HumbleT -- A humble Voron Trident build in the 150mm Tiny-T flavour

[sturdy ice]

HumbleT

As makers, we all have that one project we've always wanted to tackle -- the one that combines everything we've learned along the way. For me, that project is HumbleT, a downsized version of the Voron Trident. HumbleT is the result of years of experience in building, tinkering, and never ending modding. I wanted something small, powerful, and reliable -- a printer that meets my needs and reflects the approach I’ve come to appreciate.

HumbleT is my take on PrintersForAnt's Tiny-T, built around the trusty 2020 extrusions, keeping things compact yet sturdy. I’ve chosen parts and features that might seem a bit over the top, but they’re what I believe make a great printer. It’s meant to handle whatever I throw at it, producing high-quality prints without the long waits for preheating that bigger machines often require. This project isn’t just about building another printer -- it’s about creating something that truly works for me, using everything I’ve learned from past builds and allowing me to starting using all the gucci parts i acquired 🙂

This project wouldn’t have been possible without the support of the genuinely humble people I've met in this community. I've drawn on their modifications, wisdom, and ideas, and learned so much from their builds. Thank you all kind strangers for sharing your work, CAD source files, helping me with CAD adjustment and showing me where I can get all those cool things that I did not even knew exists!

Parts and note worthy elements (updated 2024.08):

• MCU: Manta M8P 2.0 with 5x BTT 5160T Pro, all of motion on 48v and RPi CM4 2GB WIFI compute module.
• PSUs: Dual Mean Well UHP 200W PSUs in the 24v and 48v variants.
• Frame: Misumi, speced as per Tiny-T BOM
• AB motors: custom made 55mm long D shaped shaft variant of OMC 17HS19-2504S-H-C12, shoutout to Sanity and Locki, mounted using Hartk's pinned mod and Frank.af's double sheer modification of motor mounts with a slight modification to the non-endstop model.
• Z motors: 3x V0 speced Moons motors in the TR8x8 variant, love how they look
• Printed parts: All inner printed parts in Addnorth Polycarbonite blend HT LCF (152'C Tg 😎 )
• XY joints:: Metal one that I got off Aliexpress, using bearing stack instead of teethed idlers
• Belts: POWGE 6mm Yellow one
• Pulleys:: from POWGE as well
• Rear panel: A VT's rear panel cropped to work with TinyT frame (shout out to vervurax)
• Exhaust cover: Modified version of NyxCode's Voron 2.4 Umbilical Exhaust Cover to have only PG7 passthough for reverse bowden, centred (shout out to vervurax)
• Build plate: The magbed from MandalaRoseWorks with 200W custom made edge-to-edge AC hearter from Keenovo (again, shout out to vervurax for helping with drawing it)
• Bearings: NSK's F695zz, gucci, managed to negotiate amazing price with distributor
• Rails: all Youmetong rails in the Z2 preload format. I honestly love those rails, I had only one HIWIN and I have few Honeybadger rails but those Youmetong are so great I really couldn't say which is better, cut to size by Youmetong and also got great price on the bulk order of all rails
• X axis: I decided to go with MGN12H on the X even though TinyT is speced for MGN9H, just so I can install the Voron Tap
• Probe: Vitali's Metal Tap
• Toolhead: XOL in the monobody variant, since it can be bolted to Vitali's TAP without any change
• Hotend: TriangleLabs Dragon Ace. Love those hotends, great flow without MZE and trustworthy thermistor readings with easy to route wires
• Extruder: G2SA with custom made 9t Moons pancake motor (shout out to Clee)

Lets goooooo!!!!1111111 (I am going to have SO much fun with this project :D)

Frame

To my surprise I was able to build entire frame in hour and half while preloading all the nuts as per the TinyT's CAD. What I learned from past builds is that you really want to use shims on blind joints. It makes them so much easier to build since the very last part of rotation of bolt does not relocates the extrusion. Also the absolute amazing Flystec's tool for the frame corners made squaring it a non-issue. My another nemesis were always overtightening screws and in process damaging threads or ripping bolts apart. To avoid it here I got myself adjustable torque screwdriver from WERA (0.3-1.3Nm) and started to fasten M3 bolts at 0.35Nm and the frame's blind joints at 1.2Nm. I also used 1mm 'brass precision spacers' that I got off Aliexpress that turned to not be very precise after all, but hey, they work amazing on blind joints!

[slender widget]

excited for this! never seen shims used in the blind joints before. looks cool and functional

[sturdy ice]

I cannot find where I read it but I think Misumi also recommends using shims. I recently started to use them everywhere. Even in recessed holes for M3 bolts I use M3x5x0.5mm. I found the shims help settle the bolt in and holds much more secure than it would otherwise when resting on printed part. This printer (and my in progress Micron+) have shims literally everywhere 🙂

[cursive pike]

Yay, looking forward to this build!

[surreal lava]

[marsh pier]

Looking forward to seeing this come together !

[obtuse elbow]

Oooft, this looks interesting!
Looking forward to the progress.
I saw on Steves stream once that he uses square nuts as shims in the blind joints, for the same reasons that you’ve listed here.

[brazen mirage]

I don't need another printer...
I don't need another printer...
I don't need another printer...

[sturdy ice]

You might end up being gifted a TinyT frame set so better start your budgeting now @brazen mirage

[river widget]

Which shims do you use for 2020 and 1515?

[sturdy ice]

For 2020 I used M5x10x1 brass one, For 1515 I use whatever I had in zip lock bag that I found during building the frame However it is for V0 rebuild, my Micron frame was built without shims in blind joints sadly, but I do not want to take it apart now. I would still recommend to try go with 1mm thick if they fit for you

[sturdy ice]

DIN rails? Where we are going we do not use DIN rails. (Sneak peak at electronics, panels, some more information, ...)

