Belty Boi

14/03/2026 Started a new BOM 1.5h

Originally i had been planning for this to be a corexy belted printer however i had a change of mind and want to try make a croxy belted 3d printer, which will probably make things harder

if i run out of time, this might get turned into a voron 0.1 size croxy 3d printer, it really depends how things go.

Back to what i actually did

18/03/2026 started on a parametric frame 2h

Really just getting back into the swing of parametric CAD. Last time i did a bunch of things in ways that fusion didnt like, so this time im trying to do things as proper as i can to stop fusion getting annoyed at me.

I then realised mounting rails on the bottom would be wayyy smarter so changed that.

It was much quicker having allready done it once and knowing exactly what i need to use to achive the alignment. That being not the align too, but the constrain component tool. Last time i tried to make a parametric printer i made the mistake of using the alignt tool a bunch, but i didnt realise that that didnt work well with parametric models.

im quite pleased with some of the parametric stuff i did to add the right ammount of holes to the rails.

Was there an easier way? Probably. Did i have fun? yes!

My next task is to try figure out motor, and pulley mounting.

I will probably take inspiration from the Annex K3 for parts of the gantry

i loved this method for tensioning the belts, its so elegant and compact

21/03/2026 1 hour

You probably cant tell, but when i constrained all of rails, that was done relative to the extrusions, which werent propely constrained. So i had to make sure everything was constrained

well, it still wasnt properly constrained oops

i tried moving a carridge, and it was NOT happy image (1) so i tried to fix that, which then resulted in

So i figured, since i should have constrained the frame, THEN the rails, and my rails and frame were both super cooked, I should remove all the constraints on the rails, and re constrain them on the bottom instead of building it all upside down. Wow the above sentence needed more punctuation.

Anyway YIPEEEEEEEEEEEEEEEEEEEE

23/03/2026 gantry part 1, 1.5 hour

modelling in anything with gt2 belts is allways interesting, not difficult , it just looks messy

i think ive gotten somewhere with belt mounting, i didnt want to make a 1 for 1 copy of the anex K3 , however this is heavily inspired by it

below is an image of the crossection of the annex k3

i still need to model in screws on mine, the key difference is that mine is in a bit of a smaller footprint , and also the anex team wrapped the belt arround an m3 bolt, i added a 0.8mm wall thickness hollow cylinder arround a bolt because i feared the bolt cutting into or damaging the belt.

23/03/2026 gantry part 2, 2h

Im currently working on finalising which hardware i use and exactly how i mount the rail.

in the previous entry i was mainly focussing on making sure that everything went perfectly with the belts, this time im making sure everything goes perfectly with the rails, i REALYY need to make sure i grip these rails well.

So i plan on having 6 bolts to hold each rail in

a little detail i like was making sure that the heat set inserts on the bottom that some of those bolts screw into, are recessed so that they dont stick out and cause a misalignment against the carridge of the linear rail.

8 screws of varying length later and we have a hopefully well screwed together little assembly

a mirror and some constraints later and we get this!

25/03/2026 gantry part 3, 3h

im so happy its comming together

Some of the things im aiming to achive with the cad are

have the gantry be parametric, so i could scale it from 120x120 to 350x350 or whatever else i want
have all the constraints modelled in properly so that you can move it arround in CAD and check for interference issues

part of the reason im actually making it parametric is because im not exactly sure what my travel distances will be, since some area will be lost depending on the size of my toolhead, so i want to be able to compensate for that

the parametric aspect of the CAD works arround 40% of the time when it doesnt work re defining a handful of the rigid groups seems fix it which is odd to say the least

26/03/2026 5 hours

Trying to speed up the build a lot now

i threw together the mounting for the 625 bearings

i had been delaying this for quite a while because i wasnt sure where i was going to place the lower belt which would throw off everything.

