Does a wobbly table RUIN print quality?

[@SpacePotatoMan]

Important to know that the real reason Angus's prints are not affected by being upside down is because he's in Australia and two upside downs cancel out

[@soumyadarshansahooxa9094]

😂

[@danieldukey]

I'd like to see the same test with a machine that doesnt rely on Input Shaping to see if that's the secret sauce or if it really in general doesnt matter which surface to print on...

[@ynonzohar4199]

Yap my thought exactly. I have a very basic printer and it doesn’t have any calibration/compensating for that at all

[@berribeans6593]

​@@ynonzohar4199agreed would love to see this!

[@cozmo4694]

A bit of a flawed testing technique, he should have used like an ender 3 or prusa mk3 as they are kind of the "golden standard" for testing

[@Mark_5150]

This is a really easy test to replicate. test it with your own printer

[@Validole]

​@@cozmo4694not flawed, just testing something else. In this case: "in the age of competent input shaping, does it matter what your printer is sitting on?" Although, could have spelled it out as such.

[@davidhorizon8401]

It would be interesting to see this on a printer that didn't do a compensation calibration.

[@ChrisModjeska]

Just knowing how input shaping works shouldn't. Not sure how this is surprising, you can hang your 3D printer from the ceiling and it likely won't affect the quality unless there's some specific resonance created. I spent a lot of money on creating super rigid and heavy surfaces for my printers, and I've since moved them all into a rickety folding table, and the results are better. More rigid is good if you have the dampening to counter resonances. Its difficult to intuit because a lot of other CNC stuff deals with much slower speeds and accelerations, but with much greater mass and therefore their resonance frequencies are lower and more catastrophic. I've often thought about creating a cast machine frame (mixture of resin and a very hard filler like granite) like they do for milling machines, but I'm not sure it wouldn't make artifacts worse based on what I've seen experimentally.

[@MetalheadAndNerd]

A closed system is a closed system.

[@AnIdiotwithaSubaru]

The funny thing is that that "bad" benchy is what most peoples decent benchy looks like

[@MattThePrintingNerd]

Good Job on that video! I build my THE 100 3d printer around a wobbling pla frame to even reduce vibration. It's nice to see that more and more people realise that wobbling does not affect your printing quality and also that it could improve your printing quality especially when printing at highest speeds

[@petermoore9504]

Kudos for reading the comments and addressing them in a very positive and informative manner. Cheers

[@storm14k]

So I have for a little over a year stopped on my 3D printing learning because I was worried about only having wobbly tables. I lost my setup after a move where it was on a rock solid table. I finally just now stared really thinking about it physically. I decided to search and here I am. Thank you for this video! The time I wasted overthinking and not just unpacking and going for it.

[@timbrelypearsley3888]

I have been stressing about this for my printer that is arriving in the near future. I didn’t want to have to go out physically shake testing tables and desks or modifying them for stability. This is a relief.

[@rcmaniac25]

All hail the wobbly table! I remember when Tom Sanladerer made an overly complex bouncy box to hold a printer to see if it would reduce ringing and... it didn't really do anything, maybe even made it a little worse IIRC. But like you said, this is probably the first time being done on a high speed printer (and certainly first I've seen for a cantilever and input-shapping based printer). Still interesting to see the results. You did the same prints on 3 different env. Would be interested in the extra env. of the concrete slab + checking noise (you commented it was quietest on concrete, but other audio comments seemed to be around calibration instead of the print itself). My mind now wants to see something extra: Given basically the whole Klipper community, and Bambu, have the ability to rerun the calibration for different env. But the other juggernaut in the space, Prusa, has decided... "nah" and does the input shaping calibration in factory. So, using the same A1 mini, if you calibrated it once (say, the concrete floor) and then ran it without calibration in all the same env/test... what are the results? Is the quality the same and all this calibration in each env. a waste of time... or is quality worse, Prusa messed up and I should be demanding an accelerometer from them. If fixed calibration is fine/good, it could open a whole new area where you can take your Marlin-based printer (which actually supports input shaper), have one person do the calibration, and then share those settings around to the rest of the community. Literally turning basically every printer out there, capable of running the latest Marlin, and turning it into a fast printer without needing to Klipperize or mod the printer.

