Rock Bottom  

Re: Rock Bottom

... but you are also depending on that concrete paver to be very precisely level (which I highly doubt is the case). This could introduce kinks into the machine and give you new problems.

I kept the layer of "cork" (really a spongy cork-a-like material) in my design, to served an auto-truing / anti-twisting function. Sorbothane might also be ideal in that department.

There is also an inherent "spring" action to my screw-down system, which uses a braced bar suspended between the frame and an outboard wood block. I also selected pretty small wing-nuts, to discourage cranking them down hard.

In my tests, I looked at before/after calibration and bed-leveling results, and there was virtually no change. This led me to conclude my approach did not introduce twisting.

Re: Rock Bottom

Clint,
Only issue I see with your idea is the wood, if humidity changes it can warp or change shape by swelling.

Maybe replace with some plastic, something like a cheap cutting board would work. I use that kind of material to make runners for the miter bar slots on my table saw, so jigs can run in the slots. Wood runners would swell so I use the plastic.

I like it, nice way of doing it.

Mike

Re: Rock Bottom

Only issue I see with your idea is the wood, if humidity changes it can warp or change shape by swelling.
... replace with some plastic ...

I didn't think of that!

Actually, my blocks were scrap pieces of Walnut - verrry hard and straight. I'm thinking that some spray poly would have been a good idea, but your plastic idea is preferable.

Actually, printing blocks of the exact size might also be useful. I have printed exact-height blocks for under tables that sit on surfaces of different heights - 100% infill to handle compression load - and they work great.

Thanks Mike!

Re: Rock Bottom and vibration damping and lack cabinet

I'm planning on doing a similar thing, and I have some observations for anyone reading this.

The paver could be doing a couple of things : all of which seem useful, but I'm not sure which is the most important:

1.changing the resonant frequency of the system through mass loading (lowering it with the result perhaps previously excited modes, aren't excited, resulting in smaller overall movements)
2. changing the resultant magnitude of movement for a given force being applied (f=ma), esp. when the y carriage accelerates and the forces are reacted into something which itself will deflect/ deform/ accelerate and move.

3. a structural one : stiffening the mk3 structure : the bottom section has effectively doubled in depth : resulting in less movement for a given force.

4. removing the compliance which is primarily elastic in the stock rubber feet by discarding them. The feet are not even helpfully viscoealstic to dampen out vibrations once the elastic system starts bouncing back and forth.

Note the first two are due to mass, &3 geometry., and 4 elasticity.

I also wonder if where you had your layer of foam, whether you'd inadvertently produced what is known as a "constrained layer damper" resulting in damping any transverse vibrations within the bottom extruded members. This seems a good thing, and I would think a more explicitly viscoelastic material in that layer (like sorbothane) would be better still.

I agree with the notion everything should be made as stiff as possible, but also there should be some damping : so when vibrations inevitably do appear in the system they are quelled rapidly as possible. damping does not have to be just by "sitting it on something squidgy". the frame seems an ideal object for stick on constrained layer dampers ( I intend to look some out and stick them on), so the frame goes "thunk" when hit with a hammer rather than "ting".

I wonder if similarly a bead of visco elastic glue along the outside of each belt would be good too : so any "twang" in a belt rapidly becomes "twunk". I can't seem to find a glue that is vicoelastic, and the bend radii over the pulleys make the sort of CLDs used to dampen the inside of car door panels a bad idea too : as I suspect they'd fall off. If in a heated chamber they might become gooey and drippy too (they tend to be butyl/ bitumen, with some aluminium foil) : poor persons "dynamat". Posh version a 3m product like 2552 tape :

I'm intending to use some kitchen laminate worktop to make a base for my mk3, and sit my prusa/lack enclosure on that (with the location feet pieces screwed to the kitchen worktop)

By using these feet https://www.thingiverse.com/thing:2867374 I can make the printer quickly removable with socket cap head M5 screws and an electric driver for maintenance. Into the underside of the worktop I will sink some M5 insert nuts, so it can go in and come out frequently without the screws breaking up the worktop. I'll mount the powersupply on the worktop just to the right of where the lack cabinet sits (as its meant to be resited outside the hot cabinet for longevity.

I'm also intending to fit a z brace, to stop "waggle" being excited on the Z frame, with https://www.thingiverse.com/thing:2970524 with some 10mm rods (though I don't like the feet as much).

A 75cm wide, 38mm high , 600 deep kitchen chipboard/laminate worktop seems to weigh only 10Kg though, so it may well be a bit on the light side. Darn.... Simple solution : two bits, sittting one on top of the other, with some glue in between.

The whole lot I think could well benefit from sitting on sorbothane feet to isolate from whatever table it's then sitting on, which have to be suitably sized for the weight: 4 feet would be supporting 6.5KG Mk3, lack cabinet 4--5 Kg, worktop 10Kg, or 20 Kg for the lack cabinet. Total : 20-30Kg, (44lbs to 66 lb) i.e. 5--6Kg (11-13lbs) per foot, which means sorbothane domed feet of a diameter of 1.5 inhes, duro 50 as sized here https://www.sorbothane.com/Data/Sites/31/pdfs/product-guides/Sorbothane-SPG.pdf (dimensions in inches)

Hopefully it will be neat tidy, workmanlike, and stiff stiff stiff, and damped damped damped!

Any thoughts ?

Re: Rock Bottom

This a great thread with a couple well throughout solutions. Thanks for sharing. I am going to try the plastic feet first.

Re: Rock Bottom

Wow Jonathan ... really like the feet you found!

