Hello mate, may I ask the mesures please? I'm looking for a mount like that for my BZ181 - 18X18X1mm

Thanks you for your time

Really excellent work, thank you. (edited)

Not the real Size I have a 3D scanned data and compared it to this. The 3D scan I got is closer to the actual thing but this file has cleaner surfaces so some rescaling has to be done for it to match the actual A2

The fit is perfect, providing a snug and secure hold. When using more than one, however, it would be ideal if the ends had a slight underlap for a smoother, more seamless connection.

Love these! I printed enough of these to cover all of the visible slots on my XL

Very nice design! Printed very well on the K1 max using black PETG using a skirt and some ear-brims (2h50m with a 0.4mm nozzle & 0.2mm layer height). Needed some trimming with the exacto knife afterwards and some sanding to make the scraper edge a bit more sharp but it came out great.

Update: tested the scraper comparing it to 2 other 3d printed scrapers (see picture). This Scraper has the widest removal of ice from the window if you put enough pressure to it so that it can bend. Works great for me but for a smaller person the needed pressure could be a bit too much.
Update 2: it also worked well on frozen hard snow on the windshield, impressive! (edited)

The variomatic ice scraper! ;-) Finally a contest entry based on real engineering!

Prusament PETG in Prusa Orange.

I think it needs a bunch of vents as I find the ESP32 locks up due to heat if the flash is turned on full time.

I haven't finished my enXLosure yet, so I'm just using BluTac as a temporary mount. (edited)

I replaced one of the M3x8 screws that was used on the Enxlosure to hold the PC panels in with a much longer M3, pushed that through the camera's screwhole in the arm, and secured it with a washer and a locknut. (edited)

The mounting holes for the angle adapter is not all the way through. Was this intended for glue or it was just missed? thank you :)

Printed in PLA and top mounted instead of rear mount as the pictured rear mounting option is right above the open section of the where all the tool heads are and likely not extracting all of the VOC's as intended. Also with thermodynamics, top mounting made more sense to me as heat rises potentially carrying all the VOC's. Overall was great design :)

Hi - I really like this design - but as others have noted - the doors overlap a lot and there's no play in the design to make that right. Is there a fix @Voxel3D ? Thanks!

Update - all of a day later. After reading several of the other reviews I determined that you really need to force/twist the FCB pieces to the outside to make the doors fit correctly. It's a lot of force and what's really happening is that the T-Nuts are moving around in the slot incorrectly. If I get some time I will remix the FCB parts to provide some adjustability. Printed in Priline Superhard PC CF on the Prusa XL which worked great. It's my favorite functional filament. I also used the Blurolls kit which was very nice. As others have noted - the tolerances are very tight - get your scaling factor correct and it should work fine.

New update: If you are having trouble with the upper door and front doors lining up correctly, I made a few remixed parts to address that problem. It is worth reprinting these parts in my opinion. https://www.printables.com/model/721094-cleaning-up-the-front-doors-of-the-enxlosure (edited)

A few points to start with. This is maybe one of a small handful of models (like less than 10) I've printed that I haven’t cadded up myself since I've been doing any 3D print work.

The vast majority of stuff I use a printer for are either impossible to source replacement parts (like a passenger door electric lock actuator mount for a 1995 Dodge Ram 3500 pickup), tools for other things, one off custom router jigs, little "problem solvers" for the house and garden and things along those lines. I don't spend much (any actually) time prowling a 3D model site looking for interesting stuff to print.

So my background in this particular build:

I started with 3d printing by building the MK3S from the kit, added the MMU and I used a few MMU mods to get that solid.

My first major 3D Print project was the MK3's enclosure which I used as both a functional piece and a mechanism to translate my existing CAD chops to how does that apply this FDM thing, which is here

 https://www.printables.com/model/205693-prusa-mk3-convertible-print-enclosure

When I pre-ordered the XL the first week in January after the initial announcement, knowing that the larger print volume would solve a raft of design issue, and the tool changer would hugely accelerate multi-material work, as well as save a lot of design work in not having to spend nearly as much time trying to avoid using supports.