[@JanMuell42]

> I remember when Tom Sanladerer made an overly complex bouncy box to hold a printer to see if it would reduce ringing and... it didn't really do anything yes, because Tom (and other youtubers) don't understand where the ringing is coming from: the ringing does not happen because the whole printer is moving! the ringing happens because there is some wobble between the print head and the bed! if you have a perfectly robust printer you can put the printer on the back of a pickup truck and make a perfect print without any ringing while driving through the desert (and that's actually exactly what the designers of the Pantheon Design 3d printers did: their printers use spindle instead of belts and therefore they don't have problems with ringing. to prove it they put the printer on the back of a truck and went offroading (i think this story was mentioned in one of nero 3d or cnc kitchens videos a year ago)))

[@rcmaniac25]

@@JanMuell42 I literally was just chatting with someone and made the comment about how I'm not convinced the surface a printer is on matters because everything should be relative to the frame, not the table it's on. It's also why I proposed that question of rerunning the rests without recalibrating.

[@CookieTube]

@@JanMuell42 Jup... But that is in a perfect world (or if you have a super rigid construction which doesn't flex at all). In the real world, with the common budget printer, a moving (or 'wobbling') frame is always going to have a certain influence on the rigidity of the printer arms and heads, etc. It is just plain old mechanics 101. (and also, a stepper can only take so much abuse before it starts skipping). So in that sense: yes, outside factors can and will influence it! However, the *real* question is: how much so? That is an entirely other story! If you can compensate for the 'sloppyness' with other means (like input shaping, like frame damping, etc etc), then the perceived change and visual defects would be greatly minimized, if not entirely imperceivable so it seems. Of course, the less you need to compensate in the first place the better.... As with all things, it comes down to nuance and _'what is in the small print'_ I think. I never liked oneliners like _"It doesn't have any influence"_ , or _"doesn't matter"_ . I always want to ask the how and why and see the details, especially when things go against common sense, gut feeling or basic knowledge.

[@SwervingLemon]

My printers are super-rigid box frame contraptions and they have far less ringing artifacts if I unlock the casters and let them move around a bit. I might be able to change that if I switch to aramid-weave belts. The machines move when the heads change velocity, so eliminating belt stretch (and vibration) might help clear that up.

[@tomhsia4354]

​​​​@@SwervingLemonTom's tests quoted in the comment agree. He printed some springy (much more springy than the stock rubber feet) shock absorbing feet and found those to reduce ringing somewhat. That's why he built the overly complex box.

[@Thurloat]

Emily the engineer did some pretty extreme testing recently too, upside down, on its side, hanging from the ceiling, getting kicked and swung, in a vat of oil 😂. Love these test videos

[@amarug]

I bought the first ultimaker 1 in 2011 when it was still insane to own a 3d printer. i put it on a table and printed a test object. i saw the table wobbling very very slightly and my engineering instincts instantly shouted "this will ruin your print quality". so i put in on the hard floor and reprinted the part, and indeed, it was visibly better. that trusty wooden 3d printer is still there on the floor, printing as beautifully as ever, never replaced a part, not even a belt.

[@stumpagness11]

You should see Emily the Engineer's video where she hangs an Ender 3 from the ceiling by a single piece of rope, while she swung it around. The benchy came out pretty much perfect!

[@MakersMuse]

LOL I'll have to check that out!

[@logicalfundy]

This is actually a bit of a concern for Prusa machines, especially the Mk4, which has input shaping so it prints fast - but the compensation is done at the factory, then the accelerometer is removed before shipping. So they can compensate for the printer's own vibrations, but the printers can't calibrate themselves based on the environment. It's one of those decisions that baffles me about Prusa's Mk4.

[@KaeptnKnarz]

Well that explains why my prints look like shit since I moved it from my sturdy work bench to my wobbly lack enclosure...

[@anbu94]

I put a MK4 in a prusa enclosure and got ghosting from just that. the printer just sits on a sheet metal floor that has no support under it. I stuck a scrap piece of 1 inch thick sound attenuation foam under it to support the floor and that helped but not perfect.

[@robertballerstedt5688]

I began to notice odd delaminations and wonky exterior faces once I began using the Mk4 on our wobbly table!

[@VacFink]

I appreciate the recognition that although it doesn't matter as much, as a communicator, how it looks helps keep the focus where it's needed. This kind of curious dialog between audience and creators is what is most unique about KZbin. Excellent experiment.

[@Jakob127]

Would be interesting to see how that would be on a printer that is not individually calibrated for input shaping. Like the Prusa Mk4/Mini that have Input shaping but no accelerometer.