the design wouldnt print the best in one peice , also wouldnt be very easy to use a vice to press fit the bearings

so ive split it into two parts

but i also believe alignment of these two parts is quite important , so in addition to the m4 bolts that will hold the halves together there will also be locating features

Ive added 0.2mm of clearance between the locating geometries so hopefully they should fit okay

oh yeah shoutout to autodesk for making custom length bolts a payed feature

i get the mild inconvenince of making my own parametric bolts, which really doesnt take long

It should be able to be copied and used for all the other once since i have “A config” and “B config” which flip which is the idler , and which is the pulley

spotting this wasnt great

however couplers like these exist which mean i dont need to redesign anything which is amazing

Realised i need to rework this as theses are far to close In the process of fixing the above issue i also discovered the holes for the motor attachment were slightly ofset, meaning i had to remove and readd the motor mounting holes. I am however glad that it was simply the motor mounting holes and not the motor which is aligned with a bunch of other stuff

ah this is a lot of errors

Given the ammount of errors i was getting I decided the best course of action would be to export the completed assemblies, like the ones for the motor / pulley mounting. Then go back to an earlier version of the file without the errors and import the assemblies to get a much cleaner doc,

i still need to import the rail holders again, but this looks a lot better, and takes way less time to compute when changing perameters

AHHHH 27/03/2026 6h

lovely how in fusion its shown as but once you dare to look at it it turns into this AHH fussssiiioonnnnn

it appears exporting as a step and then importing fixes the issues? im not exactly sure how that works but ok fusion

if we ignore bits of floating here and there its starting to look quite good

time to fix the floaties

Screws first

i could have just made a block, and i did but i felt like doing something a little fun just because ive been doing CAD for 6 hours straight at this point i know infil could have reduced the weight, but i just like hexagons

wow this is definetly comming together quickly now

sumarry of what ive done so far today

made progress on a bunch of stuff
broken a bunch of stuff

To elaborate what i meant when i said breaking a bunch of stuff, i managed to create circular dependancies, amongst other unpleasant things, which resulted in the file breaking or crashing any time i attempted to change the perameters of the frame.

So my solution to that was to once again export the completed assemblies as step files and then make a new file to recombine them all into again which fixed the issue

in the image below you can see me trying to make sketches to work out roughly where i was allowed to with my hotend without it coliding with the frame i could then export that sketch into the file where i was building my hotend

modelling in no go zones so i can rearrange things with more awareness for how it interacts with the frame

might just be the worst looking duct ive ever made, or ever seen but if it cools, it cools

i can prob use aux cooling if i really need

It needs to be at an angle because its a belted printer

i was also really limited on the fan placement because of the gantry soo there is probably a far better way to do it

28/03/2026 5 hours

Modelling the no go zones REALLYY helped me work out how to arange things.

ive tried to design the printed parts so that they are easy to print since the last printer i designed had some rather akward and support needing parts.

hopefully its not awful to print

Anddd i think thats the hotend hotend r1 hotend r2 hotend r3 hotend r4 hotend r5

the hotend took me a bit longer that i expected, but hopefully it should all join together well when i bring it back into the other file

29/03/2025 1h

stuff is all comming together now, well, we can ingore the parts that spontaneously broke

now its fixed i can check if my clearances are working out ok, if i did everything right the carridges should hit the end before anything on the hotend colides with the frame this one looks good

this one doesnt look like a pass however its a pass

this corner is not the best at all

then this corner is ok

overall i dont think i did a bad job

the moment of truth now is working out if i can change parameters without it breaking or crashing

it did it without crashing however it broke in some places again

an export an import later and i think its fixed

Some things i did on my hotend that i forgot to mention earlier

the first is im going to need to remove or use a hacksaw to make this tube shorter

the second is i plan on using superglue to help secure the 4020 fan i didnt believe the two screws one above the other would sufficiently secure the fan, so i added little spur to the side which i could use superglue, or double sided tape, or PVA etc

my plan for the duct at the moment is also just tape, i believe once i have the parts i will need to iterate on that in significant ways