One advantage I have come to enjoy in my design (blind luck, not by plan) is the degree of freedom in lining up the Prusa frame on the bed. I did not have to have the feet in an exact position when mounting on the paver ... and precise positioning is *not* possible when drilling through cement (at least the way I did it - with a hand-held hammer drill).

It sounds like your solution to the vibration issue is taking things to the next level. My goal was to be able to print my Breath Flute headjoints (BreathFlute.com) at 100um with no visible vibration artifacts. I have gotten there and now print without artifacts and with high reliability (can't recall the last print failure - Yea!). However, I'm using nGen Lux, which covers up a multitude of sins (at a $price$!!). Printing in Grey PLA (maybe the most telling material / color) does show some very slight artifacts, but I could live with them (except that I need to be able to sanitize the headjoints in a dishwasher - which works with nGen but not PLA).

However, if I went to 50um ... maybe another story!

Thanks for all the info an analysis!!

Re: Rock Bottom

Thank you for you rind words Clint.

I am so grateful that you have paved the way (see what I did there? ho ho) with your meticulous testing, and surviving the pain of a lot of tweaking without knowing how far good you could get. I suspect I shan't be as disciplined as you were, as my time is short, and I will go straight to a base line of lots of things being done (e.g. filament off the printer, in a dry box, on a spool holder with 608 bearings), which we now know are beneficial, even if some of the things are marginally beneficial.

Your breathflute project looks amazing. I haven't had a chance to look in any detail (though I will), as I learnt the flute as a kid, so this looks right up my street. I sing (I'm a choral tenor), and its been too long since I've been an instrumentalist.

It's very interesting that you had a need for clean prints, which drove you to this optimisation. I too am driven by a need to produce production low volume initial production components for a commercial electro mechanical tool i am trying to commercialise and bring to market.

It will be a premium device, and I don't want my finished quality to be amateur looking.

Apart from this long thread about "rock bottom", what other things did you do to maximise quality of output? have you a list of machine modifications you've made for example? Indeed are there ones you've tried but you concluded there was no point? Your analytic approach with true before and after testing with test pieces is actually very rare, especially in the arena of machine modification to my surprise, and it is difficult to sort out what is useful to do, and what isn't. I intend to update my extruder to the latest venturi design (R3?). I have replaced the motor mount for the y axis with a much stiffer one, to remove flex, and with an updated belt tensioner on the other end.

I intend in the next few weeks when I upgrade the etxtruder to upgrade my x axis components to the "zaribo" ones https://www.thingiverse.com/thing:2964026
i am thinking of reprinting the z axis tops and bottoms with greater numbers of perimeters to stiffen them up.

The quality of your prints in the breathflute documentation look spectacular.

Have you concluded there are specific things that are important to get right in the filament supply path for example (e..g the nature of the spool holder, reverse bowden details if you have such a set up) etc... I currently have the filament in a box with desiccant from which it is fed.

Lastly : what do you think are the things that should be optimised when dialing in a filament to maximise quality?

Many people on forums seem to work non analytically-whereas I suspect you've worked what really should be calibrated.

At the moment, I go through a sequence of :
1.live z height for good first layer bed bonding
2. extrusion multiplier calibration (by measuring the width of a two wall perimeter on an open cube)
2. reslicing first layer height test piece with correct extrusion multiplier, and check if that needs to be tweaked. Check that this first layer looks uniform across the bed (in case of any manual mesh level adjustments I should make).

(I haven't yet, but will soon calibrate my linearity correction, a once only thing of course ...)

For me that's about it at the moment : but I know some people advocate going on to doing:
temperature towers,
I've also read about people doing speed towers to find if there are speeds they should avoid for perimeters
and I can imagine that there could be other "towers" to optimise some other slicer parameter... though don't know which would be worth doing.

Whilst working through these tests, I would endeavour to slice each with the input of the previous optimised parameter, as the parameters seem not to be wholly independent.

All the best

Jonathan

Re: Rock Bottom

Thanks Jonathan!

I'm a software guy, so I stay away from complex hardware mods. I didn't even assemble my own Mk3 kit - I got it assembled. My hardware mods are more of the "bolt this to that with a hammer drill" ...

However, by far the most important thing I've done is work on slicer settings. I did a real survey study of the info out there, developed a detailed document on slicer settings, poured over the recommendations of the manufacturer of the material used by the filament fabricator (not easy to find), and developed my slicer settings from there. This took days. However, out of the gun, I was rewarded with great prints that needed a smidge of tweaking.

Downside is that it's all specific to Simply3D. However, the concepts should translate to other slicers ...

All of that work is available on the Breath Flute Download page ... in the Additional Resources section. The Simplify3D Settings Manual is the document, and is my core profile for the Mk3 with PLA.

Here is the related thread: https://shop.prusa3d.com/forum/general-discussion-announcements-and-releases-f61/simplify3d-settings-and-profiles-t15905.html

Re: Rock Bottom

Spectacular work Clint. Thank you.

I've just had a skim through, and you've collected so much useful conceptual info.

I will find it useful going forward, as I like to understand what I'm doing, yet understanding the effects of parameters in the slicer is opaque at best, it is generally a black box. It is telling and instructive that following the advice of the filament manufacturer you found especially useful, although time consuming to enter into your preferred slicer, and worth copying as one specific (succeful) ploy you used.

I am (currently) using slic3r as fairly new to the game, and the i3 mk3 obviously has guided me in that direction.

Although I have concentrated on hardware issues so far: this is mainly because I suspect much software fiddling lies ahead, and I wanted to dial out any obvious hardware issues before I started in earnest, in case I was optimising for a sub optimally (physically) set up printer....