PitStop 3 multi-material 3D printer in-depth design. 10+ materials at once

PitStop 3 is a 3D printer and tool changer that supports 10 to 20 materials at the same time, all with identical or different hotends and nozzles. Diving into various design decisions and the reasons behind them.
Support this project on Patreon: / mihaidesigns
Chapters:
0:00 Project overview
0:44 Old PitStop 2 project
1:56 Project goals
2:31 Tool changer overview
3:17 Universal tool design
5:41 Existing solution for multi-material
7:18 Reasons to have so many hotends
8:22 A 3D printer that just works
9:57 Extruder design
15:25 3D printer with or without the tool changer
23:17 Probe as a tool for non-planar printing
26:25 Spools and guiding PTFE tubes
28:33 3D printer design
30:59 Electronics panel
34:41 Nozzle probe
35:36 Parametric project
37:35 Adjustable lead screw mounts
39:06 Swapping parts with ease
40:03 Power and safety
42:52 Panic button
43:47 Assembling and testing the printer
44:46 Tool changer integration and custom firmware
47:43 How you can support me with my project
Twitter: / mihaidesigns
Instagram: / mihaidesigns
Buy me a coffee mihaid.com/coffee
mihaidesigns.com/
#mihaidesigns

[@rohitnalluri]

This is highly impressive, Mihai. This is not a project - it's an entire thesis and you should be awarded a PHd for it. My best wishes to you, and keep smiling! 🌞

[@marsgizmo]

impressive progress, great job! 👏😎

[@MihaiDesigns]

Thanks @marsgizmo!

[@danielletterman6562]

Are y'all actually brothers.

[@AdrianDemetrescu]

Wow! Impressive design, please make videos of you assembling it and solving issues in the design along the way - it would be very interesting to watch! Also, if you do want to make your own firmware, this software engineer(me) recommends studying existing ones(Marlin etc), as many people have put many hours into making them - it can help you make faster progress and learn from their solutions.

[@MihaiDesigns]

Yup, already looking into the basic architecture of existing firmware solutions. I intend to do shorter updates with the challenges I'm facing.

[@FilamentStories]

I was fascinated through the entire video. Wow, you have made such huge leaps forward with this massive undertaking. I can’t wait to see what comes next. Always a pleasure watching your videos, Mihai!

[@MihaiDesigns]

Glad you enjoyed it 😃

[@ThePapaVader]

Really impressive! What impresses me the most is the scalability of all of it, from a manual change to 20 automated hot end swaps! Brilliant!

[@MihaiDesigns]

Thank you!

[@texasermd1]

Amazing job with this design. Super excited to see it in person. 👍🏼👍🏼

[@MihaiDesigns]

Thanks John! Can't wait to have it built myself.

[@Vez3D]

wooooa my man !!! this is huge project. Very awesome. The details are just crazy. You have put TONS of work here. Cant wait to see you rocking this one.

[@MihaiDesigns]

Hey, thanks a ton Vez! Yeah, I'm super excited about it 😁 Stay awesome! 😎

[@scrimpyd]

This is a fantastic design. I enjoy building printers and this will definitely be added to my list of projects. Truly inspired design. Great job!

[@NeilLund]

Mihai, thank you for sharing your ideas. I would like to encourage you to not limit yourself by labels. Learning is fun and enjoyable when you take the labels away and just do all the things as needed to get to whatever next step you need to do on the project you’re working on. Beautiful work so far. I can see this is a labor of love for you.

[@MihaiDesigns]

That is indeed my strategy. Always trying out things that may not be the optimal move, but which provide answers I can use later on on my projects.

[@edwinirizarry9277]

Wow sir love to see more of the total concept! Looks amazing

[@mab4110]

Amazing! Really hope to start seeing videos on this very soon

[@bridevalley]

An awesome project! I look forward to learning more as you begin the build. Thank you for sharing.

[@MihaiDesigns]

Thanks for watching!

[@johboh]

Very nice! Excited to watch the build series 👍

[@changerdevie5421]

Very amazing, i was hypnotised during all your video. Really deep and well thinked project. I'm eagger to see the 3D printer finished and working.

[@ricardoresendes2859]

This is amazing! Very excited to follow along on this build. = )

[@jakeharms1386]

Love this! This is the type of innovation that will really change the space.

[@tvathome562]

That is the work of a man truly following his passion, so many complex problems for one person, I take my hat off to you.