Here are some renders of what its like now main r4 first render of it

29/03/2026 working on the bed/belt assembly, 3 hours

started working on belt assembly again

using those mesurements i can find the perimeter to be

93+93+ (( (360-153)/360) x 57 x pi) + ((153/360) x 12.8 x pi)

which evaluates to be 306mm so my minimum belt length is 306mm , and therefore my max should be arround 318mm

back to the rollers i could have left them as plain cylinders, however that feels rather boring

for the sillies i also decided that the spoke count, width and radii for the chamfers should be parametric

so putting all of that together we get this bed assebmyl current state

3/03/2026 belt time 8 hours

Im aiming to design some sort of mechanism between the two flats that push against each other to acomplish bed tensioning

more refinement

ive treally tried to keep these parts printable so far

im a bit scared of the shafts wiggling out of the bearings

this with a screw could be a solution, especially since it would add a splash of colour lil render whatever

i added some little inxdexing indents to help align the blue parts

i love how it looks, however adjusting the bed screws, or the motor belt tension looks like a rather large pain

I might just have to accept that

my plan for tensioning is to have two blocks, the screw really tighly into the frame, and then push against two of the belt mounts to slide them.

its not the most elegant, or elegant at all, however it should be functional

luckily the tensioner should be able to be removed after youve finished tensioning

i think this has just proved that there is very little space to mount the bed with this method of tensioning

I think a tensiong method that pulls the belts tight instead of pushing them should be better for the printer

I think this is good now, its bigger and less compact than the previous tensioning design, however it grants me more room to try and configure the bed mounting and other things.

extending out to the other side gets us here, ive decided to leave an 8mm gap between the tensioner and the bottom peice, i think this is enough room to tension the bed belt

Putting everything together gets us to here

weirdly though more stuff has become misaligned, and i cant work out which constraint has messed it up which is annoying

03/04/2026 continuing to try make things work , 4 hours

turned out fixing those screw alignment issues went a LOT deeper than i thought so im going to remove the hotbed assembly and try and do it again, hopefully a good chunk better

my strategy for restarting is export the thing that worked, and dont have much that can go wrong, then build from those

i thought the bed would be an easy thing to nab from the previous version but it turns out it had a ton of unessesary positionings that caused it to cause issues

ive fixed those issues now making it actually usable. It doesnt look very different but it is different in the fact it doesnt break everything if you import it into another project.

“It doesnt look very different but it is different in the fact it doesnt break everything if you import it into another project.” It still breaks everything somehow. Im not sure why fusion is remembering stuff about it, so im going to export it as a step and them import it to hopefully remove this sillyness.

Just exporting and reimporting the bed did not fix it, I believe its the belt thats causing issues for some reason. fusion throws errors like this when i try and export the belt which definetly means ive done something weirdly with constraints.

I believe the best course of action now is to export it as a STEP file, and then import it again, and reconstruct the rigid groups and constraints only where needed in a smarter way.

interesting how after importing it the only part that wasnt preserved was the belt. I think that reinforces my hypothesis that it was that belt causing the issues

20-30min later and ive rebuilt all of the assemblies without any issues

I just realised ive made an “interesting” choice, the drum that moves is the ones thats drive by the motor, so by moving that drum it adjusts the bed belt tension, but also the bed motor belt tension.