As I'm doing more and more vehicle and outdoor parts, the move from mostly PETG to mostly ASA and CF enhanced ASA was going to make the "probably should have" an enclosure just to avoid draft and corner lift issues to a "absolutely will need an enclosure".

Knowing how much cad time I burned creating my I3 enclosure monstrosity it dawned on me that somebody may have solved this one, so for the first time in years I start browsing this site looking for a solution and Voxel3D's was it.

First, he'd already done all the heavy lifting (and having designed one for the MK3S I knew just how much work THAT was, and this was utterly self contained to boot.

Second I wouldn't have to figure out every little bit of hardware I was gonna need, source that, and get it bought. Another big plus.

Third, I had a limited timeframe wise as I just didn't have enough free time to toss as the padding up a new design from scratch.

Fourth: There was a kit. Big bonus.

SO WHY THE BIG LONG ASSED POST?

So being as Voxel3D has already done all the heavy work I figured the least I could do was offer a fairly detailed writeup of what I encountered during the build, some things I changed (and why) and some pointers that may help others in their own builds of this thing.

SO as a result, this is gonna be a lengthy read. I'll try to break it up onto semi logical sections prefaced with the ALL CAPS thing as you can't do formatting on comments.

ABOUT THE KIT:

I ordered the kit of AliExpress (first time I've every ordered anything off that despite having an account for over 10 years), as the total cost was way less than I could have picked up the same amount of Plexiglas at Home Depot.

Further I would have had to tie up my neighbor's xTool P2 to cut them, and his schedule is as shitty as mine is.

I was very pleasantly surprised at the shipping speed and packaging, and they used the same style of heat set inserts I've come to prefer for a number of reasons.

THE SCALING:



The next step was to ensure that all this printing was going to work. The scaling tool helped but it was too dicey eyeballing a dimension with nothing more then my 58 year old post cataract eyes and a tape measure.

I was going to model a "test" part that ran up against some real world dimensions to verify the scale % but as it turns out Jeggo beat me to the punch and already had done so. Futher he picked the single best point to do it with, between the bed and the CoreXY point on the XL, and as that is all machined metal stuff.

Tf the test part doesn't fit exactly, your scaling is off:

https://www.printables.com/model/664844-shrink-tester

So I printed that after printing Voxel's shrink tester to confirm my hunch that it wasn't accurate.

The shrink test 475mm part showed a 100.42% correction factor, but Jeggo's part didn't fit (it bowed out) indicating the correction factor was too big.

So I measured that part against the CAD dimension of Jeggo's part, and did a correction factor off that, and reprinted it, which was again a bit too large (again due to measuring against a tape).

This indicates I need a really big set of good quality calipers for this sort of thing.

However, by doing some interpolated math between Voxel's correction factor and Jeggo's more accurate part fit, I did a 3rd print at 100.18% and that snapped in dead nuts on so I was good to go at that point.

Rather than start with the top of the frame I did one of the front columns first as a triple check against the scaling factor.

While that fit fine, it was my first time using Prusaslicer's paint on multi-material (as I've always modeled this stuff in on my own work).

The orange trim looked really dark and just didn't pop, and this was due to how the slicer handled the selected face.

There were places where the trim material color was only one layer thick, and against the black of the frame it just got washed out.

So that led to cracking open all the parts and cutting in the trim with a 5 degree draft angle to get at least 4-5 layers so the trim didn't get lost in the sea of surrounding black.

That took a fair bit of time but thanks to Voxel supplying the STEP files at least it was doable, as that would have been impossible if I had nothing but STL's or 3MF's to work with.

More on the "WHYS" behind that in the mods section.