[@rogerknutson9258]

I was just about to comment something similar. The test here was done using a printer that is pretty well setup out of the box. What about running something like a Wanhao i3 or something similar as these weren't the best printers to begin with from the start.. Very curious to see those results. From personal experience i've since moved away from wobbly tables as much as possible and continue to advocate against them.

[@JohnUllrey]

I was running my Original Prusa Mk.2S on a plywood table built into a closet wall, the table was amplifying the printer noise and making it really loud. I used a cement paving tile and a partially inflated air cushion under the tile to quiet it down. This made an incredible difference and now I just hear the cooling fan.

[@slowcyclist4324]

Glad I saw this. I was just about to remodel my entire store room just so that I could clear floor space to place my incoming A1 on the floor. Thankfully all I need to do now is just clear some storage space, and that is magnitudes easier than organising the entire storeroom.

[@pianoomann88]

Great video. I vote to keep the wobbly table. It helps the average consumer/enthusiast like me know that the print quality won't change too much even if they don't have the ideal setup. Nice work 👍🏻

[@I_EpicsStudios2]

This was an interesting video for me because when I was getting into printing, I always heard that printing on wobbly tables or non-level surfaces was bad for quality. Despite this, I set up my first Ender 3 on a 120 year old table that belonged to my great-great grandma. It is a beautiful table, but it was so wobbly due to the table's connecting joints becoming loose with age. When I got my second Ender 3 a year later, I upgraded to a much sturdier table that could fit both printers, but I found no difference in print quality, leaving me questioning everything I had heard about wobbly tables and non-level surfaces.

[@Simon-L-B]

Thanks! This stopped me worrying whether my slightly wobbly chest of drawers would be affecting my prints

[@oasntet]

I suspect machines not using linear rails for every axis will have worse results. There's a lot more flex in something like a rubber wheel running directly on v-rail. So vibration damping is probably still quite useful for the majority of machines out there.

[@KryaDiere]

Just wanna say.... I don't think it's visible with the big stuff you guys usually (test) print, but I print small items with little nobs (smaller than a benchy) that are also ironed on the top surfaces to sell. I always see issues on the little nobs. My 2 bedslinger printers are on the same table as 2 (paper) printers and a silhouette cameo (plotter), and these 3 tend to shake my table when they're working. I've gotten layer shifts when I have them all on and working at the same time. That's why I was very interested in this anti vibration tech by bambu.

[@ExtantFrodo2]

When I saw the video title I thought you meant "Wobbly print bed". Your test was well worth performing.

[@armandoa2484]

Excellent video! The speed on that little printer is impressive. I think the slight ghosting could be a combination of both the increased speed and belt vibration. All things considered still impressive prints. Thanks again for the great information!

[@markbooth3066]

A neighbour left an aged parasol base beside their driveway. After it wasn't collected for a couple of months, I took it, drilled out the wobbly steel upright (obviously the reason it was thrown out), cleaned up the slab, slipped it under my printer, and it made a world of difference. Being decorative granite, it look gorgeous on my worktop. Since the top side is highly polished, there's no cable abrasion, and the printers rubber feet grip it nicely. As you say though, the big thing is how much quieter the whole printer is.

[@sweat-lord6261]

Thank you for this, Iv got the A1 and i have it on a VERY cheap Garage shelf that iv put weights on to reduce the wobble, but it still wobbles so much, good to know its not really an issue, even when we are upside down(fellow aussie)

[@Oddman1980]

I have been running a Creality K1 on a wobbly folding table for months, and I haven't noticed any problems from it.

[@UncleJessy]

You mad man. Honestly impressed with that hanging test

[@MakersMuse]

It's a little tank tbh!

[@UncleJessy]

@@MakersMuse I have a fun little test with it for this weekend for the ERRF conference

[@AirsoftAbominations]

One of the early anto resonance mods was to put the whole printer on freestanding linear rails, so if anything it should improve

[@SirTools]

Fun and interesting results. Can't say that I'll be looking for a wobbly table for my 3d printers. I was thinking about putting one in the back of the car with a power station and run a print while driving around town....🤠

[@CGwatcher]

It would be much better to actually look at the parameters that were set by autocalibration in each particular case. Because even if the quality may look similar, effective printing speed might suffer if the acceleration is reduced due to calibration detecting more vibrations (not sure if this is the case for Bamboo autocalibration though, it might only change the shaper frequencies and that would just lead to more artifacts with the same acceleration).

[@Zettymaster]

its a bambu, so tough luck getting those variables out of the printer.

