Smoothie M2

[innkeeper]

has anyone tried to address it?

Yup...
http://softsolder.com/2013/05/31/makerg ... ilization/



The springs have zero damping and the platform weighs the better part of a kilogram, so there's a distinct tendency to flop around atop those springs (before the screw heads reseat themselves) under high acceleration. The silicone pads provide a rigid mount with good damping: all those problems Go Away.
.

Ed, i dont know how / why i missed this message before...i'll give the dampers a go...and could aslo be assisted with simply using stiffer sprngss.
but my major complaint is the actual virtical play from the single linear rail. .5mm is a lot...

frankly a solidly built printer as this is ment to be should't have a hot build plate that has so much play that can flop aorund like a fish on the deck of a boat.
its a major deign flaw with using a single linear rail on the Y axis..

I've been looking into how somting like a redi rail could be retrofited, but, not easily done without a completely new z stage platform..
besides... if i'm going to bother getting a new z stage cut, one could also just ad a second linear rail.... which mght be the best aproach.

Plus this all would probably require reloacting the electronics outside the unit, or reducing the build height.... thats mabey not that its much of an issue as to ge the proper z stepper in there your going to also reduce build height anyway.

[innkeeper]

Rick, are there cad files available for:

stanless steel spider mounting plate
and the aluminum
z stage?

id like to see what i can do about adding a second rail.

[ednisley]

flop aorund like a fish on the deck of a boat.

Except that it really doesn't do that.

When you apply a vertical force to the edge of the platform, there's a tremendous torque on the rail. You can easily apply 2 kg = 20 N of force with a fingertip that's 100 mm from the rail; call it 2 N·m. The resulting 0.5 mm of deflection amounts to 0.005° of angular deflection.

However, during printing, the only deflecting torque comes from the offset of the platform's center of gravity relative to the rail. In round numbers, the moving part of the platform weighs a bit over 1 kg, it's offset from the slide by barely 30 mm, and it accelerates at 1/3 G, so the deflecting torque is maybe 0.1 N·m. Now we're down to 0.00025° and maybe 0.025 mm.

I may have misrepresented the numbers, so run 'em yourself. I think you'll come to much the same conclusion, though.

It's trivially easy to design a rock-solid 3D printer that sells for $200 k. Selling one for $20 k requires some effort. Selling a $2 k printer means trading off components that seem nice for ones that actually work.

The M2 has had some implementation flaws, but not many and nothing terrible. A floppy platform isn't one of those; the silicone pads I added made me feel good, but didn't really change much of anything.

[innkeeper]

flop aorund like a fish on the deck of a boat.

Except that it really doesn't do that.

Well that's not true for me...maybe my linear Bering bad???

yes sure you can make things better by slowing acceleration and print speeds.. but not the goal really

I'm not sure the silicone spacers will help me...I can;t actually push the plate down against the springs... now I originally bought this built, so I am unsure if they backed the screw with a nylon lock nut or something so it can't move vertically .. but I have zero play on the spring, I just cant compress manually. I've never had that part apart to see how they accomplished that...my other bot I can compress it... so I know what your getting at, and on that bot I put much heavier springs to solve the same issue.

I actually measured the play the bearing has with a dial indicator on the end of the arm and it really is .5mm at the end of the arm, and it is mostly the slop in the linear bearing.
you can watch the linear bearing move when you do this and feel it as you push up, certainly with less force then the movement of the carriage would create..., I've considered adding mass tot he ends of the carriage to reduce the effect.

the distance from the linear rail to the end edge of the platform is about 120mm..and the rail being 12mm wide so if my math is right, that equates to about +/- .025 mm of play at the linear bearing on each side...that doesn't seem to unreasonable a spec. ok so figure some of that is deflection. but really it is academic, ... if your seeing the same play then its not a bad bearing . What I am thinking is adding second rail closer to the end of the spider, it will reduce both any play and deflection. as your maybe 30mm from the edge then. now your down to about +- .05 mm of movement at the end.

that would require new parts to be milled more mass on the z axsis ... and I am willing see what it would take to do that if it is an inherent design issue and not a bad bearing.. I've already asked for the cad drawings.

if you don't have the same play then maybe my linear Bering is out of spec, well id not be surprised considering everything else that's been a problem with this m2.

it really does flop around (and yes all the screws are tight ), especially with those tight infill's. (typically run 80mm/s) and yea you are right...I can make this better by lowering down the initial speeds but I am looking to actually fix it.

If your curious, you close enough to me you could come see it for yourself if you like.

Who knows...maybe I just have a m2 built on a Friday as this m2 has been a problem since day one... but still have an affinity for it.. go figure..