THE HEAT SET INSERTS:

I fully agree that heat set inserts are the way to go on a project of this size (I use them pretty much all the time in my own work anyway).

There's good heat set inserts and really horrid heatset inserts. These are the "good ones" as they have the smaller diameter with the "ledge" at the bottom.
Initially I'd been doing all my heat set insert stuff with my Hakko bench station soldering iron, and some custom full length heat set inserters made for me by my machinist buddy.

And while this was fine for occasional one offs, doing masses of them called for something a bit more accurate in terms of placement and repeatability.

After a lot of research and pondering building my own a while back I settled on this one:

 https://www.amazon.com/dp/B0C7D34CL4?psc=1&ref=ppx_yo2ov_dt_b_product_details



I picked that one mostly because it was all machined parts, and thus the 90 degree's would be solid and keep the iron perpendicular to the work surface, and the iron they used would accept my existing heat set tools (as the ones they supply are pretty cheesy and the threaded bits aren't exactly going to keep the thing perpendicular).

That one gets it's own set of mods but that's outside of the shop of this write up.

If you do get one of these, do yourself a favor and treat the bearings the same way you would on an MK3S.

Give em an alcohol bath (as two of the bearings are packed in the rust preventative, and the other one is already installed (also still in the rust preventative) and lube them.

I made a set of SuperLube adapters for the bearings as the 8MM's are 8x45 and are longer than any of the Prusa bearings, and the vertical rod is a 12x30mm.

If anybody needs these let me know and I'll post them.



The press helped a lot and I ended up clamping it at the edge of the work table to allow me to use it for the larger parts.

However, due to the odd angles of Voxel's design, some of these had to be sort of freelanced with the press and I couldn't really use the depth stop as much as I'd liked.

Before I had the press I'd sort of developed a technique of sinking them, then immediately screwing in a very long (60mm) cap bolt into the inset to check the relative perpendiculars in question.

If they were off I'd back out the bolt, touch the inset again with the iron just enough to slightly soften the plastic, then screw the 60 back in and bring the bolt (and thus the heatset insert) in perpendicular.

I've found that for the most part a general eyeballing of the perpendicular with that long of a bolt screwed in with usually ensure that the shorter ones will go in without issue.

THE FIT:

Once that was done I assembled the side panels first, as it was the portion of the entire assembly that was most tightly coupled to the frame and thus would verify if my Jeggo tool derived correction factor was actually right.

I figured this would be much faster, and much less wasted material and print time to make sure the rest of the build went well.

I was happy to find that every one of the heatset inserts lined up exactly with the precut holes in the kit plexiglass.

Throughout the entire assembly I only had one that had any resistance to going in, due to my missing the heatset perpendicular check with my 60mm tester.

With the side panels in (still left the protective liners on them just in case I needed to dissemble anything) I proceeded to print the rest of the parts, working on the draft angle solid trim stuff for the next part while the previous one was printing.

This would have been easier with source models but I was grateful to have at least have STEPs to work with as I just have zero experience doctoring triangulated polygonal meshes.

All my CAD experience through the years has been with some variant of NURBS based modeler (SolidWorks / ProE / Rhino / any number of ACIS based derivatives) and I've always kind of ignored the polygonal mesh stuff as it's more of the purvey of CGI and game development than actual dimensionally accurate CAD for parts.

You'll never see a subcontractor for say automotive stuff designing parts off a triangulated mesh, and there's real good reason for that.

THE PRINTING (AND SUPPORTS):



One of the primary reasons I GOT the XL5TH was both for multi material stuff and the nifty ability to use dissimilar materials (PLA for PETG and PETG for PLA) as supports.

As the XL was literally just built I hadn't had time to play with any of that stuff but I found Prusa's article on the subject ( https://help.prusa3d.com/article/combining-materials-xl_498103 ) and cooked a specific set of filament profiles for doing so (to avoid the heatbed temp errors during slicing), and went to town with it.