[@jort93z]

I feel like he'd have mentioned if one of the prints took longer?

[@spoolheads]

Thanks so much! I have six P1S printers and I still haven't made a desk for them yet out of fear of vibrations from one printer effecting the others. I've got a couple on concrete slabs and I think I'll buy a big multi-shelf unit and just put all the printers on the concrete slabs. Thanks for the video!

[@gpmilz8]

I appreciate the honesty of reviewers !

[@BitterSweetYTB]

My Prusa mini also printed perfectly fine on an unstable desktop, although that was an accident, and it didnt seem to be as noisy as the A1 mini. Im very satisfied with it

[@calaphos]

I think the main problem is the calibration sequence on a wobbly printer. The goal is to measure resonances within the printer structure (e.g. between bed and toolhead), but measurement is done with an accelerometer which will measure overall (unintended) movement of the toolhead - even if some of that is the whole printer structure moving. This of course depends a lot on the actual resonance of a hanging printer or an unstable desk and if/how much the calibration sequence ignores such low frequency resonances.

[@victornpb]

I think you’re onto something, solid mounted printer will measure vibrations on the tool. When it’s wobbly the frame will also move in the opposite direction but that’s not being accounted for.

[@TardyTheTurtleOG]

For some reason I've always loved the little cantilevered printers and always wanted a Cetus, but man I adore that clear A1!

[@TardyTheTurtleOG]

Also you should try suspending a second printer from the first suspended with both printing to see if quality gets substantially worse. 😂

[@edwardpaulsen1074]

I never had any doubt. As long as the printer has high rigidity and low flex between the axes as well as a fairly tight movement system with low backlash, then the printer will not care. It is striving to go to very specific points in space and it will do so with a tight build. Many, if not most, artifacts are a result of loose or stretching belts and /or loose or wobbling axes (especially the X & Zed axis riser) and other backlash or imprecision issues... even the flex in loose rubber wheels. Many of those can be corrected with proper and consistent maintenance. It *IS*, after all, a machine tool! Even expensive lathes and multi axis milling machines are only as good as the latest maintenance. Proper homing and leveling checks are necessary on at least a weekly basis during daily use. Still, an excellent example of what a good solid base and a bit of compensation can do. Thank you very much for testing out the theory and proving that 3D printers DO have the capabilities as long as they are rigid and the slicer software is accurate with the commands being sent.

[@HandleGoesHeree]

Great video! Have been wondering for ages whether my table is actually a problem

[@levyshai]

Great video buy I think the real thing someone needs to test is the effect of wobbling on MTBF or other printer wear and tear. As well as stats about model detaching from surface, the likeliness of that in correlation to wobbling

[@PaulG.x]

Usually "louder" is because the table top is acting as a sound board Hanging that printer from the gantry column probably was not the best option . The joint at the base of the column was designed to support the weight of the column and the gantry/print head. Hanging the printer from the column makes it support the base , power supply , electronics etc - much more mass and the joint flexes more. It would be better to hang it from the base using a cradle of cords or bungies from the corners of the base.

[@stevesloan6775]

It would be good too see a 3d scan comparison. Definitely a perfect example to show off the differences.

[@musca8803]

Thanks for your tests. In my experience (without input shaping), the rigid and massive base (or the floor) together with a thin and soft layer of rubber to adsorb and dissipate vibrations is the best solution.

[@Nebby_99]

I suspect my Ender 3 Pro won't be as resilient as this printer 😆

[@kaizoor]

3D Printing in space: What are saggy bridges?

[@DonsArtnGames]

I use the concrete tile to dampen the noise from my 3D printers. The 3D Printers were a lot quieter since the vibrations don't make it to the surfaces that amplify the noise (my tables and desks). But like many others, I'd like to see a printer's results without input shaping on the different surfaces.

[@CookieTube]

I'm very glad to see this test in the way you did it Angus! And also the comments you make about all this. It DOES confirm that vibrations/wobbly underground DOES have an influence on the print quality! It makes sense, it passes the smell test, it confirms the gut feeling. I have seen many people/youtubers in the past stating otherwise. Stating it doesn't matter at all. Which doesn't sit right with me; I always had the feeling: it doesn't mean that if you can't see it with a quick test, that it isn't there.... And your test confirms this thought. However, it also shows that with 'normal' usage, those influences are extremely small. Up to the point they are as good as imperceivable when you use an ordinary decent normal table or underground. And all this taken in account that with the printer you used there is input shaping/frequency calibration being done. So, I suspect, without such a 'modern' feature, the result might be a little more pronounced. How much so? Accordingly to those 'older' test from those 'random' youtubers, not much more difference either. But I strongly suspect it does have a bit more influence. Conclusion: use any decently stable enough underground, but you don't need to go overboard with it; a normal table will do. (unless you're of the kind who paints flames on the car to make it go faster.... in that case: go all in, it will not hurt :-P )

