa V1, which is *very* similar to i2, basicly i2 is an update to v1, i've had nothing but trouble, constant trouble. It was rare to have it working for more than a day without issues. 90%+ of prints are failed in one way or another, not always because of the printer, but most of them.
Important calibration such as Z-endstop level would walk away at times after every and each use - and this could be even upto millimeters at times! Seriously: I've had the bed just right the 10 prints before, and then all of sudden 3mm gap between bed and nozzle, next attempt it's where it should again.
And that's the least of the worries, on the v1 many parts are simply too weak of a construction, they've cracked under usage, some of the parts has been strengthened for i2 tho.
There is fundamental flaws in the design as well, from ground up, to the ways they have been implemented, for example, the Y axis belt has barely more than 90 degree hold on the pulley -> it will slip rather easily unless your belts are really tight. Thing is, you can't get them really tight with Prusa, there is *no* tensioners what so ever in the design. And one should have these belts tight to avoid backlash and missed steps ...
Thing is - the v1 worked *so* poorly, that i never got a full set of spare parts printed without one kind of a failure or another, some of that was due to the experimentaiton, some was because i needed to get actual work done with the machine too, rather than constantly trying to get spare parts done before it breaks again.
I'm now at the point i will probably throw the frame away in trash, just take off all the precious parts. Shame tho really, i wanted take a baseball bat on it, but it wouldn't be very entertaining at all after having to disassemble it to take precious parts away
Yea, it works *that* badly.
EDIT:
You will start blaming the workmanship on my machine in any case: Mechanics and electronics is something we know very well with whom we built it, we used professional tools, and built it in a professional setting, everything was true, properly torqued, properly aligned. In the end the assembly accuracy didn't seem to matter, because after a while pattern emerged: It always breaks before it gets any work done, and we started to "just slam it together", until every single issue that arose was solved, rinse and repeat style, and eventually the frame wasn't square probably neither because the vibrations would get the nuts loose, being broken so often couldn't care less anymore, and just put them back in without even checking if it measured still up as long as the X axis was true with bed (Y), and bed was traveling straight on the frame, and the frame was sitting straight on the desk.
Broken printed parts list is something like:
Wade's small + big gear
Several bar clamps, each type
Z-motor mount
X-carriage
Several belt clamps, both types
Several endstop holders OOHH which reminds me, your X endstop will never stay, since rotation of those smooth rods is not stopped by any means, you need to get it pinched against something for it to stay in that place for any period of time. Which reminds me that your Z axis can jump off if there is pull upwards, or your Z-endstop has failed again.
Good luck trying to get the bed stay leveled & tensioned with the default vitamins ... Makes one feel like, what, no one heard of nylocs before? Would have removed possibility of so many issues from get go
and oh yeah, to get the full Y travel you need to mount bed so high that you loose significant amount of Z *and* your Y belt will have kinda high angle (might not matter, might matter in your setup). That's because you need to still mount it somehow *and* tension the leveling springs (which should be double springed to begin with!) you need quite long screws for that, and they will hit your vertex threaded rods unless you battle with sweat, blood and tears to get that high tension with short screw *or* make yourself studs and raise it all so high that the bed can move "beyond" the frame partially, and the bolts wont hit the threaded rods, that's the only way to get even close to full Y travel. With proper Y carriage for the bed, you will have hard time to find good place for end stop holder to get full travel - with the aluminium carriage there is no way to find such a place.
You will also need to upgrade several portions of it, and i2 won't be stable on X direction because there is 0 support on X direction.
Tbh, the Prusa V1, and therefore i2, is one of the crappiest examples of mechanical engineering i've ever seen. It's clear Prusa is not a mechanical engineer, he has done many great things for the reprap community, but he would have saved so much pain if he would have even consulted a real mechanical engineer during the design. And what's with all that trying to save plastic crap, making things way tooo thin for them to be strong enough for the use. Plastic is cheap, all it *Really* costs to add that extra millimeter to the bar clamps, or x-carriage belt clamp area is just a tiny bit of printing time, which could be reduced significantly by driving high speeds on a well designed structured which can support those speeds.
Without Prusa however, the reprap community as a whole would not have grown so fast probably - but mechanical engineer he is not.
Edited 3 time(s). Last edit at 01/19/2014 01:31PM by PulsedMedia.
|
Re: Trying to figure this all out... January 19, 2014 02:01PM |
As a solution, all he would have needed to do to make that machine *much much* better would have been to angle the main vertices so that they support both the X and Y directions, just something like even 5 degrees would have done MANY great things, then make all the parts taking much forces, such as the X carriage belt clamp area a bit thicker, and infact, the V1 has spot on the belt clamps *designed* to make tightening the belt a maddening endeavour, a slot through which the belt slips unless bent over double size, and way overtorquing it, easy fix, takes just couple mints to draw up better clamps, but still. The carriage side is way too weak - it's the portion which takes ALL the lateral forces, and it's the *weakest* part of the carriage.