Most of parts will arrive during the coming weeks (PSUs, gucci bearings, tiny SSR) leaving the only one thing that I still not have being a bed heater. Keenovo should ship it in next few weeks hopefully. For now I've made sure my Y extrusions are at the just right dinstace from top using printed spacer and clamping them with some spare 2020 I had on me just to make sure they are as parallel to each other as they can. I am sure once I mount X beam one or two spots will be more difficult to move and I will loose one rail and make it tram to the other, after all I went with all Z2 preload rails. 🙂

Then the bed carrier. I went out of my way to align it perfectly to the rear extrusion, like actually perfectly, I rest it on the bottom panel posts (few mm below the side extrusions and tested if it is aligned with the rear beam. I did it few times until I got it perfect. It is absolutely against what I tell people that good enough is both good and enough, and I will never recommend what I did here to anyone, especially that you have wiggle room after you mount it later, but hey, that was fun! I also put this sick printed brace with total of 4 bolts, I think using 5 of them would be excessive. 🤣

Then the star of the show, electronics. No din rails! Thanks to the amazing work of @marsh pier I got my hand on the mount that uses 2020 extrusion instead of DIN rails. It does move it slightly below the deck panel and then with the combination of Annex K3's part I can mount dual UHP PSUs as well as an MCU there with Manta rays. I need to print those yet. Then in between the extrusion and first PSU (the only PSU in photo) I will have SSR, the tiny ED24C3, since with 200W AC heater I will be switching less than 1A on it, and it's good up to 3A (good thing t have 240V at the wall!).

Panels! I finally got my dibond 3mm panels, got them really cheap, 2x deck and 2x rear panel for total of 120 PLN (+36 PLN shipping). the rear panel is actually VT 250mm cropped to fit on TinyT speced frame. (Thanks @brazen mirage)

The last thing that I found interesting, turns out I actually can use the standard Voron Trident Z chain mount, all I need to do is mirror the lower mount for it. I wonder why TinyT changed it so much, perhaps for MGN9H the bed needs to go higher? Hard to tell, but from my very brief test fit, it will work just fine! I used the spare chains I had from the VT 250 build that I never used on XY, as I went with umbilical from get go. I really hope I will not run into issues of using magbed that will have sheet lower, if I will, I will add some neat spacers under M4 thumb nuts, but this is something for future Piotr to figure out.

[remote star]

Love to see another Tiny-t variant in the making! Idk how keen you are to take apart your frame... but I highly highly recommend you replace your shims for these actual blind joint washers https://www.aliexpress.us/item/3256804723997016.html and add these corner brackets wherever possible https://a.co/d/ffEyrEh.

I recently built my Tiny-t Sidepack using shims in the beginning and the frame still moved over time. The washers I linked are specifically made for blind joints and your frame won't move. After adding the washers and corner brackets, my frame feels extremely solid now

[marsh pier]

Really like how the corner wago mount fits in!

[sturdy ice]

That leaves me with only need for NDC wago so I can put it next to the SSR and suddenly everthing will fit! Will also see if I can squeeze one on the oppoite side of the corner wago

[sturdy ice]

Thanks for the information. I might grab them for another build, this one will stay with shims. For brackets I actually am remixxing one I found and like on Printables, the lower extrusions are held together by the legs (and at the front by motorm mounts) and for top I will use printed brackets, I have some on VT 250 and there no extrusion twist or move even if I really try so this will be all covered. I also noticed that the linked part are for M6, with 6.4mm opening, the screws are about 9-10mm so this should still work well. I will see them on VT 250 rebuild!

[remote star]

The M6 size will work just fine! Thats what I used on my build

[sturdy ice]

How much preload on linear rails? All of it!

I chose some fantastic rails from Youmetong on Aliexpress, all with Z2 preload. The GIF here is just a joke, as the rail cart won't roll down just from the force of hitting it with a wrench, but it does move incredibly smoothly. It's the smoothest rail I've ever used. The rails are packed with Mobilux EP1 grease and are very tight, which prevents them from free-falling under their own weight or the weight of the printed part. However, they still glide exceptionally well.

Working on the Y-axis will be a challenge. With rails this tight, there's no way they are perfectly parallel to each other, even though I've gone to great lengths to ensure the rails are mounted on the extrusions as accurately as possible. Later today, I’ll build the X beam and then fine-tune the alignment by loosening a few bolts on the right rail and letting it naturally align with the left one.

I absolutely love this project and find it so much fun!

(The gif is meant to paraphrase another one from @marsh pier 🙂)

[sturdy ice]

Two steps forwad, one step back

Yesterday, I set out to work on the electronics bay for HumbleT, but things didn't go as planned. I spent a considerable amount of time trying to wire the AC inlet, only to realize after damaging the third one that my 4.8mm crimps might be the issue. It wasn't until today that I discovered there are actually two different thicknesses for 4.8mm crimps: 0.5mm and 0.8mm. I mistakenly assumed this referred to insulation thickness, so I ignored it. As it turns out, my SSR requires 0.5mm crimps, while the AC inlet needs 0.8mm. This small detail has been quite frustrating. I ended up ordering more AC inlets and, unfortunately, more of the wrong 4.8x0.5 crimps, thinking it was a particular brand issue. An hour later, I finally ordered the correct 4.8x0.8 crimps. Thankfully, with TME being local to Poland, I should receive the correct ones tomorrow.

On a more positive note, the gantry is assembled! I opted for the ChaoticLab CNC XY joints, as I wasn't satisfied with the unbranded ones I had previously sourced from Aliexpress. All bearings are top-quality NSK with metal seals, and they spin much more smoothly than the RainDew bearings (though I never regreased the RainDew ones, so that could make a difference).

For the AB motor mounts, I used Hartk's pin mod with Frank.af's dual shear modification, along with custom-made D-shape long OMC motors. Combined with precision brass spacers, they look impressive and seem capable of handling much higher tension than standard motors.

The front idlers feature Hartk's pin mod of the stock Voron Trident front idlers, with my own modification to reduce the notch from 1mm to 0.4mm. This adjustment makes them much easier to snap in when printed with Addnorth PCCF and assembled.

Initially, I planned to use POWGE belts, but then I found Mellow offering GT3 belts at a good price, so I decided to go with those. I’m still figuring out the ideal tension; they are supposed to be run at 180 Hz, but I might just use the PF Makes belt tension meter and set it to 2.5 on its scale. I still went with POWGE pulleys on the motors.

I encountered some challenges with the Z chain placement. I wanted to use the regular VT chain location, but it interfered with the B motor at an awkward angle. I tried to modify the part in CAD, but it didn't work out. Fortunately, @marsh pier came to the rescue with a modified Z chain design and an updated mount for the MCU. The power supplies I chose are shorter than his, so his solution needed some adjustments for my setup and were kind enough to help me out, absolute MVP.

Overall, I’m excited about the next steps. It feels like I’m getting very close to having this printer up and running!