The incredible ease with which the PLA just peeled right off the parts (including the really tall vertical ones), along with the quantum shift in tool changing speed compared to the MMU2 made (to me) the entry price of the XL worthwhile.

That's a huge game changer, and I can now model stuff without having to figure out how to minimize supports via part orientation to the bed.

So I liberally applied supports using the smart fill and enabled supports for generated material only (turning off auto generated supports).

What qualms I had about how tight I could fit the supports to the material vanished after the first couple of parts.


Settings wise the SNUG Style works fine, the XY separation can be pretty much anything (the default 50% works and I've taken it down to as little as 10% and it's still fine).

I did turn off solid sheath as recommended and while it wasn't a problem on the one part I forgot to do this one, you'll save a fair amount of support material (and print time) by doing so.

The only ones I actually used the wipe tower on was the tall BBR BBL parts and in reality probably could have got by just fine without the wipe tower.

I haven't used a wipe tower since and haven't had any issues (yet).

I didn't use any of the supplied plate layouts as I was having to substitute my own models with the cut trim for Voxel's anyway and was doing the print order pretty much ass backwards in order to get the doors in first to verify the correction factor.

ASSEMBLING THE TOP:

So once I'd cut in all the trim as closed solid poly surfaces and printed the upper stuff I did the top assembly.

Having read the admonishments about "being careful" and having seen several posts on the broken parts (and someones clever part to stick the broken parts back together) I took the whole "will be flimsy" bit very literally.

One trick I'd learned from my previous enclosure design I did for my MK3S (it's in one of the pictures here and I released the models for the printed parts for that a while back) was the need to keep things somewhat together and aligned during assembly, and with this frame being far less substantial than my ludicrously over engineered enclosure build for my MK3S this would be even more important.
(ASSEMBLY TIP - GRUB SCREWS AS ALIGNMENT PINS):

If you haven't assembled yours yet I'd highly recommend you pick up a few M3 Grub screws in the 10-12mm length range. These will be invaluable as you can partially screw them into the heatset insert side of the rail parts and use those to act as locater pins going into the non heatset insert opposing rail during assembly. It made the initial bolt up infinitely easier, particularly for doing that part solo without any helpers.


This allows you to get a bolt in with the remaining bits more or less in place where they will be, then move on to the next, removing the grub screw and replacing it with it's actual bolt.
(ASSEMBLY TIP - MAGNETS):

Glue in the big magnets first before assembling the parts. The reason for this will be evident in a bit.

Also. "pre flight" the other magnets polarizations before final assembly.

There's a couple ways you can do this (putting shapie dots or lines on one of the poles for example but these marks are very easily rubbed off).


The easiest way I've found to not get them messed up is to:

1) Glue one magnet in. For each pair of parts I figure out which part will be easiest to manipulate and then glue the magnet into the opposing part.

So for the door magnets I glued the top magnets in first. (Note on that in a bit).

2) Put the other sides magnet directly onto the magnet already glued in. Just let em snap together.

3) Put glue into the opposing parts place where the magnet goes.

4) Place this glued part over it's already properly oriented magnet attached to it's pair part.

5) Slide the freshly inserted part laterally away from it's partner parts magnet.

You slide rather than pull as the magnet will just stick to it's other magnet on the other part 6) Push the magnet into the part to fully seat it with the glue.

That worked remarkably well, with one caveat:

The door top magnets, while correctly polarized to the opposing part's magnet are incorrectly polarized to the ones for the other door, which leads to them opposing each other when the front doors are closed but the top is open.

In retrospect I should have matched up both sets to preclude that from happening.


(ASSEMBLY TIP - HOLD OFF ON TORQUING THE SCREWS DOWN):

While it's tempting to crank down all those little screws, don't.

Put then in enough to hold stuff together but don't torque them down until later.

There's several reasons of this:

1) Your initial assembly is NOT on the printer, and the base rails lower surface are NOT square to your assembly table.