• had to replace the electronics, as they were bad from day one, the hot ends kept jamming ... and after having MG run me around with pid's, and moving around the temp sensor on the head, replacing temp sensor...etc.. and actually trying 2 different hot ends on my own before giving up and shelving it... I find out recently the Rambo board is not reporting the temperature changes accurately and consistently. someplace during the print the hot end would go under temp and jams......(I didn't 100% confirm this till I took it off the shelf after sitting around unusable for 2 years and decided to use it as a donor chassis for smoothiewear conversion)
• Now going for a new pinion gear on a bot / stepper that for the with probably 200 hours on it, and as I understand it, the worn pinion is from a bad original run of steppers that used soft metal on the pinion..(ok I am not alone on this one)
• z stepper from this generation was changed with a higher current model post this production.. so I have the older z stepper. (new one which is still under spec'ed, and also ran over current) (not alone on this one either) going with your mod as I am adding auto level and ...

I don't agree with you on what you can accomplish for 2k... have a bot with dual extruders I paid ~2k for in 2012 sitting next to the m2 that's rock solid been running the heck out of it for 3 years and in the last two I have been running it at 100-120mm/s most of the time with none of these kinds of issues and very good quality prints, run it at 80mm/s for fine quality and it's chassis is made of wood for crying out loud. and yes it has some mods, but probably less then 200 usd worth. I actually got the m2 as an "upgrade" to it... so I don't buy that a 2k bot can not be solid....heck that's this bots claim to fame, I'm just trying to get there with it am I using the other bot as a benchmark for the m2.... sure....

Its also a hobby, I don't make a living out of running my printer, its strictly a side thing ...... so as such I am willing to work on it and make it better.....hobbies are supposed to be fun right

ok soap box off ...
sorry for the long post ...

please let me know if you think the linear bearing is out of spec.

Thanks Ed.

[Tim]

I think I followed that conversation properly, in which case I think you really may have a bad bearing, because I have no play at all like that on the Y axis rail.

If you have no play on the bed springs, check the screws. The screws that are called out in the instructions are about 2mm too long for the job, in my opinion, because if you tighten them down too much, the ends of the screws will be hitting the heated bed plate.

[jimc]

ok i have 4 m2's sitting here. 2 are older. about 2.5 years and 2 are basically brand new. if you take the corner of the bed and hold it gently between 2 fingers you will always be able to feel play up and down. there is just too much there to give. springs, bearings, flexing and it all accumulates. the up down wiggle isnt the issue. take that same corner of the bed and gently wiggle it side to side like your rotating the bed as if the bearing was the center point. you will feel an ever so slight bit of play. it should not be much. the older printers i have here have more play but they are the ones i print on daily and give me fantastic prints. the one new one that is 3-4 mos old has play as well but not quite as much as the older machines. i can see they are using a different brand bearing. the brand new machine i just got last week has no play at all. that one hasnt printed anything yet though. the bearings are still probably packed with grease. my older machines have prob a couple few thousand hours on them. in all that time the slop in the bearings have never gotten worse. i find it hard to believe you would have a bad bearing but stranger things have happened i guess. i guess what im getting at is unless the play is really severe i dont think you will get bad prints.

[innkeeper]

thanks, that's good information. that also helps put things in perspective.
generally I don't have bad prints, in fact I get really god prints... I can push it / aggravate it to make bad prints.
but I guess you can push anything to not perform well.
but in general, if I keep the acceleration and / or speed within reason, it is fine.
It just the first limiting factor when I do start pushing it.
who makes the new bearings?

[innkeeper]

If you have no play on the bed springs, check the screws. The screws that are called out in the instructions are about 2mm too long for the job, in my opinion, because if you tighten them down too much, the ends of the screws will be hitting the heated bed plate.

wow I practically leaped up and checked the printer when I read that lol. the screws do not stick out of the top of the spider... now I really have myself curious about why there's no movement.. i'll have to unscrew them and see what the deal is. it could just be that I have them so low now, there not much movement possible.

interesting that you have no play, comparing it to what jim posted... how old is your machine?

[jimc]

new bearings are marked cpc. not sure if that a mfg or not. for your bed springs you generally want to tighten the screws all the way down until its squished down to the nylon spacer then back off 1 turn and level the bed from that point. this way the springs are compressed as much as possible and you still have slight movement. you dont want the screws out 5-6 turns. then everything will be floppy.

[innkeeper]

CPC is the same brand as mine.
I'll do that for the springs since I am going to take them apart anyway to have a look see...... id say I am probably pretty close to that now, but not from actually following that. just out of pure experience with other bots and not wanting it loose and bouncy ... though frankly I didn't try and see if it even could compress until today.. I just assumed it could, and was surprised it couldn't. usually, its good to have a little compression left so when you crash the head into the plate it allows some give. .. not that its ever happened to me