[sturdy ice]

Heater has arrived! New challenge -- Mandala Rose Work build plate seems to not follow TinyT's front holes position

Just had a mail man bring me the HumbleT custom 200W AC heater. It was modeled using the build plate model from TinyT repo, from the look of it the front holes are shifted closed to the middle of plate than it was supposed to be making the heater not compatible. However while it is a edge to edge heater, Keenovo did left a a solid 10mm radius from the hole without the heating coil, under right angle of light it is very visible where the heating coil is, with this in mid I will just use a sharp knife and make the holes 2-3mm deeper after confirming few times that indeed there are no coils int he area I will be cutting (visible under right angle of light and easy to feel with the pressure iof finger).

Cutting heater?! Safety first

I had a wall plug around with two wires that I used in the past to hook Tasmota Light Switches in a way that allow me to configure them before I install them in the wall, and confirm all is working as expected. so I just took it, 2x WAGO 221 422 and hooked the heater into wall for about 3s to take photo with thermal camera to confirm where are the coils and make sure I will not hurt them while widening the holes. Seems it will be easy. Done with P2 Pro USB-C thermal camera, nice thing, software wise it is not great, but much cheaper than alternatives. Also great to zee that it is a legit edge to edge heater. 🙂

[sturdy ice]

Heater adjustment was a success!

I got worried that becuase I heated it up the Lohmann glue might start to cure now, so I cleaned the magbed with IPA 99% and applied heater there alinging it without any tools or DMZ pins, since those would be less useful after I cut by hand the cutouts to make them bigger. Tested with M4 knurled nuts and everything is okay. Very nice, great success.

[wicked rose]

That was a genius use of a thermal camera! Now I need one.

[olive gale]

Super exciting build. Hoping to read more soon

[sturdy ice]

PSUs are finally mounted

I have a bit hard time focusing on this build lately, getting stressed out by all the things that are left to be done and eventually not doing anything. Figured I need to just start doing random things I know I will need and eventually I will get where I need to be. Installed two PSUs, 200W 48V and 200W 24V Mean Well UHP. That was quite a challenge to mount them, as the PSU closer to center have very difficult to access crew terminals once it is mounted, I had to wire it first then mount. Pushing wires into WAGO was also a difficult but now I have both AC and DC side wired along with Earth PE for frame (tested it with multimeter, all works as expected!). Next step will be to fit Manta M8P built with 2x Noctua A4x10 (The humblest fans you can get!) and then passing the wires there, I have just about enough space to go with HE screw terminal as SSR signal there.

I will also need to go without screen as I simply cannot fit it there unless I find alternative mount, for now headless + maybe two anti-vandal switches in the skirt as emergency stop and graceful shutdown button.

Wires might look thin but they are actually 0.75mm2, Heluflon in the teflon coating, very nice to work with. Stripping wires from them with automatic tool usually cuts off a handful of strands too so now my go-to method is to tap them with Engineer PA-06 using the 0.95mm cutout and then using fingernails to wiggle the coating off, this way I have no broken strands however if I try to use the tool to rip the insulation with at times I get ugly stretched strand of it left or if I use smaller one I get cut strands... so tap method it is.

For fork terminals I went with some quality one off TME, I had to get both M4 and M3 fork terminals, as the AC side uses M3 bolts and DC side uses M3.5, however the quality M4 fork terminals I have does not fit on the AC side at all, funny enough, the cheap no name one does fit there if fitted from the top without screw. Comparing it visually the cheap one I got as a kit with crimping tool are only half as thick as the one I got from TME. I have a thing for quality crimping terminals now.

[sturdy ice]

Brace yourself!

Some time ago I found a nice model on printables with corner braces and a step file. Modified it to suite my taste more (made it symmetrical, redid chamfers, enlarged holes among other things) and finally installed them using shiny black 'ultra flat' M5x10 bolts that past Piotr has ordered 'just in case it might be useful' 😎 Used M5 hammer T nuts that I sourced long time ago, as unfriendly as those can be at times, the M5 variant are really nice to have for something that you might want to later remove and not fight with extracting the roll-in nuts.

I hold such printed braces in high regards, have similar one in my VT 250 and they prevent the to extrusions from twisting when frarme is lifted by just one of them, highly recommend. Will tomorrow post remix to printables.

[sturdy ice]

PSA I've put the corner brace on printables if anyone is interested -- https://www.printables.com/model/1012482-2020-corner-brace-m5

[hardy cradle]

curious on why you went with the stock front tensioner instead of BFI since you have double shear on the motors and cnc xy carriage? wouldn't BFI get you more tension on the belts?

[sturdy ice]

This is the pin mod of those. I first went with BFI but the stock one have difficult to pop in notch so then I found the Tetsu usermod of them that has smaller notch and printed it just to realize all of them including VT one are the V2.4 specific that uses both M3 and M5, then I found on printables the mod of this mod but only as stl files that was proper M5 however they were too long, after adding M5 bolts I couldn't wish to mount front doors there. I could've use slicer cut tool to tame it but decided then that I am tired of it and I do not enjoy tensioning 2 bolts because I find it difficult to confirm that carrier is not angled and reverted to Hartk's pin mod and used it, the pins i had were too tight so I run them though drill with a sandpaper for a few moments and built it after reducing the stock notch from 1.0 to 0.4mm they pops in just fine. The notch is an issue as parts from Addnorth PCCF are unimaginably stiff. All of this gives me the single bolt to tension and I have high expectations for the durability of the Addnorth PCCF so this is the best out of the options I had. I might change them in future if they end up failing me but even with 6mm GT3 Gates belts I do not plan to go crazy high on the tension, maybe upper range of the PF Makes tensioner tool. 🙂

It is also just so much fun to mix and match things to see how they gets together, same applies as to why I went with 6mm belt if I dual sheer instead of 9mm. I just like it like that and the GT3 belts that I sourced supposedly stratch less than Gates GT2 and POWGE belts, time will show.

[hardy cradle]

oh i didn't know you printed it with PCCF so i think you will be good since it's much stiffer than ABS
and i get it it's a PAIN to balance 2 screw tensioner lol

the build looks good 👍 , and for this size of a printer i think 6mm is the best you don't need 9mm it's overkill imo

[sturdy ice]

Many parts of this build are overkill hehe, 2.5A long shaft dual sheer motors being one of them when belt path is just 120cm, but such is life when you build humble things

[sturdy ice]

Hotbed installed, toolhead updated

After a little break from tinkering with HumbleT, I finally made some progress! The past few weeks have been a rollercoaster ride with my VT -- dealing with motion system hiccups and even losing two hotends along the way. I knew it was time for a rebuild and update because the parts I printed with 3DO ASA GF warped badly at the motor mounts from the heat. To top it off, I got a bad spool of Addnorth PCCF that clogged the hotend and nozzle non-stop. After printing over 3 kg of this stuff without a hitch, I ended up with two completely clogged hotends that I couldn’t save. Once I finish HumbleT I will then finish Micron+ 180 and finally rebuild this VT.