2) No matter how well you did your correction scale, there will be slight dimensional inaccuracies and these will be compounded across the multitude of parts going together.

3) There is absolutely NO way to keep the frame square to itself during the assembly.

4) There will be some cumulative slop on the fitting of the plexiglass panel into the slots.

Therefore it's best to just get it together, leaving the parts somewhat loose (I did "fairly" tighten down the rail edge to rail edge bolts to try to keep the facing surface twisting to a minimum and try to get the plexiglass recess faces as square to the next rails recess face as much as possible).

One I had the whole roughly together I put the entire assembly on the printer (thus getting the bottom rails aligned in the XY&Z axises with respect to the printer). THEN I began tightening things up.



(ASSEMBLY TIP - TREAT THE ENTIRE ASSEMBLY LIKE A WHEEL WITH A HUGE NUMBER OF LUGNUTS)

Rather than just cranking down all these 100+ screws work around the entire assembly a bit at a time.

You'll decrease the "racking" of the assembly a lot, and find that stuff fits much better as a result.

What I did was tighten them all down to the point of having a little bit of resistance, and then worked a pattern of skipping as close to the alternating opposites like you would when torquing down a tire, except across multiple planes at once.

I'd do a few on one plexiglass piece, move to the other side, then the lower back panel, then the upper, and back to the "missed" bolts from the first trip and so one.

Each iteration I'd tighten it maybe a half turn beyond the previous trip that bolt..

Doing this way allowed for a gradual increase in perpendicular pressure and which was then transferred to the plexiglass panels and ultimately the frame under them.

What I found is that every time I came back to a bolt I'd previously thought was as tight as I could safely get it, by the time I got back to it again on the next round there was yet more play I could comfortably take out.

In all I probably made 9 trips around the entire enclosure gradually tightening everything up. This takes a bit more time but you'll find that you'll ultimately get a much better overall and more uniform torquing of the screws and panels to the frame.

You'll find the result to be far more structurally solid than just starting at a screw and moving on the next on all the way around the panel and moving on the next from there.

During that entire process I didn't have but one bolt I had any difficulty getting tightened (and that was the one I'd missed with my perpendicular check during the heat set insert process).

No squeaking, groaning or noises whatsoever, and every single bolt lined up exactly with where the pre cut holes in the plexiglass were.

Kudos to the AliExpress place, they pretty much nailed the laser cutting of the panels.

OBSERVATION FROM POSTS ON OTHER'S BUILDS:



I suspect that a lot of the problems other folks have had in getting this thing to go together falls into 1 or more of the following issues:

1 - Incorrect scale correction factor: This was described previously but I think it's probably a bit part of the overall fit when done.

2 - Not Perpendicular Heat Set Insert Setting. If these are skewed from perpendicular to the surface, you'll have a miserable time getting some of the bolts in. If you hear a horrid creaking noise while trying to torque a screw down, it's most like the the insert isn't square to the frame.

3 - Doors not fitting / overlapping etc

This is the most commonly reported issue, and I think it starts with the incorrect scale factor, continues with non perpendicular heat insert setting, gets compounded by completely tighting a bolt and moving on to the next one.

When done I was pleasantly surprised I didn't see the massive overlap of the doors I'd seen in other builds.

I was real happy with all that. However while there wasn't any overlap of the doors, there was a very noticeable height delta (Like almost 3mm) where the left door was 3MM down vertically from the right one.

I knew my fit was good just due to how all the screws like up exactly with the plexiglass, so it wasn't due to a parts not to size thing.

Looking at the hinges, the right one's mount to the beam is upside down to the left one (as the beams are the same part) and the threaded nut is opposite on the sides, and that to some extent contributes to the Z height they pull in at but that wasn't all of it.

OTHER MODS:

2 PART PANEL TOP REMIX:

I really liked PeterM's 2 part panel top, as I knew that initially I was going to just do a solid top but at some point will be adding some sort of temp controlled ventilation and filtration as well, and the top is most likely the bast place to land that stuff.