[@Cretten]

Welcome back wondered were you had been. Love your videos!

[@MihaiDesigns]

Hey, thanks! Been busy working on the design :)

[@thehappyextruder7178]

Great implementations of simple processes, keep up the great work !!

[@TheITsoftware]

Great job! 👏 The swappable bed level sensor is wonderful 🤩

[@Marc42]

This looks seriously awesome!

[@bernhard_derProtoTyp]

extremely cool! can´t wait to see this built!

[@Repkord]

Incredibly well thought out design. Fantastic work! I look forward to hearing more about your progress hopefully in person next month at FormNext!

[@hatless-cluncky-capsize]

Great work! Thank you for sharing your work!

[@AndreaNassisi]

Wow, it is excellent to see a well-thought design! Looking forward to seeing the build.

[@MihaiDesigns]

Thanks!

[@Creative_Electronics]

I watched the full video and you did an incredible job! I'm looking forward to see your whole printer and pitstop design working! I'm right now working on a new fully 3D printed RC car, for me it's already version 4 and I'm designing it this time to be modular :)

[@MihaiDesigns]

Sounds like a cool project. Best of luck! 🤞

[@nickbuhling]

Awesome project! Can't wait to see it in action.

[@MihaiDesigns]

Me too 😀

[@emberprototypes]

Very cool project! I love that you've integrated a probe for XY and Z offsets. If it ends up taking up too much space you can consider using a camera assisted method for the XY that can be taken on and off the bed instead.

[@t1mmy13]

I've been wondering what bigger thing you've been working on for a while now, with all the cool little stuff you've shown. This is very interesting!

[@lukaswernetz3258]

I've just watched the whole video! Soo genius! You might become the future Prusa.

Impressive and incredibly well-designed! I'll definitely want to see it being built!

My biggest problem with this design is that it is eminently not enclosable. And building a tool changer that cannot handle ABS is maybe not the best idea?

[@ruzzcraze1862]

@ many people i know have never printed with abs. Small tradeoff to make for many people. As you said its a cool design, I also want to see him build it.

[@lukaswernetz3258]

Extremly skilled and amazing Design!!! Keep Up!

[@MihaiDesigns]

Thank you! Cheers!

[@MakerMeraki]

You are absolutely MAD, and i love it!

[@teitgenengineering]

This really cool when you told me you where working on a tool charger for that printer i didn't think it would be this impressive

[@valcocora5689]

Nice design, Mihai. Bafta!

[@KR33tzu]

Bravo ! Foarte tare ideea . Te urmaresc de ceva timp si apreciez munca ta . Astept cu nerabdare urmatorul update :)

[@Ahadun]

I'm speechless. I love it

[@THEOGGUNSHOW]

This must be your Magnum Opus! Such a sublime and intuitive design. Kudos 👏 my friend on a great accomplishment.

[@iamrobot9319]

Great, so much work on this project... Very impressive... Keep going :)

[@MihaiDesigns]

Thanks a lot!

[@user-lj1us6iv2v]

Hell yeah continue the great work 👍

[@marcus3d]

With CoreXY you could also have kinematic bed coupling (like the RatRigs), enabling tilt during non-planar printing like your printhead already supports. Because if you do non-planar printing with the nozzle at a different angle you'll easily get very bad results, depending on the direction of the movement (going uphill vs going downhill vs going across). Even with airbrush nozzles it matters if the filament is extruded to one direction or another. You really need either the nozzle to tilt or the bed to tilt or it'll be utter hell trying to slice it (even more so than non-planar slicing is anyway). Also, with CoreXY it would be easier to make an enclosure, for printing tricky filaments that easily warp if they cool down too much.

[@MihaiDesigns]

Yeah, that's the plan for a future version, but it'll be a little while before we get there.

[@DiomedesDominguez]

23:11 keep it simple by setting that every hotend has their own position in the stand by station, hotend 1 to the position 1, hotend 2 to the position 2, always. Another thing: trying to keep pre-heated hotend will create a lot of complexity because it's too far from the printer (during the movement the temperature will drop, some filaments will drop because of remnant pressure in the hotend, having such small amount of pre-heated nozzles will require a lot of movement creating waste of time just in the standby side, ...), and even if you resolve them, the user will need to keep in mind specifics gcodes for their specific filament creating more tuning than necessary. I suggest you to keep it simple by watching how the Prusa XL will do it: first you should think in a CoreXY printer to reduce complexity and have a lot of optimizations like using a single power source instead of two and not using bed slinger that could cause a lot of artifacts in the prints. Aaaannnd you made bassically a big resolution for a lot of problems explaining the nozzle probe at 34:45, so you can forget almost everything I already said, but the CoreXY approach must be implemented.