Finally, I decided to crack open another spool of the same material, dried it out, and started printing again. To my relief, I didn’t run into any issues this time! It was clear that something was off with that faulty spool since the clogs hinted at uneven carbon fibers. I reached out to Addnorth but didn’t hear back. Luckily, since I got the spool from 3DJake, they reached out to Addnorth for me. They agreed to replace it or give me a refund, so I opted for a replacement. But given my experience and what I’ve heard from others, I think I’ll switch to ezPCCF instead, I did not heard any complains about quality of ezPCCF unlike Addnorth PCCF.

Now, let’s talk about the hotbed. It’s finally wired up and installed! I added three WAGO 221-422 connectors on each sides, and they look pretty slick. The IGUS polymer spherical bearings are held down with some ultra-flat M3x4 bolts. I decided to use M5x12 bolts instead of M5x10 for mounting the printed parts to the bed carrier because the shorter ones just didn’t feel sturdy enough bacause of the limited surface area. The 12 mm bolts feel just right, a bit too right, to give them some extra space I added M5x1x8 precision brass washers, which bumped the length from 10 to 11 mm in the end. I’ll be swapping these out for some nice black nickel-plated bolts once I get the 12 mm ones in.

For the bed surface, I went with a 16 cm HoneyBadger textured plate. It’s 1 cm bigger than the bed, which makes it easier to mount when not using any sheet stoppers, and it lets me tweak things a bit, which is great for TAP -- I will call it “analog TAP probe points fuzzing”

As for the hotend, I initially wanted to go with a Dragon Ace, but that didn’t pan out. The heater blocks were super soft on all my units, and I accidentally pulled out thread even though I was following the torque specs. After a bit of frustration, I switched to a Dragon UHF. I put one on my VT while I was fixing it up, using a 50W heater without MZE, making it a Dragon UHF Mini. It’s working like a champ, letting me print at a comfy 24 mm3/s with ASA. I tried pushing it to 26 mm3/s, but the heater was running at 90% PWM just to hold 265°C, which wasn’t ideal. So, I decided to stick with 24 mm3/s. I will eventually install there 70W heater instead but I really do not want to take it apart now when it's work since I know I need to completely rebuild that printer anyway -- I will let it be as it is now...

Now, looking at my HumbleT, I’m a bit worried that my hotend might not sit low enough since the VT-style Z chain is limiting how high the bed can go. I could switch to the stock TinyT Z chain since I have the three-point compatible one, but for now, I’ll just print the full Dragon UHF XOL to lower the nozzle. I’ll probably mount the Dragon UHF with its MZE there to ease my worries about the sideways mounted hotend possibly blocking some part cooling fan ducts with the silicone sock extending to sides in X axis. The longer sock that accounts for the MZE should be way less restrictive as MZE itself is so much smaller than entire heater block and with 70W heater it shouldn't run out of thermal capacity and potentially allow me for 28 or maybe even 30mm3/s of ASA, as with 0.5mm nozzle and 0.2mm layer it will be exactly 30mm3/s at 300mm/s print moves.

[sturdy ice]

Did some testing of be height. My VT 250 with LDO bed, graviflex sticker and HoneyBadger textured sheet is about 12.7mm thick when checked at the edge just before heater starts. My HumbleT bed from Mandala Rose Works (with embedded magnets) with HoneyBadger textured sheet is about 8.2mm (+/- 0.2mm when I check on it) meaning my print surface of HumbleT is about 4.5mm lower than on big Voron and this is why I will have hard time without the MZE added. I could raise the bed by adding some mods to the bed mounts but I am not sure if I really want to go this route now. Addng MZE will add enough to make it work

[sturdy ice]

German language have word that would describe me going with integrated mangets bed

Verschlimmbesserung
an attempted improvement that only makes things worse
I just cannot can I even 🤣

My humble options are

• Hope Metal Tap have toolhead a bit lower and does not require too much movement to trigger probe
• Use full height Dragon UHF with its MZE
• Add some more spacers on build plate to raise it by 4-5mm, currently using M4 nuts there, so I'd need to add on the top of it or something printed under it.
• Modify printed parts that connects to Z motor to have the notch for spheral bearings longer
  Raising bed will also means my neetly cut to length wires for wago might be too short and will require to slide wago mounts closer to rear by few mm.

[olive gale]

Thanks for the update, I enjoy reading these.

Interesting to read about the soft ace heater blocks. I have an upgrade to an ace hotend in the works, I hope it will not be a Verschlimmbesserung

[sturdy ice]

It is worth to be careful when tightening those guys with nozzles and inspecting the heaterblock if you can after you remove nozzle. If you stretch it it will end up getting loose on the heatbreak as it will no longer keep block from rattling around. If your heatbreak gets loose I would definitely suspect deformed heaterblock. Have 6 of replacement blocks to fix my hotends

[olive gale]

Oh boy. I hope this hotend will not be a lot of trouble. I should pick up a torque driver

[sturdy ice]

I feel like I should use the normal wrench and just make it finger tight instead or use my WERA adjustable torque screwdriver and bolt it to something like 1.0 Nm

[olive gale]

You have this one?

[sturdy ice]

Nope, 0.3-1.2 Nm WERA 074700

[sturdy ice]

If I only had 1 cm more to work with

The MCU is finally installed in the electronics bay, alongside the AC inlet and both PSUs and wired just enough to remote into Raspberry Pi CM4 there. I also managed to install DangerKlipper, Moonraker, and Mainsail. For motor drivers cooling, I will go with two Noctua A4x10 fans, while building it I noticed that the cooling shroud pressed a bit too hard on the screw terminal cover, so I had to remove it temporarily. In the future, I’ll likely modify the design and add a 2mm cutout to avoid interference, though even with that, access for maintenance might still require getting to the MCU from below. For now, I think I’ll just leave it as is.