[@Davila_Designs]

This is the exact video i was looking for! Thank you! I just bought a 3d printer and Ill be using it in a trailer 😅. I wasnt sure how that would affect print quality

[@glennfelpel9785]

Thank you for running these experiments. I have been curious about all of this too. For an X1 Carbon another thing to consider is how you have the cabinet door when you run the vibration calibration. The first time I did the calibration the door was open. Later I began thinking that would certainly change the dynamic responses since we run the printer with the door closed. So of course, we reran the calibration. I really don't know if it made a difference but still it is more realistic to do the calibration with the door closed.

[@aaronstestlab]

That benchy from the suspended machine still looks better than what my Voxelab can do.

[@Zardoz66]

Awesome fun video. it goes to show how much this thing is doing. I am a firm believer of cnc kitchen take on this, and even though Bambu does a fine job of stabilizing. I believe anything you can do to help stabilize a printer will help with print quality and certain noise.

[@Dia1Up]

A wobbly table is ok if the printer has input shaping imo. (I'm the guy who gave you the maker coin at LTX!)

[@RoseKindred]

"No one in their right mind will do this." Meanwhile, I was already thinking of a way to install hanging chains in the overhead beam to do this myself.

[@morron2474]

Considering the feedback, its crucial to research thoroughly before investing in the A1 mini

[@charlesballiet7074]

oscillations are tricky things. sometimes its about particular harmonics creating interference nodes at inconvenient locations rather than raw displacement

[@LostApathyDotCom]

I've definitely had issues with vibration. I have an X1 Carbon and had it on a small cart on casters, so it wobbled a lot. The parts looked great, but were off dimensionally (printing honeycomb wall parts - they couldn't snap together). Put the printer on the floor, and the parts are fine. Took the casters off the cart, and it prints fine too.

[@brichardson7884]

This makes me feel much better about my setup lol. Here I was worried about perfectly leveling my workbench 😅

[@daliasprints9798]

Like all forms of damping, a wobbly table replaces high frequency resonances that input shaping could cleanly eliminate with minimal positional inaccuracy/smoothing, with lower frequency resonances that might show less obvious ringing patterns, but which are impossible to correct with IS (because the filter window would be so long it would ruin accuracy). If you want high quality without slowing down a lot, you HAVE TO stiffen everything up.

[@teammattchew4657]

as a newcomer to the 3d printing world this is super helpful since now i have to stop blaming my issues on my filing cabinet and actually calibrate my machine -_-

[@fatboy1271]

Quality for hangin in the air looks pretty awesome for what it is 🤘🤘

[@swells1]

I was looking to get a more sturdy table… no more! Thanks for this. Subscribed

[@mistertechnik]

Now I wanna see the A1 Mini printing in zero G, I think that would be really interesting to look at

[@jaadotech]

Two (2x) A1 printers inversely opposed mounted on a bracket, printing identical models in perfect synchronisation suspended from an industrial spring. I’ll wait for the follow up video!

[@ausfoodgarden]

Very entertaining and interesting video Angus. I was worried about printing on a wobbly table and took extreme action too and saw no difference. Back when I was working with large machinery we actually solved manufacturing issues by putting machinery on flexible mounts to let them move more. Physics can be very complex. 😁 Cheers!

[@keithcress1335]

I'd bet that in this case the "more noise" is because the drivers have to actually supply more power. It would be interesting to see what a watt-meter shows between the three.

[@conorstewart2214]

What is your reasoning that the drivers have to supply more power?

[@keithcress1335]

@@conorstewart2214 Possibly because the drives now have to contribute to the machine's inertial motion in addition to the normal axis inertia.

[@MisterkeTube]

@@keithcress1335 Now they don't as the only thing they need to drive is the relative movement of the printhead or bed vs the printer's main body. The global movement of the printer is irrelevant (well apart from how that might cause waves of deformation going through the printer and the printed object, but apart from tall narrow prints, I think that effect will be minimal).