This just means that when tensioning the belt you need to remove the tension from the motor, and add it back after. This could be improved by moving the tensioning mechanism to the other side. For now i will stick to this design, however i plan on changing that in the future

I found a solution to the above issue. The two bed rollers are the same in every single way, nothing stops me from swapping them, so its a very quick fix which i will do now

before:

after:

A subtle difference that should save a headache down the road

at some point when making a 3d printer with belts, i allways seem to run into something that requires solving a maths problem

like the above one, in that problem the perimeter of the shape is given, and you need to find the length of the tangent.

that was a surprising chunk of maths to go through

i got it in this form where p is the perimeter of the shape, the length of the belt, and then x is the length of the tangent

for l=308 it solves to ~63.32

now that i know the belt loop size and i can make it in CAD i can place in the motor. my random picking of the bed orientiation seems to not be the best as one of the bed screws is going to interfere.

rotated it 180deg, fixed!

im working on getting something to hold the motor in place for the bed belt, i do need to make it look a bit better though

a small fillet goes a long way in making things look better

so does some colour

here is a render of how the bed assembly is comming along new new belt assembly v5

ive decided i want to see a bit of progress, so im putting together the gantry

wow waiting for fusion to compute stuff is terrifying, im so scared it will crash. After adjusting those ofsets fusion took 58 seconds to compute all of that, which is scaryily long for it to be unresponsive.

Ive also adjusted the positioning of the linear rails since there was some additional build that i could get by simply better positioning the rails

and a little bit more optimising later and the frame is 6mm smaller

component count is also starting to go up now its all in one file

wow it looks, like — something, something top heavy. oh no

the width of the bottom part is 170mm the width of the top part/gantry is 264mm

While attaching the gantry to the bed, i needed to add heat sets and holes into parts in the gantry, which meant i had to unlink the gantry. It took fusion 2minutes and 24 seconds to break the link which is, a while. Fusion manages to keep me on the edge of my chair

fusion has suddenly decided to be REALLY stupid

I think the best option now is to just export as a .step and import, and make it one big rigid body to prevet things moving. not the way i wanted to do it, but ive allready validated the build volume will be good in a different file.

On the upside i started on a klipper config while it tried to compute.

ive managed to get fusion to behave

currently im just trying to get an extra aluminium extrusion into the design so improve the rigidity of the printer

i did spot something questionable about my design, which is that the hotend in theory can print wayyy out away from under the bed, which i imagine probably would be great, however i do have a voron 0, so if i take the build plate from that and run two beds to extend it, it could work perfectly fine, that would also be pretty easy since the bed will just attach to the aluminium extrusion.

i did a chunk of BOM optimising, given the short time frame, JLC MC seems like a REALLY good choice for the linear rails, and because that means im allready getting stuff from there buying bearings and extrusions from there also saves a bunch of money, even if the shipping is expensive.

back to working on the frame

my strategy for joining everything together has been trying and get the holes in place with everything being 1 body, so its easier to reference features, and then use the split body tool to split things off after im done.

some work later and those extrusions are in which should make the printer a lot more rigid comming together

Its really taking shape now

The last functional thing to design is the bed mounting, but thats mostly done at this point, there is just a little bit left mounting the bed holder to the frame. Then its just the electronics enclosure and cleaning up this repo and gettting it shippable.

onto the bed mounting

I realised i could make the bed mounting plate bigger, as it shouldnt interfere with the tensioning mechanism for the belt as long as the bits i add arent super thick, which is fine by me!

Before:

After:

I had to be very careful with how i positioned the mounts as to not obstruct the tensioning for the belt for the bed motor

also made a few changes such as this ridge now attaches with bolts to enable a far better print orientation.

Electronics Bay

My idea is to get all of the electronics on one 3d printed plate and then go about attaching that somewhere.

Like this

Given this printer will have 48v i really need to take saftey seriously, its a printer that will be at a 3d printing convention, i need to make sure its safe.

My plan to do that is to have all of the mains in a seperate chamber, and then run the 24v and 48v into another chamber above it

Ive also added PSU brackets beacuse those are probably a useful thing to have, it gets a bit tricky without them.

my strategy at first was make a big box , however i might need to fly with it, so now its a big box that can flatpack down, which sort of made me restart in how i designed it, but its fine.

Im trying to make it a nice box at least, making the bolts recessed where possible, that should also help me if/when i mount it

Im oddly satisfied with the enclosure

Now i get to count all the screw for my bom!