[@MihaiDesigns]

CoreXY planned for early next year, after I get this working.

[@ahmetemindilben]

This is incredible!

[@mattprobst6911]

This is awesome. I would totally buy it.

[@Golf3gti1]

Tolle Arbeit. Bin sehr gespannt / Great work. Am very excited

[@janzugic6798]

badass. genius design!

[@kamilx6188]

Woow, awsome project 👌

[@iAmTheSquidThing]

I'm glad to see someone researching ways to reduce waste in multi-material prints.

[@MihaiDesigns]

Big thank you! Yes, I believe the future is to reduce waste and I'm doing my best to move the technology forward.

[@aron.mp4]

Amazing! Even more people should watch your content! Great job Mihai!

[@MihaiDesigns]

Thank you so much! Just starting the build for the newer version.

[@ycy-technologies]

brilliant design. thank you for sharing!

[@MihaiDesigns]

Thank you too!

[@thebuildbay]

Impressive engineering! I tip my hat to your Fusion 360 skills 👍🏼

[@MihaiDesigns]

Wow, thanks! I should perhaps do a video on how I structure projects in Fusion.

[@bridevalley]

@@MihaiDesigns I would find that very interesting! Please do it!!

[@thebuildbay]

@@MihaiDesigns Absolutely!

[@CraigParkerD]

I love this project idea! I hope it’s designed so that you can start with 2 or 3 and build up. Love this!!

[@MihaiDesigns]

Yup, you can have fewer hotends and add more over time if needed.

[@CraigParkerD]

@@MihaiDesigns can you post when you have documentation up

[@bw_boulder_problem3878]

Impressive. Thanks for Sharing.

[@anastasiosgoumas]

Great job friend....keep the great work

[@MihaiDesigns]

Thank you! Cheers!

[@Doogleraia]

I see further potential for even larger beds, belt-beds, or even chamber. But One thing that would be a very interesting evolution of it for a "v2" (or whatever name that sounds more fancy..), taken directly from tool-changers from CNC machines, top-mounted tool-changer on a rotating magazine, so the print head goes up to it and the magazine pulls out the current tool into an empty set spot for that given tool, and then the next tool only moves down to slot into the head/mount and back down we go again for more printing fun! To further build on that style of system, with a top-mounted magazine like that, the tool heads could even be automatically z-tested in the magazine to calibrate them up there prior to being put into the slot for printing, only needing to reference machine z-offset after that.

[@LuisGarciaportillo]

I sincerely missed the "stay awesome" at the end of your videos. But this beast is more awesome than I can describe

[@MihaiDesigns]

Thank you 😀

[@henninghoefer]

Wow, what a project! Not only is your design awesome, but also the Fusion360 modelling is top-notch. For the tool changer I'd suggest to make all slots active slots and just do a temperature control multiplexed for all hotends (probably at a lower standby temp). Yes, this might need a few more MOSFETs, but overall I think it's way easier to put every hotend always back in its same slot and not having to deal with tangled tubes or shuffling around hotends between slots. For control I'd suggest a setup similar to how most Biqu displays work: They act as a separate controller using special firmware and can pre-process the G-code they read from their own SD card reader, which is then sent to the main board running Marlin via internal connection.

[@MihaiDesigns]

Thanks! I find the tube untangling to be a fun challenge to solve. If it doesn't work, then you are right that I can fall back to running all active slots and figure out the custom hardware for that.

[@florinlazar3190]

@@MihaiDesigns maybe use klipper as an firmware, with it you can add more then one controller board and heave all the heaters you want, also with some python scripts and klipper macro you can solve tool selection.

[@MADEIN3DPRINTING]

Very informational video and great job👌💜

[@MihaiDesigns]

Thanks for watching

[@user-en2gh9fs1f]

Still checking back weekly for updates on this!

[@flightradarfinds]

Awsome job wow

[@surajbiswas8642]

Nice wonderful, As expecting Last time at the end of hotend video.

[@nikscha]

Very impressive! At minute 20 or so you mentioned the problem of preheating. I suggest to just have a single connector located in the arm. So whenever a certain tool needs preheating the arm can just dock against that tool in advance and power it up.