[@justinc2633]

@@keithcress1335 the force on the machine is a by-product of moving the tool head, if anything the steppers dont have to work as hard as on concrete. also its definitely not the drivers, modern stepper drivers are the quietest part of any decent 3d printer, regardless of load

[@KareemFloat]

the table surface itself also acts as a speaker panel basically. sorta like those resonance speakers that can turn "anything" into a speaker. as even though there is a arm supporting the table surface, there isnt anything else to dampen the noise. if you were to add eva foam under the table even, it would make a big difference and that's not even with direct contact with the printer itself. it's how i dropped quite a few dbs from mine.

[@ericmary1338]

I have 3 printers running sometimes 24/7 on a sailboat. Waves, bouncing around in wakes at dock and even heeling suddenly with a passing squall. No problems, no changes in quality.

[@reforgedcriterion1471]

This is hilarious. I was just showing this off on my only available flat surface in my 23+ printer farm (which is a table that swings over your lap, or bed, or whatever on wheels) and showing off the flawless results with this table moving a full 2-3 inches! haha.. I was going on chuckling about a torture test and someone posted your video!!! Great minds do think alike! Although I can't say I'd have gone as far as hanging it lmao. Well played sir!

[@falxonPSN]

One thought occurred to me regarding hanging it from the ceiling. If you had a second cord attached at another point on the machine that was elastic but not too bouncy, like an exercise band, that would help to provide some damping and might even reduce errors to the point where it would be the same as the wobbly table.

[@rodneysmith1750]

What a riot Angus, you have, as always done a super job on this one!

[@dtibor5903]

For me the woobly table looks better a bit, than the concrete floor. I did similar tests and found that the best is to use a big tile (or any similar sheet of rigid material) above 2cm soft PU foam. That foam dampens all vibrations. Yes the printer wobbles a bit but there are no oscillations like when hanged with a thread. Rigid surfaces does not dissipate vibrations either, especially if the feet of the printer are rigid.

[@orange-micro-fiber9740]

I didn't know that model printer has resonance compensation. That's neat. Calibrating for the resonance is a smart idea. Now I have to see if my printer (sv06) has that.

[@stevekay6895]

I had my X1C on a bench that wasn't completely supported on the front edge, it was attached to the wall so I figured it would be fine...Then I started Printing and It Constantly threw resonance errors till I supported the front edge of the bench with a random pile of things I had around the house. So I have to say it makes a difference.

[@6kunio8]

There's something about watching the suspended printer printing that just makes me laugh! This was a pretty interesting experiment

[@vynaltheworld4092]

I love this community so much.

[@williamwilson4049]

the feet of your printer makes a huge difference too. having solid feet lets vibrations resonate through the entire machine, if you use soft materials on the feet it absorbs a lot of it I put some impact foam under each foot and it works great. Also if you want to print new legs for the machine Racket Balls work amazing to absorb shock from vibrations

[@geoff3610]

i got a neptune 4 recently, first printer, and it is proudly a member of team floor printer. I had trouble even after bracing the table on 3/4 sides with wadded up socks. Eventually just put the table on the ground and printer on that. Dunno if it truly helps, but it makes me less anxious lol.

[@ArtisanTony]

The biggest question is not what it does to the print but the printer. The stress on the structure created by that energy no dissipating into the surface has to wear on the joints and other parts of the printer. Of course this would not be the case in space :)

[@se.real.kiliong]

I think the problem with the hanging test was actually just the point of the printer you hung it onto. If my logic is correct, the print quality should be just fine, if you hang it from the base, for example using 4 strings or some kind of hanging gantry

[@thesledgehammerblog]

The top of the gantry is generally one of the strongest points on a bedslinger, so I think it would be a valid test. In fact, the instructions on my Prusa Mini actually tell the user to hold the printer by the Z stepper motor at the top of the gantry when carrying it.

[@se.real.kiliong]

@@thesledgehammerblog I think you don't get the problem. The issue with hanging it up on that point is, that the swinging weight flexes the gantry, therefore creating an inaccurate print. This flexing is well within the flexible deformation limits of the material and therefore not a problem when carrying while turned off. Running it like that leaves some ugly artifacts though, which have nothing to do with vibration.

[@tyrannicpuppy]

Fascinating. I'd love to see a repeat with a machine that doesn't do input shaping and other calibration to compensate for such things. My only anecdotal evidence I can provide is that when I bought my Ender 3 Max, I stuck it on the cheap nasty wobbly computer corner desk I had free at the time, and I couldn't get anything but spaghetti out of it. Bought a small lamp table the next day with four diagonal outwardly pointing legs fixed at each corner that doesn't wobble at all and suddenly I was getting successful prints. So I think there may be a threshold where the machine cannot compensate at all.