The space... I'm constantly surprised by how tight the available space is. I still need to wire the motors, hotbed, and toolhead, and I can already tell it's going to be a bit of a spaghetti situation in there. Thankfully, the SSR is compact enough to fit perfectly in the gap between the center extrusion and the first PSU, so no issues there.

Then the gantry. I had a bit of a mishap while installing Vitali's Metal TAP. The carriage went flying, scattering bearing balls everywhere. The other 50mm rails I had for the TAP were Z2, which were a little too tight. I was concerned the toolhead wouldn’t be heavy enough to move smoothly, so I ordered some Z1 rails instead. Those should arrive tomorrow. Though I did installed Gates GT3 belts so progress there too!

As for the AB motor wires, I did not found any solution that I liked to hide those wires, the extrusion channel covers were a bit too restrictive for my taste so I figured I’ll be sleeving them with neon green sleeving instead. if I can't conceal them, I might as well make them stand out and look good. The more neon green the faster printer will be.

The outstanding points on my list remain:

• WIre hotbed (thermistor, heater, Earth PE)
• Toolhead (umbilical with 2x 0.75mm2 for power and 2x 0.25mm2 for CAN network)
• Wire the Z motors into MCU
• Sleeve and wire AB motors into MCU
• Noctau fans on MCU for motor drivers
• Side electronics fans, decided to go with 2x 6015 24v dual ball bearing GDSTime that I will run at 12v 100% PWM to make them quieter and not develop PWM cycling noise
• Side panels
• Rest of Vitali's Metal TAP
• XOL and G2SA
• SSR
• Skirts
• See if I can find a way to mount 12864 screen outside since I do not really have space for the stock VT/TT screen hinges
• Flashing M8P 2.0 and EBB36 with Klipper
• Doing my klipper config based on my VT 250 one
• Front doors

If I can stay motivated and chip away at this every evening, I think I can have it all finished by the end of the coming week. Or... I might end up shelving it for another week or two.

[marsh pier]

Looking good! For the display the PiTFT might fit as an alternative

[grand beacon]

That's really disgustingly clean looking...

[sturdy ice]

After I wire 5 motors, hotbed and toolhead I bet it will not be this clean, but since I can hide most wires under the PSUs it will not be awful I think

[marsh pier]

Also deck plates and bottom panels 🙂

[sturdy ice]

I need to order bottom plate! Might be a good moment to order dibond panels for future VT rebuild too and maybe Micron+ while doing so

[agile garnet]

You even color matched wagos, you absolute mad man

[hot cedar]

I used M5 washers instead of shims for the blind joints. I remember this trick from watching one of Steve's Trident builds. Since washers are made via stamping I found it easier to mate the raised side with the BHCS which made it easier to install the blind joint.

[wicked rose]

How’d the wiring turned out?

[sturdy ice]

Will be sharing update later this week. Racking up progress for another update post. 🙂

[sturdy ice]

Almost done!

Over the past few weeks, I haven’t had much time to work on the printer, but these last two days have been super productive! The entire printer is wired up except for the toolhead, and I’ve installed Vitalis Metal TAP. I’ve also completed the belted gantry and added rails as backers on top of the Y extrusions.

About the Y Extrusions

While checking over the Y extrusions, I noticed that I’d used M3x8 bolts, which fit just right for the MGN9H and Misumi extrusions. I started to worry about tolerances -- if any of the bolts were longer, they might not go deep enough or hold tight enough. So, I swapped them out for M3x6 bolts. While doing this, I realized one of the roll-in M3 nuts had failed. I tried to roll it out, didn't pan out, then using flat screwdrivers on each side, but even then, I couldn’t remove it. Finally, I had to resort to pliers, grabbing the nut from both sides and rotating, but unfortunately ended up damaging the Y extrusion. Lesson learned -- make sure to source good-quality roll-in nuts for future builds.

Thankfully, my past self anticipated something could go wrong, and I had a spare extrusions on hand. Dismantling and re-squaring the frame was a bit of a pain, but the corner tools from Fysetc helped. A printed tool then allowed me to set the Y extrusion height with the top one as a reference point, making sure the extrusion will not twist was challenging. I’d previously used wood clamps, but they’re a bit unwieldy, so I tried a new idea. I used my Knipex smooth parallel-jaw pliers along with DIN 875/2 squares on either side of the Y extrusion just as something that is flat and strong and I do not have to worry about force from pliers on, which let me clamp it securely while I fastened it in place. This approach worked perfectly, and everything was aligned in under a minute! When I reinstalled the X beam, I was pleasantly surprised to find that I didn’t need to tram the rails -- they moved smoothly together, despite all being Z2.

Electronics

On the electronics side, I’ve wired all the motors, SSR, and bed thermistor to the MCU. I also installed DangerKlipper and did some initial configuration. Two Noctua A4x10 fans are cooling the stepsticks, and all that’s left in electronics bay is wiring the umbilical, installing skirt fans, and fitting the remaining skirts.

I ran a PID tune to test the SSR, and the 200W AC heater on the 150mm bed is fantastic! The bed heats up quickly, and the top surface of the spring sheet is only about 2°C cooler than the set heater temperature. The lack of a magnetic sticker and the strong magnets in the build plate really seem to help with heat transfer. I think my preheating process will benefit significantly, as placing the bed at the bottom while the part cooling fans blow on it from above should help heat the circulating air quickly and evenly. Although this might make the PWM work harder, it should be great for the chamber overall, especially that without the magneting sheet the transfer of heat is so much less restricted.

Tap, tap, tap

For Vitali’s Metal TAP, I used a Honeybadger 50mm rail, carefully cleaned and greased with Mobilith SHC 220. Luckily, I had a spare MGN9 rail from Honeybadger, which gave me plenty of room to slide it back and forth

Next steps

On my quick next steps list remains

• Finish basic klipper config with all motors
• Re-tension belts, this time with motors holding after some movement, so the motors are not holding in some weird half-rotation position
• Install electronics bay fans and skirts
• Crimp umbilical
• Build G2SA and XOL toolhead
• Flash EBB36
• Install panels and doors

[cursive pike]

AtomicVectorPlotter, do you have any additional info on those PSU and MCU mounts or what you used to get the additional 2020 mounted?

I'd like to do something similar but maybe with a LDO Leviathan.

I'm loving the build PiotrK and it inspired me to do my Tiny-T.