So rather than re-cad (end re-print) the top in is entirety for each iteration a fixed top "base" and interchangeable panels made good sense and PeterM had already done that bit.

However since I'd already had some corner lifting (like the scale test part) because the printer wasn't in an enclosure yet, I was a bit leery of printing the insert with that big of a flat surface on the bed.

This was compounded by brittle cold (for TX anyway) resulting in the HVAC cycling a LOT and when it first kicks in it was blowing markedly cold air.

Since I was already going to have to touch that model to model in the solid trim I also markedly increased his corner fillet radiuses to give the print the best chance of succeeding without issue.

You can find that remix here: https://www.printables.com/model/676169-enxlosure-2-part-top-panel-top-r2/remixes

TEMP / HUMIDTY SENSOR:
I also wanted to have a humidity / temp display IN the cabinet so I'd ordered one of Prusa's ones they use for their enclosure.

I modeled up a replacement for the top door stop / magnet part to accommodate the display.

I also modeled in the continuation of the two color trim notion for that part.

Since that was a simple part which I whipped up off of some dimensional takeoff's from Voxel's STEP file (and it was a bit of a tight fit landing the display in without running into the horizontal door rail right above it), I cadded that up, exported the 3MF, sliced it and sent it to print.

Then due to the weather and TX's joke of an electric grid, the power went out and I lost my cad file for the part.

All I have of that is the PrusaSlicer project file which you can find here:

 https://www.printables.com/model/742218-enxlosure-magnetmount-door-top-stop-w-display

If there's enough need / interest, let me know and I'll re-cad it and post the STEP. If I do it again I'll probably add heatset inserts for the mounting point as well.

ABOUT THE COLOR TRIM (AND MODLED IN COLOR)

This project was my first actual use of the XL as it'd literally just been built and I knew I had to get it into an enclosure ASAP or I would have never ending issues with corners lifting etc. as we were in the middle of a very nasty (for TX) cold snap that was causing the AC to cycle a LOT and on startup was blowing cold air in from the ducting in the attic.

I spotted Voxel3D's two color trim and thought, well that's neat, as I often tend to do the same kind of thing albeit on the very slow per tool change MMU2 for my MK3SS+.

When I pulled the model files down I noted there wasn't actually any solids for the color trim in the model and wondered how that was gong to work, and only then noted an oblique reference to the multi color stuff being done in PrusaSlicer.



This was my first encounter / usage of the multilateral "Painting" in PS. I've been doing all my multi material work as modeled in solids both in Rhino and Fusion360 since I first got the MMU2 as there wasn't any other means of doing so in those days.

So any place I wanted different materials I'd literally model them in, always using some multiple of the target layer height (for xy plane stuff) or the single wall thickness (for Z axis stuff).

Further I'd tend to model those in with draft angles for printability purposes.

I've found over the years of using the MMU2 that in many cases you not only need a few layers or wall thicknesses of the text / logo / trim stuff to show up well and have crisp defined edges, you may also need a few layers / wall thicknesses behind THAT printed in solid white just to get the colors to match what you're choosing filament wise or your bright yellow may come out turd brown because it's over solid black.

I've done this sort of thing so much that I've written a plugin for Rhino where I can drop in my logo, pick the center location and orientation, and set the overall scale to an integer number multiple of single wall thicknesses the "lettering" portion would be based on the print nozzle size.

The plugin does the logo, the logo surround and a single "backing" object behind the logo and surround.

I do this because I've found when using highly contrasting material colors (like this one where the bulk of the print is black) that you need a couple wall thicknesses of white behind the logo / text elements to not have them end up way darker than you like due to the translucency of the filaments and the "cuts" being so thin.

I have a similar script for fusion 360 that does essentially the same thing.

For solid expanses of just "trim" color (like on this model) I'd just model the stuff in from the get go.