The bar clamps needs a bit extra material at the threaded rod side, and slotting it for easy changing without disassembling whole machine, thing is with certain slicers your retract will be so bad you get barely any material on the end side which is shaped like an arrow (wrong version of slic3r, duh! Wish cura had the finetune options of slic3r and cura plotter would work right on my laptop with it's old integrated graphics with whatever opengl windows came with...), further it would need thinning on the bar side top to make it a bit more flexible, that's the spot where it most often fails, so experimentation, trying thinner, and trying thicker and finding out which gets the bending right.
Endstop holders should be mounted on the frame, and come with somekind of easy finetuning option, a screw and bar in a slot would have been all that is required really.
Y motor mount is total crap too, lessening the number of teeth hitting the pulley, all it needs is lifting the motor to the 608zz level to partly fix it, proper method would have been to use 1-2x 608zz to do "serpentine" to ensure as much contact as possible - one can integrate finetune tensioning on this very same bearing mount too! Tight belt *finally*
X motor mount works quite well, but same thing here, could ensure much mroe grip on pulley PLUS tensioning by adding a 608 screw adjusted on the top belt.
X carriage would preferrably need some belt tensioner too, because eventually that belt vibrates baaad, tho the X motor mount tensioner might solve the issue this causes: Slipping from pulley.
Underneat Y carriage, mounted in the frame should be dual 608 belt tensioning system *if* the Y motor bracket changes don't solve tensioning + vibration issue.
Z ... Well for one the printed couplers are crap - they will never hold. Further adding stabilizer 608 on the bottom of the threaded rod is kinda a must (here you can mount Z endstop microadjustabel too!) - the threaded rods go all around without them, i ahven't measured this but juding from the angles, the whole X carriage was wobbling as per which layer it was until adding 608s to the bottom of threaded rods.
Plus the Z-motors should be on bottom, as it is now it's kind of top heavy --- not too good of a thing for something which has great forces and kilograms moving at relatively high speeds and accelerations.
X smooth rods are not really held in place with anything but big wishes, amazingly enough, it works sufficiently. However, worst place ever to mount X endstop because your smooth rods will rotate as mentioned earlier
Vertexes are too thin: There's barely material on the sides of the threaded rods, 0.5-1mm just for safety's sake (tho i haven't managed to break these apart) OR adding a whole 5mm and dropping infill during print -> Stronger, yet potentially even less plastic used and faster print times. Any mechanical engineer should know that a lot of strength comes from circumference/distance... This is called leverage Don't know the real english word for this about calculating forces and strengths.
The 2 per side threaded rods going upwards to form triangle, for strength i would make them M10 or M12, on a angled triangle shape for the ultimate in strength. Cost difference here is totally negligible.
Integrated vibration removal: Rubbery washers (can be printed, just means it needs to be shaped and tensioned right) on certain key spots to remove part of vibrations BUT not introducing too much flex, just to take the highest frequencies out.
Y axis smooth rods should be wider apart - see my mentioning about forces & strengths above - it would make it slightly more stable, sometimes (or often, depends on calibration) your nozzle will hit tiny pieces of plastic while doing fast travel, this would better smooth them out, while most is done by the springs.
Same goes for X axis - but cannot be helped without doing better Z axis, beause the Z axis smooth rods and threaded rods are all lined up, your X carriage will rotate by the amoung of flex for the force & the tolerances (which in my V1 were *huge* for the threaded rod + nuts), oddly enough, this didn't seem to affect the quality of prints, maybe just luck.
What for Prusa V1, therefore i2 too as well is good for: If you like tinkering, repairing machinery etc. not doing the actual work, it's brilliant! You get to tinker a lot with it.
It's extreme strength on the Y axis, it's incredibly strong in that direction, that's the benefit of the triangle shape. Shame all the strenght benefit is ruined by the equally bad Y motor mount.
You probably think that i'm being ruthless on the design - but i really am not.
It's a bad design, it's a very bad design. Makes a machine of nightmares.
Only way it works if you run it at extreme slow speeds, with very low accelerations, with very low forces, then it might just be durable enough to get a good day's print out of it. But when you know a good frame could sustain 5-6x the speeds, with 1/10th of the maintenance required, with reproducibility 10 times greater, with steady good prints ... It becomes maddening machine, when you know the technology, the parts you have invested in, could do so much greater but the bad design of the machien is holding you so badly back.