[@joncorso6103]

This is just wild. Better prints from a machine tarzan-ing from a garage ceiling than I ever got from my Ender 3 with countless tuning.

[@Radioman-pv5np]

Ok, next test. Printer on the hood of a car while driving through rural Texas! Great content! Loved this video.

[@Eff917]

My guess is the effect really comes down how well the printer is put together. If you have some play or backlash in the components, then the effect of the wobbling will be way more pronounced, as it will exaggerate them.

[@GruesomeJeans]

Hanging a printer and having it print is wild... I have a cheap sub $200 Anycubic printer that sits on a wobbly dresser on carpet. I feel like the print quality is on par for the quality of printer I have. I did add the concrete paver and thick squishy foam and the sound difference is quite dramatic.

[@WwAiRtRhIiOnR]

Planned to put my printer onto my old, no longer used, DIY diaper change table, which is pretty sturdy and additionally Anker it to the wall corner of the room with two rods on both sides. Might reconsider it now, but probably do it anyway for noise reduction. Something you could have measured would be power consumption. As far as I understand how inputshaping works: it uses the motors to conter the vibration (not unlike using acceleration in the opposite direction instead of just breaks) Could need more energy on a wobbly desk compared to the concrete floor.

[@goatboy150]

Your walls are gonna resonate.

[@PasswordincorrectStudios]

Emily the engineer also made a video similar to this she tested how far she could push a ender 3 printer both hers and yours are both great

[@Graham_Wideman]

3:45 I like how Angus' hair is so advanced it has its own wifi antenna! I guess it's a hair-ial ! 🙂

[@YoSpiff]

Thanks, that was helpful. I've had my printer on a pretty solid workbench but wanted to put my new one on a desktop next to it that is nearly as solid but not perfectly level. Sounds like it wont affect anything.

[@fdm225]

lol, Truly impressive. Hanging the printer was unique. Question for you on the auto-calibration, you mention that you ran it between changing what you were printing on; it brings to mind the question as to when exactly should you run the auto-calibrations on the bambu lab printers? I am getting my first one (coming from a 5 year old printer that did most things manual) and am wanting to know what (and when) it's recommended to calibrate. great series btw

[@johnford7847]

Very interesting video, Angus. Thank you for sharing. I hope you keep investigating this issue.

[@alexmecanique]

Happy, you talk about this, I tough about this many time watching your videos :D But I think you would see a more negative effect with a cheaper DIY 3d printer.

[@tophatvideosinc.5858]

If you want to, get a granite surface plate and a steel roller table. Surface plates are heavy and perfectly flat and the roller table can support that and allownyounto move it out of the way or around your shop for other projects. And having a perfectly flat surface can be useful when showing off printer accuracy. Like the corners or elephants foot, as well as a good surface to use guages on for reference, tuning or repeatability.

[@thesledgehammerblog]

My Prusa Mini lives on top of a large Ikea Kallax shelf which definitely has some wobble to it when printing, especially with the newly added input shaping in the 5.1 alpha firmware. I did print out the 16 minute "Bonkers Benchy" Josef Prusa put out with the firmware update and it definitely has some artifacting, so I am now curious as to how well it would work if I printed it on a more solid surface. That said, I've generally had no issues with print quality before the input shaping was added, and prints came out looking nicer than the ones from my Ender 3 S1 that was sitting on the ground.

[@oseaghin]

This is truly remarkable. Thanks for the testing!!

[@Roobotics]

For the suspended one, I suspect because of X/Y movements, you are -mixing- the resonances between the two axis, while the software only compensates for the resonance of a particular motor on it's own axis, also that violent shaking is likely causing Z bounce which further cuts into layer accuracies, but it hardly shows that.. Maybe if a fast cylinder was printed like that. One slow layer, then a fast layer, then a slow layer.. but that's oddly specific. But the idea is that fast layers would squish and spread in areas due to XY messing with Z. A cylinder at just one speed would be mostly uniform in those distortions, so harder to detect at only the first layers.