So on first glance I thought, well this would save a bunch of time, and why hadn't I payed attention to the fact that feature's been out for a while. So I was super hopeful that this was going to save a fair bit of time in my own stuff going forward, and set up slicer to do it's thing.

That's when I noticed the pitfalls:

You have ZERO control over the extrusion depth of these "painted" areas.

Depending on Slicer's whim, this may be anywhere from 1-4 layers or 1-3 wall
thicknesses "deep" depending on the orientation.

Further, and far more distressing was that slicer will (depending on where this occurs) only print 1-2 layers or wall thicknesses of SOLID material and then use the remaining depth as infill.

So I test printed a couple chopped up sections of the rail edges and found my initial assessment after looking at the sliced output was correct.

Much of the color "depth" of the painted stuff was actually just infill, and when printed the solid color areas had noticeable optical "blurring" from the translucency of only being a layer or wall thickness deep, and further the infill pattern was clearly visible.

So my perfectionist OCD Aspbergers stuff kicked in and I knew I was gonna have to touch every piece that had the color trim stuff on it.

Again, fortunately, Voxel3D had supplied the step models so I was able to open them in Rhino and perform the necessary surgery to get actual solids that would print without infill to get the uniformity to the trim color I was after.

So while not posted as a remix, every part in my build has been touched to get "real" solid trim in place.

POST ASSEMBLY OVERALL RACKING:

On several openings and closings of the top door I noticed that the upper and lower door plexiglass panels were landing off kilter and was not settling into the grooves on the frame. It was then that I noticed the entire rail on one side had slide back from the front frame.

While the big magnets keep the assembly from just falling off they do not preclude front to back sliding.

That wasn't helping. I'd slide the entire rail forward to it's "stop" notching while doing so the entire structure was changing as I did so.

There just isn't enough cross bracing on this thing to "lock" it into "square".

That led to creating a set of "clamps" that mount to the perforated vent panel on the back that intersect and hold the two top side rails in place from front to back.

Just due to where they land on the existing structure they overlap one of the 8MM nuts used to bolt down the plexiglass so I derived enough measurements to line this up with Voxel's full assembly STEP file to ascertain the exact angle of the plexiglass bolt as well as the offsets to the holes on the perforated plate.

The resulting clamp matches up with his R5 version of the back parts:

Side note:
Only after I did a set and couldn't figure out why the didn't line up only to find he'd rev'd the part in the standalone part file but hadn't updated the overall step file with that revision).

You can use that Plexiglass mount bolt as a "lock" to hold the entire top assembly in place.

No need to overly torque it down it'll just keep the whole thing captive and preclude a lot of the lateral and axial shifting when opening and closing the top door.

Due to those screw locations there's a separate L & R clamp.

The bases though are identical, just make sure you orient the base plates so the profile of the base plates matches the perforated vent panel (the top, bottom and outside edges should like up exactly with the metal Prusa plate.

You can find that part here: https://www.printables.com/model/740982-enxlosure-back-clamps

POSTNOTE - THE DOOR IS STILL A JAR, OR MAYBE A BOTTLE, OR A CANNISTER. SOMETHING IS ROTTEN IN DENMARK AND IT"S NOT THE MODEL.

So after all this very careful attention to all these details, and even with my anti-racking clamp mod in place, my doors still didn't line up vertically.

There's no overlap due to all the other stuff I'd done as detailed above, everything fits exactly right and lines up but the one door is markedly down from the others.

(There's two pictures here that show the before and after).

At first I thought it was due to the door hinges, and the beam being inverted from the other beam and thus changing the "hang" point of the doors.

I could either rework the beams to have the same vertical orientaion as far as the bolt and nut to thus get the hang points the same on both sides.

Or I could redo the hinges and put in solid liners for the bolt "pin" location or enlarge the mount screw holes to allow some xy axis play room and hope that when torqued down they wouldn't move (much).