[@DennisDay58]

Using the SPACEX method of failures and redo again is very smart. I love your work even though my brain hurts trying to follow along but keep it up.

[@SLcompany_watch]

Can't wait to see this awesome project actually working in real life

[@MihaiDesigns]

Oh, me too! There is more work to do before it's all ready, but I'm hoping to have it functional by the end of the year.

[@scruffy3121]

Very nice project. You might want to place a strain relief bore in the back of the Z adjustment slits.

[@uprisingsnake7152]

Great Project. Try to build a simple 1:1 mockup as soon as possible to prevent any flaws overseen. Please check your center of gravity of your construction loaded with 10+kg of spools. In front of the Screen it looks dangerously

[@hayden9944]

This is cool. If you are pre-heating ur nozzles - perhaps an ooze sheild like the ones seen on the snapmaker J1.

[@messinger123]

Brilliant.

[@santiagoblandon3022]

Mindblowing! The attention to detail is really impressive!. Autodesk should be sponsoring you >=(

[@I3enjiman]

Dang, what a project! I really feel you wanting to do everything from ground up. Even though it might take a little longer and be quite intricate, you can make exactly what you want and how you want, all while learning a heck of a lot. I really hope you manage to achieve all the goals you set for this project and I'm very excited to see where it goes! Stay awesome!

[@MihaiDesigns]

Thanks for watching! Yes, that is my strategy. I'll post more once I have updates.

[@MAmosov]

Great job! My opinion is to upgraid the core xy or simular highspeed printers that allows to use modern boards and pressure advance properly.

[@danifem]

amazing

[@Yeltrah]

Very interesting

[@yehudabillet8145]

The nfc tag system is actually genius

[@arismegas3732]

Wow!!! This is amazing👍 I really like the idea to use nfc tags🔖 I think you should make the station a cylinder so you can rotate it. That will enable you to have a smaller setup and include more tools

[@MihaiDesigns]

I did consider a large revolving unit with all the tools, but there is a risk there with filaments snapping after being twisted a little bit on every tool change. Good point though. I might make a prototype of that one day to either prove or disprove my assumptions

[@kyleyoung2464]

Briliant.

[@amazac9869]

Outstanding CAD work ,have a more simple idea ,that is ; place all your extruders at the top X gantry and place a Z slider above slides on a rail that will push the new extruder downwards (latching) -Z Z, you need another input for controlling the new ABOVE X axis ,powering the hot ends can be done by SSR sharing same ground with an adequate power . Good Luck.

[@ericlotze7724]

7:31-7:35 Not that many models…YET This printer design is seriously going to lead to some interesting stuff!

[@mattweger437]

Nice, good luck

[@cleisonarmandomanriqueagui9176]

Amazing video ! very very impressive . Could you share with us your progress on the firmware making ? from scratch as you said

[@spankdag2948]

tare, Mihai

[@modmike4406]

Amazing project! You could use a Beacon probe permanently in place. Its higher than the nozzle and solid state.

[@dgfpv1988]

not finished watching yet but that's a masterpiece

[@JM-ep7sj]

"Stay hungry, stay foolish", you are. I wish new generation IT hipster had 1% of your attitude. Amazing job

[@KK3DTechnology]

After very long time... 👍🏻😍❤

[@MihaiDesigns]

Indeed. I wanted to get this design ready to build and I postponed video production for this time. Now I'm back :D

[@NicksStuff]

Seeing a working prototype would have helped a lot

[@maximuschapman7760]

I love you're perfectionism. I would buy a printer like this for $1000 for my print farm.

[@christianmarkussen6412]

What a great project. I would consider using Klipper for both the printer and the tool changer. Both controllers can be connected to the RPI and you can drive the tool changing function through Kipper macros with perfect synchronization. Many other tool changer do this.

[@MihaiDesigns]

There's an advantage in running motion asynchronously from the printer, otherwise it can't do any move while the printer is printing. And I need a custom interface. So even though your solution would work, I'm willing to try a different approach.

[@arturstanczuk5448]

@@MihaiDesigns in short the klipper ecosystem looks like this today: mainsail/fluid (web user interface) -> Moonraker (API Web Server for Klipper) -> klipper (3D-Printer firmware). For the needs of your project, we only need a klipper - to which you need to add your own custom interface. Klipper provides: multiple micro-controllers, multiple extruders, multi-device support (kinematic, sensors, sensors, stepper motor drivers, etc), that's why there are so many comments supporting the selection of a klipper :)

[@add859tankionline]

Awesome model I am in awe of this CAD detail!! but all of this work to just change the colour. I'm sure the future will involve some dye that can mix into one master filament. What will be important is to change filaments that require different temperatures on the fly.