[sturdy ice]

Dual PSU is unchanged from https://github.com/Annex-Engineering/Gasherbrum-K3/blob/main/Release_1_2/STLs/Electronics_Brackets/universal_dual_meanwell_uhp_psu_mounting_bracket_a_x1_rev2.STL and https://github.com/Annex-Engineering/Gasherbrum-K3/blob/main/Release_1_2/STLs/Electronics_Brackets/universal_dual_meanwell_uhp_psu_mounting_bracket_b_x1_rev2.STL and for rest best to highlight @marsh pier so he can notice ping 🙂

[cursive pike]

I didn’t want to @ as it was not super important. Thanks for the links.

[marsh pier]

This is the 2020 extrusion mount to replace the DIN rail. Controller mount is based on Ramalama2's Octopus Pro Fan Case but adapted for a Manta M8P, attached to the 2020 extrusion using the universal controller mount from the K3 (might be that I shortened it a tad)

[cursive pike]

Thanks. I’ll keep it as an option whenever I roll out the eBay.

[sturdy ice]

Back to the project

Been a while since I worked on HumbleT, had little motivation to touch it and pretty much shelved it until SMRRF. During the event I once again felt the desire to build it and now did few quick adjustments to get it into running condition.

Idlers and belts

The idlers I were using, Stock VTs but pinned, did not worked well with Xol so I swapped them to BFI, Then had a bit issues with tensioning belts, those are Gates GT3 6mm and I when toolhead was in the middle it gantry was deracked but with toolhead to either side it was getting all kind of weird. This has been addressed by adding more tension. I use the PF Makes belt tensioning tool and always were going with 2.1 or 2.2 on it's scale with Gates GT2 and Powge belts, however the moment I went to 2.5 at 150mm my issues with racking gantry suddenly stopped, interesting, but I do not feel like going too far understanding it, perhaps those belts just need more tension? Good thing that I have high quality bearings and dual shear motors.

Toolhead

I've built Xol with Honeybadger 2510 axial, Dragon UHF without MZE (Dragon UHF Mini), stock heatbreak and regular v6 0.5mm nozzle from tungsten carbide. Together with G2SA. All of toolhead are printed out of 3DO ASA CF, I had so many issues with Addnorth PC Blend HT LCF that I gave up on this material and just went with ASA CF for now. I will be switching to ezPCCF anyway soon enough. The toolhead is mounted on Vitali's Metal TAP and this also gave me some worries. Initially the toolhead were not going down by itself even under own weight, after few dozens of up and down it finally moves well. I suspect it is both the fact that this 50mm Honeybadger rail is very tight and also that I packed it with Mobilux Mobilith SHC 220 and slid few times on a longer honeybadger MGN9 rail before getting it back on the metal tap.

The EBB36 is mounted on a bit too long standoffs, this will be changed in near future when I will be changing toolhead, maybe to A4T.

Umbilical

For umbilical I went with 2x 0.75mm2 heluflon wires for 24v V+ and V- as well as 0.25mm2 for CAN L and H, that should pretty much solve any possible voltage drops even if I put a high current heater there. In the past I did had issues with EBB36 running low on power when 125W Rapido2 heater was going to temperature, that was on machine that I had less than ideal power delivery and alos 0.50mm2 for power.

Leadscrew Z motors

I went with Moons motors as I love how they look, what I did not anticipated is that those guys are 200mm of thread and then about 5-8mm of this non-threaded part. Xol crashes into it at the front. Best idea I have is to just modify the front motor mounts, add 5mm there and use longer bolts. I have plenty of space there since those are the tiny motors and will also be nice first print on itself out of ezPCCF.

What's next

Now that I have it somewhat running, in the sense that all is wired and all the MCUs are visible in Kalico I need to finish configuration file with pins, setup sensorless, setup and test TAP that I never used before and put some panels on it. I feel like I might push out a Voron Cube tomorrow.

Then some other fun things, like installing rest of skirts with my modified fan mounts, replacing front motor mounts, adding exhaust cover, printing MF Nano and so much more!

[cursive pike]

Maybe just add a 5mm spacer to the mounts instead of reprinting the full one?

GT3 does require much higher tension for proper registration.

[sturdy ice]

Nah, I need something to test this machine anyway, that might be a good real test with ezPCCF

[sturdy ice]

It works. It. Freaking. Works. Guys!

All right, so Humble-T is operational. Prints, screams, melts plastic, all the good stuff.

Around the time my VT250 broke down I only had a V0 on my hand and this is what I used to print Xol, out of 3DO ASA CF. Yesterday I noticed my Xol getting loose, I had it before but mostly with 3DO ASA GF one. Not a good sign. On the other hand the motor mounts out of Addnorth PC Blend HT LCF are in great shape, not a single bolt got loose and motors read on the outside low 90'C with thermal camera. Very nice. I push them at 1.4A

Right after I got it running I was surprised by how violent and loud 5160 Pros at 24V are on the Z motors. Enabling autotune made them even more loud, even in the silent profile. I gave out on them there and quickly swapped to some 2240 and holy smokes it went from "I cannot be in the same flat as this printer when Z moves" to "I think I can sleep next to this printer as long as only Z make noise". In the future I think I will abandon 5160 Pros all together and go solo with 2240, for AB motors I will hook 36V power supply as the HV option and first to get it will be most likely my Micron, or VT250 that I will be rebuilding.

The TinyT doors are really weird one. I though my front door panel was cut wrong but not, it was cut to spec out of DXF, using the very DXF from TinyT repo, but it somehow was too tight to close doors when magnet holders were jamming into it. I took dremel, the bit meant for side cutting/grinding and added a recess there. Does not look great but is 100% hidden by the handle/magnet holder so that's not a issue. What is an issue is that the hinge have 2mm thick part, so I need 2mm foam tape, which I do not, I have only 1mm and 3mm. Used 2 layers of 1mm one as a stop-gap solution then noticed that front doors magnet holder that goes on the opposide side are taller than that and want a 3mm foam there. I am not sure if my printed parts are way out of spec or the design could use adjustment but I gave up and ordered extrusions from Misumi for click-clack doors, so I will be installing those in a week or two when I get all other parts for it.

Already printed some ezPCCF parts, 5mm longer front Z motor mounts to lower those guys so Xol stops running into the tip of the leadscrews at the front edges, will be installing those tomorrow maybe.

Currently I am using Annex/Voron 2.2 panel clips and I like how they compress the foam and hold panels but the look of them is just not my thing. I will be looking for some alternatives in this regard to hold panels later.