[@RaoulEvilD]

Nice test idea! I am convinced that CALIBRATING in mid air was the issue, not PRINTING in mid-air itself. The calibration process aims at mesuring the motion of the nozzle/bed relative to THE MACHINE'S FRAME. However accelerometers used in the machine measure motion RELATIVE TO EARTH. When the bed does not move relative to the earth, the motion of the nozzle/bed relative to the machine's frame is the same as the motion of the nozzle/bed relative to earth. But when the machine's frame is constently accelerating in various directions relative to earth then the motion of the nozzle/bed relative to the machine's frame is no longer the same as that relative to earth. Therefore the data acquired from the accelerometers become unreliable, leading to faulty input shaping calibration. If you hadn't recalibrated the machine in mid-air I am pretty sure that hannging the machine would have been unconsequential on the print quality for any practical purposes.

[@Hughie1987]

I've printed almost 30hr prints on an ender 3 on a wobbly ass TV table sitting on carpet and the prints came out perfect. No layer lines or anything

[@carlchong7592]

The reason that the stiffness, or inertia of the table, doesn't matter is that the print head is quite light compared to the mass of the base. I expect that the control systems guys from DJI are familiar with the issues of vibrational impedance matching. The reason that the hanging test showed new defects is that the beam loading scheme for the vertical column is so different than for when the heavy base is supported at the bottom. By hanging the printer from the vertical column, the stiffness of the vertical column is now supporting the acceleration loads of of the bed. In this configuration, oscillations are resulting in increased beam bending loads on the column that weren't present before. In previous (base loaded) tests, the column is only resisting the inertia of the the print head and slider, but in the hanging test, the additional swaying load from the moving bed is imposed on the column. If one were to consider the system as three major assemblies: print head and X-Z axis, bed and Y axis, and base you see two very different loading schemes between base supported and column hanging. With column hanging we see that loads from the print head and XZ are transmitted into the column AND the inertial loads from the bed slinger also transmitted into the column. This results in more variable bending moments transmitted into the column. In the base supported config (normal config), bed slinging loads are transmitted into the more rigid and heavy base, print head and XZ loads are transmitted to the base via the column. This puts less variable load into the column because the mass of the base can more easily accept the oscillation of the bed slinging than the column can just by being heavier and very stiff. A more representative test would be to support the base on a thick soft foam block than to hang it from the column. Noise reduction on the concrete floor is probably more to do with the floor not being a very good "drum skin" for transmitting vibrations to the air. If this were true then a good test would be to support the printer on a thick block of squishy foam. A squishy foam block would not be a very efficient drum skin for resonating vibration into surrounding air.

[@dblackburn81071]

I have a FL Sun Q5 on an old end table printing at 60 mm a second and it prints just fine

[@davidconner-shover51]

what surprises me is why they are not adding counterbalances on bed slingers, one to compensate for the print head, another for the print bed, Granted, this would introduce a twisting torque, and double the forces the stepper motors require to move..

[@felsinferguson1125]

Guess I'm in good shape - I've got my Ender3v2 sitting on a retired laser bed. A big ol' honkin' beast of a steel thing about 8 feet long by 5 feet wide, with cranks and wheels to raise, lower, and move both an inch and a half thick perforated (1 hole per inch in a grid) steel slab and the laser-mount in the X/Y for lining up an old CO2 laser for cutting work. The thing might as well be a block of bedrock poking up through the concrete floor it sits on. I'm told that it took 3 good sized guys using two handcarts to get it to where it sits. By myself, I've moved it a grand total of about 4 inches, and only managed that by repeatedly slamming against it. Pushing was a hopeless cause. At a guess, it probably weighs more (and I'm certain it's more stable) than my Mazda pickup!

[@jme2006]

you should print large blocky objects with lots of corners. That shows how vibrations can affect surfaces in more apparent ways.

[@drew79s]

If you look really closely you can see how the input shaping has actually reduced the accelerations in each case and the resonance indications change pitch in line with setup stiffness... It's an interesting one :) The reason they're bigger on the hanging one is because the amplitude should go up as the pitch goes down, and it's dependent on the target parameters of the input shaping. Given the fact that the machine, when hanging is converting its motion into linear and polar motion, you could probably get an even lower print quality by placing the printer on a flat plate which had ball bearings underneath it, limiting polar motion, but allowing all the linear motion. Regarding the printers energy, there's no net energy gain for the 3d printer, except for a very small amount related to the displacement of the filament and the raising of the z axis. All of the rest averages to zero, because the printer isn't moving, ie; if you place it on a bench, part of it's design isn't to displace it's self along that surface. All the motion is internal. The other thing I suspect would be interesting is the print times, each version of the print would take a different period of time, because the input shaping is obviously reducing the accelerations.