[@billyjoe3309]

FINALLY!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

[@mrfxm55]

I don't know anything about any of this but it's very impressive like having an industrial grade factory in your home garage. I suppose a radial printer is on the horizon.

[@the_futur1st]

Hi Mihai. First of all, awesome project, as a little proposal, please consider CPAP cooling system(Vez used this on his VzBots), it is adding a lot smaller weight to the printing carriage that any other.

[@ThePhilbox]

Mihai!!! Im still going down the extruder wood grain rabbit hole.... I found with my ASX1 it was the little plastic gear on the stepper that prevented the wood grain. Changing to metal I got wood grain. Also tried a variety of big and small gears there. Different combos produced wood grain or not. I ended up with a mellow (Ali) large gear) and random plastic small gear. Still more to explore there.

[@MihaiDesigns]

Wow interestig insight.

[@zoidbergVII]

A small custom PCB could be designed for heating the extruders and use CAN. Then you could use as many or as few of them as you wanted and have software control of the extruders. It is a relatively simple/low cost PCB to do this.

[@MihaiDesigns]

That could be done indeed. I wanted to avoid anything custom for now, but in the future I'm also considering custom PCBs

[@rickfischer5116]

Looks like someone is ready to make a how to video on fusion 360 that's amazing work!

[@armoth6958]

Awesome just awesome! Love your vids. This is mind blowing, can’t imagine how much time this must’ve taken but your brilliance is clearly evident. What about core xy? (I can see how this would be difficult it it seems to be the future of you want to print faster and have good quality. Would love to hear your view?

[@MihaiDesigns]

I already have plans for a CoreXY, some time early next year most likely.

[@ericthecyclist]

@@MihaiDesigns A corexy with tool changer?

[@Da-Real-Gigachad]

Nice

[@xisop]

Really impressive. Mind is blown 😄 Maybe it's an option to make all the parking bays active. Park all hotends in their own spot at all time so the tubes won't get tangled. Add relais to the power of the parked pays so they can preheat when the gcode trigger is met. There are lots of 24V SFF relais

[@MihaiDesigns]

Thanks! All active is a recurring theme and I do get the reasons for it. It's a fallback solution if the current one proves to not be as easy or as reliable as I'd like. Personally I find it a fun challenge to solve the untangling and is why I wanted to go with it, but I'm ready to discard it if it doesn't work perfectly.

[@vizionthing]

Rather than risk tangling the reverse bowden's, consider using Pi Pico's to control the preheating of each nozzle in its normal storage location, Pico's cost 1/10 the price of a normal Raspberry Pi and are perfectly capable of controlling two extruders, for power switching very cheap and small MOSFET boards are available that can handle the current needed for 40W hotends, each pico can be configured to run on an SPI bus as a slave so a single bus can be used to link them together with minimal wiring. Outstanding design, looking forward to seeing this progress

[@wjsbruwer8206]

Wow, my goodness you did a great job! I was wondering, if this printer fully works in the end, would you ever sell kits of the printer or would you just have like fully build instructions of the printer? Because I think when you get this to fully work, a lot of people would like this printer :3

[@MihaiDesigns]

Ideally we'll have both full build instructions with BOM as well as pre-assembled units that come tested and tuned.

[@wjsbruwer8206]

​@@MihaiDesignsNoice, I will defiantly be ordering one when available :3 And around what price range will the printed be? I guess probably around 700 to 1000 dollars?

[@MihaiDesigns]

@@wjsbruwer8206 The project requires a lot of parts, unfortunately. For this money, you could perhaps source about half of them off the shelf for DIY 🙁. I'll be looking into better sourcing options later on because as it stands, only small businesses would afford it.

[@davincifpv3842]

@@MihaiDesigns I realy like your printer and would be interested of buying one? Would it be possible that you sell a kind with all the non 3D-Printed parts? And then you can 3D-print the parts at home and fully asemble it by your self? Would be a price of around 2k$ possible for that? Or is it way to cheep?

[@MihaiDesigns]

@@davincifpv3842 it's way too early for me to know a price yet because the design is not even finished, so I don't know if that price can be achieved. I'll know more as the project progresses. Cheers!