The only thing that annoys me is the Honeybadger 2510 in Xol. I have same fan in the Yavoth toolheads and those are okay, but in Xol it seems to get in weird turbulance whine noise that I just cannot handle. I have some Delta 2510 to pick up tomorrow and will definitely be swapping those soon.

Right now I am printing some more parts out of ezPCCF, new toolhead, that I will rebuild with Delta 2510 and update my build, all in all this build is everything I was hoping for, freaking amazing printer I am telling you all!

Next steps will be toolhead, replacing front motor mounts, adding skirt fans and rest of skirts, adding lets once those arrive and new panel clips. I will be updating this thread as those things go.

I'd like to thenk all of the people that aided me on this project, it wouldn't be possible without your involvement, custom parts, random CAD help and words of encouragement. I love this printer and am very glad I got to build it. Cheers guys!

[sturdy ice]

From other things that I just freaking love in this build is AC mag bed. The speed at which top plate gets the heat is unparalleled, which also means my preheat routine where part cooling fans blast air at the top plate much faster exchange the heat with air (faster gets back to higher surface temperature), the temperature of plate is much more even according to thermal camera and after print I can slide the plate off instead of lifting it up and sometimes force-detaching prints. I use single-sided honeybadger plates where the non-printable side have smooth finish from me going on it with steel wool 0000, does not seems to damage surface of magbed this way and I can quickly swap plates and resume printing

[surreal lava]

Looks great! GJ!

[sturdy ice]

Hotend failure. heater cartridge failed after single print in 300'C.

One of those moments when a working printer is not end game, it is just point in time on the journey. While chasing a way to get as good layers with ezPCCF as possible I somehow managed to kill heater cartridge. It is 70W from Mellow and rated at 350'C. Chances are that it really did go over it in Dragon UHF Mini to maintain 300'C reading on the termistor on the other end.

The toolhead was supposedly to be rebuilt anyway, since I printed it out of 3DO ASA CF. I had parts printed out of ezPCCF on hand, all but EBB36 carrier which I will just use the 3DO ASA CF one. The Honeybadger 2510 fan went into awful air turbulence and I already have a recommended 5v Delta fan with tacho on me.

New hotend, kind of

I am still going with Dragon UHF Mini. Initially I wanted to go with Dragon UUHF Mini, which is just a Dragon HF heatbreak and either a volcano nozzle or 8.5mm volcano adapter and regular nozzle. The more I was thinking about it yesterday after taking hotend apart the more I did not wanted to go this route. The draft shield on the Dragon HF heatbreak rattles around and it would annoy me just knowing it is there. Instead I decided to stay with the stock long copper heatbreak that in a way could be called reverse Volcano nozzle due to it's length.

This gave me idea, if I murdered the heater due to temperature delta between thermistor reading and heater itself I will just put Boron Nitride thermal interface material that I sourced some time ago from Slice to the heater and heatbreak to hopefully replace the air gaps with TIM that will better move heat from heater block to melt zone, hopefully helping with melting capabilities, and since I will not be removing heatbreak when swapping nozzles, it kind of make sense. Boron Nitride paste is rated for over 800'C operation temperature so all should be good.

I had a 100W Mellow heater cartridge that from what I see is rated for 550'C, so that should give me some breathing room and hopefully not kill it. While at it I decided to replace the heater wires. The silicone white wires are really thick, breaks the blackish aesthetic printer and a bit annoying to route. I so happen to have BN 0.5mm crimps and tool for them that I wanted to test. Initially I took one of the 70W/350'C heater and tested it there, everything went okay and it survived pull test just fine. I was unable to rip it off which gives me confidence it will work well. Eventually I crimped it onto the 100W heater and installed in new Xol out of ezPCCF together with 0.5mm2 heluflon as heater wire, So much thinner and easier to route, and on the top of that, all black.

Currently...

Now I am in process of building the G2SA out of ezPCCF parts, carefully opened the filament path with 2mm drill bit to ensure it will work as specced (I prefer drill bit than sliding a filament though it really), coated the Xol bolts with Vibratile VC3 and about to replace the OptoTap PCB printed part with one from ezPCCF as well as slightly longer bolts on both the OptoTap PCB side and the one that bolts it to Metal TAP to give me as much surface area as I can for thread. My 3DO ASA CF printed part got a bit loose on the self-tapping side but I am pretty sure it is from me yanking it around trying to remove the connector, which I eventually failed to do and instead disconnected it on EBB36 side.

[surreal lava]

https://tenor.com/view/cereal-breakfast-national-cereal-day-gif-17915350

[sturdy ice]

Aww yisss. HumbleT got VT.1838 serial

[sturdy ice]

Panel clips

Initially I went with the Annex / Voron 2.2 kind panel holders but those not only did not fit my taste, they were also pretty weak, perhaps because they wer from ASA GF that is not very flexible in the first place. I decided to see some other options and eventually I arrived at conclusion to just go with the regular VT panel clips. However I wanted to seal the chamber better and dampen some noise from the inside so I remixed them a bit, took the 6mm clips and made them be fit for 5.5mm instead, this way with 3mm polycarbonite panels and 3mm foam tape I am quite heavly compress them, I did not installed chamber thermistors yet but it is very hot on the inside and frame on the outside is borderline to hot to touch too.

After toolhead rebuild

After 70W Mellow heater failed me from print in 300'C I now rebuilt the toolhead out of ezPCCF with new 100W Mellow heater rated for 550'C and things are looking great. Entire day I've been printing ezPCCF at 295'C and not only heater did not failed, the average PWM for hotend heater is 42%, so about 42W top maintain that 295'C, this is very nice.

While I was building the toolhead things got a bit weird. I broke off my newly replaced heater wires. Turns out it was not the crimp that failed but instead the heluflon right above crimp broke off. It is entirely on me since I did not put heatshrink there that would limit it's wiggle, I do not trust the heatshrinks I have on hand to work this close to a heater. if I manage to find some high temp heatshrink tubes I might attempt another swap but this time I just went with the stock white super flexible wires and another heater.

While I was installing Dragon UHF Mini I took it apart and coated the long-heatbreak-meltzone-insert with Slice's Boron Nitride, this is TIM rated for 800'C and Slice uses it on their hotends for heaters, thermistors and nozzles so I figured there's absolutely no reason for me to not go this route as well, since bottom part of heatbreak function as meltzone, I will not be removing it often like nozzle, so no reason to not use TIM, that would later act somewhat like threadlocker too. I have no point of reference given that the previous heater gave up on me and was of different kind, but I see that my ezPCCF prints amazingly so chances are it was a good choice.