Yes, now i was being a bit ruthless - but that's my opinion of using one for a while. I'm just glad i will receive both a Makerbot, and an MakerFarm Prusa i3 for comparison soon - and i'm really hoping i start to get some actual work done with these machines, and i can settle on for a design, as we need probably 10 of these machines by late spring running 24/7
The bar clamps needs a bit extra material at the threaded rod side, and slotting it for easy changing without disassembling whole machine, thing is with certain slicers your retract will be so bad you get barely any material on the end side which is shaped like an arrow (wrong version of slic3r, duh! Wish cura had the finetune options of slic3r and cura plotter would work right on my laptop with it's old integrated graphics with whatever opengl windows came with...), further it would need thinning on the bar side top to make it a bit more flexible, that's the spot where it most often fails, so experimentation, trying thinner, and trying thicker and finding out which gets the bending right.
Endstop holders should be mounted on the frame, and come with somekind of easy finetuning option, a screw and bar in a slot would have been all that is required really.
Y motor mount is total crap too, lessening the number of teeth hitting the pulley, all it needs is lifting the motor to the 608zz level to partly fix it, proper method would have been to use 1-2x 608zz to do "serpentine" to ensure as much contact as possible - one can integrate finetune tensioning on this very same bearing mount too! Tight belt *finally*
X motor mount works quite well, but same thing here, could ensure much mroe grip on pulley PLUS tensioning by adding a 608 screw adjusted on the top belt.
X carriage would preferrably need some belt tensioner too, because eventually that belt vibrates baaad, tho the X motor mount tensioner might solve the issue this causes: Slipping from pulley.
Underneat Y carriage, mounted in the frame should be dual 608 belt tensioning system *if* the Y motor bracket changes don't solve tensioning + vibration issue.
Z ... Well for one the printed couplers are crap - they will never hold. Further adding stabilizer 608 on the bottom of the threaded rod is kinda a must (here you can mount Z endstop microadjustabel too!) - the threaded rods go all around without them, i ahven't measured this but juding from the angles, the whole X carriage was wobbling as per which layer it was until adding 608s to the bottom of threaded rods.
Plus the Z-motors should be on bottom, as it is now it's kind of top heavy --- not too good of a thing for something which has great forces and kilograms moving at relatively high speeds and accelerations.
X smooth rods are not really held in place with anything but big wishes, amazingly enough, it works sufficiently. However, worst place ever to mount X endstop because your smooth rods will rotate as mentioned earlier
Vertexes are too thin: There's barely material on the sides of the threaded rods, 0.5-1mm just for safety's sake (tho i haven't managed to break these apart) OR adding a whole 5mm and dropping infill during print -> Stronger, yet potentially even less plastic used and faster print times. Any mechanical engineer should know that a lot of strength comes from circumference/distance... This is called leverage
Don't know the real english word for this about calculating forces and strengths.The 2 per side threaded rods going upwards to form triangle, for strength i would make them M10 or M12, on a angled triangle shape for the ultimate in strength. Cost difference here is totally negligible.
Integrated vibration removal: Rubbery washers (can be printed, just means it needs to be shaped and tensioned right) on certain key spots to remove part of vibrations BUT not introducing too much flex, just to take the highest frequencies out.
Y axis smooth rods should be wider apart - see my mentioning about forces & strengths above - it would make it slightly more stable, sometimes (or often, depends on calibration) your nozzle will hit tiny pieces of plastic while doing fast travel, this would better smooth them out, while most is done by the springs.
Same goes for X axis - but cannot be helped without doing better Z axis, beause the Z axis smooth rods and threaded rods are all lined up, your X carriage will rotate by the amoung of flex for the force & the tolerances (which in my V1 were *huge* for the threaded rod + nuts), oddly enough, this didn't seem to affect the quality of prints, maybe just luck.
What for Prusa V1, therefore i2 too as well is good for: If you like tinkering, repairing machinery etc. not doing the actual work, it's brilliant! You get to tinker a lot with it.
It's extreme strength on the Y axis, it's incredibly strong in that direction, that's the benefit of the triangle shape. Shame all the strenght benefit is ruined by the equally bad Y motor mount.
You probably think that i'm being ruthless on the design - but i really am not.
It's a bad design, it's a very bad design. Makes a machine of nightmares.
Only way it works if you run it at extreme slow speeds, with very low accelerations, with very low forces, then it might just be durable enough to get a good day's print out of it. But when you know a good frame could sustain 5-6x the speeds, with 1/10th of the maintenance required, with reproducibility 10 times greater, with steady good prints ... It becomes maddening machine, when you know the technology, the parts you have invested in, could do so much greater but the bad design of the machien is holding you so badly back.
Yes, now i was being a bit ruthless - but that's my opinion of using one for a while. I'm just glad i will receive both a Makerbot, and an MakerFarm Prusa i3 for comparison soon - and i'm really hoping i start to get some actual work done with these machines, and i can settle on for a design, as we need probably 10 of these machines by late spring running 24/7
Sorry, only registered users may post in this forum.