I seriously toyed with both notions.

Then it finally dawned on me.

What if the entire XL Frame is racked?

I never had this problem on the MK3S.

The initial build was on my wife's beloved kitchen table with it's 2" thick solid marble top.

Right after I'd completed my enclosure build of that I noticed that the during printing, not only was the MK3S 23+ db louder than it was on her table, but the entire cabinet (being on casters) would visibly and violently shake during printing.

I'd tried the paver routr but the surface finish on them (understandably) sucks for this application, and they were neither level, nor sure.

So I looked into finding some solid marble or granite scraps from a local cabinet fabrication place (my wife's brilliant suggestion).

While there were places that had scraps (and they were cheap) they were either too big or too small or too irregular to be useful.

They COULD take a way to large of piece an cut it down, but that would be inordinately expensive.

So my lovely wife stumbled upon the perfect solution:

A 4" thick piece of precision surfaced solid granite (it's actually a cutting stone for industrial knife sharpening purposes that has no more the 0.05 micron's of difference in surface height over the entire surface and weighs well over 100#).

This thing was an off the shelf part at woodcraft for $99.95.

Sold. Problem solved.

My MK3S is on about as good of a foundation as it can get and weighs less than 1/4 of what the XL does in a substantially smaller footprint.

Looking at my XL platform (an off the shelf Home Depot "Husky" variable height rollabout cart), while it has a "solid" 1" top it's hardly solid wood hardwood and further isn't cross laminated but rather has a veneer over what is probably MDF under the hood.

As such, it's more than susceptible to humidity variations and is probably not exactly flat.

Borrowing my neighbor's rotary framing laser I leveled the table up as best as possible then set the laser to sweep the surface, and sure as hell there's about 1.5mm of height delta across the surface that the XL sits on.

Surely, you'd think, it can't have that much of an impact?


Uh yeah, it does.

The XL's frame is very solid, is as well as it's pretty spot on accuracy of the cuts involved on the materials)

Couple that with my OCD about "lug nut pattern progressive tightening" of the vertical uprights and the core XY install on top of that when it should be sure.

However, when I put a shim under the low corner of the XL's footprint I found with the rotary laser?

The doors pulled into almost an exact match in the vertical offsets between them.

Further, I'd noticed during printing all this stuff that the entire printer has gotten markedly louder, particularly on certain moves (long diagonal xy's, straight front to back Y's and a few others).

Like a loud ass SKRONK where the whole thing resonates and visibly shudders.

So rather than try to find what is probably an insanely heavy and expensive piece of 24x24x4" solid granite, I've got a different solution for the long term.

I'll design a set of feet with known reference points at the bases that have some fairly fine pitch threads connecting to the actual supports under the XL, sort of like a more precision version of what you have under your washer and dryer.

Then I'll use those (again borrowing my neighbor's rotary laser) and get the feet base Level with respect to the table's top, and land the XL on that.

It doesn't matter if the table shifts or isn't exactly level, just that the supports under the corners of the printer are in the same plane AS the table so the weight of the printer doesn't pull it out of square.

Will keep ya'll posted on those developments. (edited)

How did you make that model? That's really neat.

Great model and design, prints well! Didn’t quite work to get me the angle on my trident but should be perfect on a V2

Seems to work quite well, although I had to recreate it and make some adjustments to fit my 12mm WS2815 strip (144 leds/m). The carrier snaps into place with a satisfying click, and can be removed by slighly bending the upper arm outwards (how well it works depends on how many walls you have there - more solid walls equals less bendiness). Printed with ASA (Fiberlogy Red), 0.4mm nozzle, 0.2mm layer height.

Hiya, just wondering if you're happy to share the stp file for this? Thanks

Resized for test input shaping on my prusa mini. nice!

Just tried this on the Prusa MINI with new 5.1.0 firmware!
Perfect print in astonishing speeds!