Lastly, I've coated the toolhead bolts with Vibratile VC3. The one I had on me was really thick as solvent evaporated so I mailed the European distributor of them, the PLTight from Germany and they responded within hour that I can control viscosity by adding either acetone or MEK. I do not have any MEK on me and I rather avoid it, but I had acetone, sure enough adding about what looked like 1/5 of acetone into my VC3 and mixing it I had now very thin liquid that was perfect for coating M3 bolts. so far everything seems to hold very tight with those bolts and ezPCCF so I am full of hope it will not suddenly get all loose on me.

For toolhead I also went with 15mm standoffs and 1x 1mm brass shims, this gives me about 0.5mm-0.75mm of space between motor and the toolhead PCB carrier and allowed me to crimp all the wires without any pigtails, which is really nice.

The new hotend fan, 5v Delta ASB02505SHA-AY6B is freaking awesome. Not only it seems to move more air, just by checking how much it blows with an hand, but also is like 1/10th of the noise the Honeybadger 2510 made. I will definitely use only those Delta fans as hotend axial fans moving forward.

Front motor mounts +5mm

I also swapped the front motor mounts with a parts that are 5mm longer, to lower the motors itself. It did worked great and now I cannot hit the tip of leadscrew with Xol even when I am at the very front corner, Y-3 X-3 or so, however there's like paper thin space between leadscrew and Xol at those coordinates, if I were to do it again I think I would lower it by 2-3 mm more. When installing those motor mounts I realized I need to slightly slide them in from the side as the front Z rails were about 2mm above them, this worked without issue now but I wonder if 3mm longer variant would too. To hold the bed mid air I used bottle of Ketchup as a jack. Worked well.

Next steps

I wanted to fix my doors, the front frame magnet holders were all kind of weird, the recess in Misumi 2020 extrusions, the BOM one for TinyT, is 6mm however the standoffs for the magnet holders were wrong, a single printed part have two standoffs and one of them had 5.3mm and another 4.8mm, they were not even the same length! When bolted down they would tilt to the inside and then the doors would sit all weird there. I quick hacked the magnet holders out of CAD and printed however after just one day my 3DO ASA GF got so moist that entire print was overextruded and looking terrible. I thrown that spool into dryier and will resume tomorrow with the print, perhaps also mount the skirts and side fans since I have no other plans for it for now, it is pretty much feature complated build, maybe beside some thermistors and MF Nano

[sturdy ice]

Goodbye TMC 5160 Pro, you loud, violent sonuvabitch

I gave up on TMC 5160 Pros and 48v all together. Those drivers are way too violent and loud at low speeds. whenever doing first layer or small perimeters they had so much vibrations and noise that made me think that my bearings are failing apart, exactly the same noise as my VT250 does but there I have very different motors, bearings and belts yet the noise was exactly the same. Went ahead and swapped Mean Well UHP 200 48 to 36v variant that was meant for my Micron+, swapped last two 5160 Pros to 2240 and everything started to be so quiet. My sensorless works at exactly the same settings (80mm/s at 0.5A on 2.5A rated OMC 2504s), or it started to work right after I swapped diag1 with diag0 pin in the config.

TMC 2240 are just head and shoulder above any other driver and I will be selling out all my 48v PSUs and 5160 Pros that I still have.

Skirts and fans

Skirts are finally installed together with fans. I am limited with the skirt fans to just 20mm all together. This is because the recessed middle 2020 extrusion that all in electronics is bolted on is right at the fans location. This made me source a pair of GDSTime 6015 dual ball bearing fans.

I've then remixed the TinyT fan mounts that offset fans by about 10mm into the chamber, most likely to deal with air turbulences but this is no way for me to use it, instead I made fan bolt to the printed part without heatsets in part but instead with heatsets in the fans itself, much like standard VT would have. Problem started with the intake fan, the air turbulence whine was too much. I played around and I'd need to offset the printed grill by at least 7mm to not get this awful noise, but then the skirt would look not that great.

Past Piotr did ordered a bunch of 60mm fan grills, I eventually ended up printing the fan mounts without grill at all and bolting grills instead, no noise at all and the offset is not that huge either. I potentially could make it work with printed part but I also lack understanding of the fan blades profile and how to design it to not have air turbulences so I just did not bothered in the end. I will source some proper black grills from Aliexpress and swap the silver BHCS to some nicer looking black one and call it a day.

I am also not a huge fan of the PWM modulation noise that fans make when running with PWM, simply running fan at 100% is so much quieter than running at 99%. Some fans do not have this noise, or perhaps one can tune the PWM parameters in klipper to not have it? Either way the route I went with was to set the voltage on the skirt fans to 12v. Driving 24v fans at 12v 100% PWM makes them reasonably quiet and they still move surprisingly huge amount of air. Doing unscientific test with a strip of paper I even had airflow around 36v psu. I bet it will be even better then moment I mount bottom-most panel there.

Electronics with skirts mounted

Fan wires and canbus toolhead could use some shortening, will get to it eventually

Squeaky leadscrews

This one give me some worries. whenever bed was moving from Z5 to Z10 I had weird squery noise there when printer was heatsoaked. No amount of tramming bed helped here, it was the front right motor. I did loosened the 4 bolts holding it, slid bed few times, let it find it's most comfortable position, did not helped at all. I eventually just took some unblessed greasae, Mobilux Mobilith SHC 220 that I use in linear rails and put some of it on the leadscrews, all 3 of them, had bed ride up and down with a looped macro 512 times, then wiped the goo from leadscrews and top of POM nut and surprise surprise no squeaky noise. This is both good and bad. Good that I got rid of the problem, bad that I had this problem with supposedly POM nuts. I have no idea what material it is really made from and perhaps I should source some gucci PEEK TR8x8 nuts. I need to find someone that sells those. To the best of my knowledge, only LDO has them, but like 4x more expensive than quality TR8x8 nuts directly from OMC. I currently run some Aliexpress randoms

Kinda finished, kinda

There are two outstanding things that I still want to do here.

• Chamber lights. I already have Rainbow on a stick, so I will need to figure wiring and haul the wires for it to the very top.
• Chamber fans / air filters. KyleGB was kind enough to take a look at TinyT CAD and provide me side mount for 3DPF for it. This should solve both the air circulation during preheat and also deal with some smell from the printing.