Jump to content

My First Voron - V0.2r1 (Formbot Kit)


Recommended Posts


I was one of the many Kickstarters who bought the Creality CR-6 SE in 2020 as their first 3D printer. However, I am probably one of the rarer individuals who only unboxed and started printing with it 2 months ago. After that was upgraded with Octoprint, modded firmware, PEI, dual gear extruder, a new toolhead and various other quality of life improvements, the time came to move this new hobby obsession to the next level. £430 (£70 import tax) lighter, and two boxes arrived to the UK from China from Formbot within a week via Fedex. I am hoping to use the experiences here for a Voron v2.4r2 350mm in a few months time, as well as use this for v2.4's functional parts in ABS.


The Build:

Preparation / Resources:

I have been closely following @Perkeo's build, trying to absorb everything.

Dmason's V0.1 blog has also been useful: https://3dpandme.com/category/printer-builds/v0-1/

Ballistic Tech has a vlog on this kit here: 


I am pretty happy with electronics, linux command line and mechanicals. Although had to purchase some crimpers and a couple of 123 blocks for the build. Additional things not included in the Formbot kit: Lithium EP2 grease, Isopropanol Alcohol 99.9%, threadlocker, various connectors (Wagos, XT60, JST).

Voron Discord Profile: Lik#5169

Edited by indi8six
  • Like 1
Link to comment
Share on other sites

The Formbot Kit:

ABS parts come in a separate box with phenomenally good packaging. Each part has it's own customised cut-out.


The quality of the parts themselves was impressive. No warping and uniform adhesion.



The larger box contained a couple of layers. The top layer comprised of loose bags of fasteners, motion systems and wiring.


The second lower layer contained customised cut-out sections for the extrusions, motors, electronics. Despite these cut-outs, some of the extrusions managed to free themselves in transit and scratched up. I will try and hide tactically hides these on the build! The 1515 extrusions are anodized after being machined/cut. Whilst the Formbot BOM states "Makerbeam XL", this is not actually the case. The profile is similar to the LDO 1515 extrusions ("D" channel rather than Makerbeam's "T" channel).


The electronics included in the kit: BTT Klipper Pi (v1.2), BTT SKR Pico (v1.0), genuine Phaetus Dragon SF, Meanwell-branded PSU and Moons steppers. The BMG gears seems quite generic and might be a target for a future upgrade.


Overall I am very happy with the quality of the kit and the USD$60 outlay for all the ABS parts (functional and cosmetic) seemed a good decision. Trying to print ABS in my open-air Creality bedslinger or causing a fire-hazard in a cardboard enclosure would not have been worth the hassle!

  • Like 5
Link to comment
Share on other sites

Good luck with your build. Really enjoyed my building process, but still need to print a proper voron cube. Both part cooling fans died on me in the first hour, so awaiting the arrival of some Honey Badger replacement fans from 3DO. You may want to consider replacing these fans already in your initial build.

I will follow you with your build! 👍

  • Like 1
Link to comment
Share on other sites

Welcome to the community - happy to have you on board.

I am sure you will enjoy this build immensely. 

Love you list of mods - go canbus from the start - less wires 😄

Happy building, modding and printing.

  • Like 1
Link to comment
Share on other sites

Finally made a start and have been at it daily for the past week


I stole the idea of labelling up the extrusions for easier picking later!


As previously mentioned, the Formbot kit has "LDO" style 1515 extrusions with a D-shaped channel and not the Makerbeam T-shaped channel.


Thankfully this is good news as it means the No-Drops mod works out of the box. Unfortunately despite my best efforts of trying to print ABS on a bed-slinger (eSUN ABS+), there was still significant issues with warping and bed adhesion. I therefore printed them in PETG, knowing that any glass-transition would not affect this non-load bearing part what-so-ever. I did this in 4x plates of 30x drop outs and loaded up all 120.


I used the remixed HexTray made stackle and printed it out in vase mode in only an hour...


Final bit of preparation was to clean the rails. For this I printed off this linear rail bath, again in vase mode. I could have used the included bag but the bath made it easy to slide the rail up and down. I used mineral spirits for first pass and 99.99% IPA for the final clean. Lubrication was with this lithium EP2 grease applied via a luer-slip syringe.


  • Like 3
Link to comment
Share on other sites

The Frame

The rail centering tool in the original Voron repository was very useful for aligning the rail to the extrusion. I followed the Voron manual of alternating for bolts for the Z and Y rails. In hindsight, the kit has enough fasteners to complete all the rails. I think this would be overkill for Z but I would definitely populate the Y rails fully next time.


The printed linear rail nut holder from Formbot was so thin and frail that it snapped in a couple of the insertions. I suspect some of the parts are a little over-extruded.


The flattest thing I could find in the house was the induction hob. I hope Neff have their glass tolerances right! A couple of 123 blocks were also useful but in hindsight not crucial. I double and triple checked the pre-loaded nuts, but I also added more for Slim Handles, camera, Nevermore / MFNano, Din rail connectors, and Kirigami PCB.

In hindsight I should have loaded up some for this PSU mount, but some thin square M3 nuts are on their way from Aliexpress which ought to drop in. Also I have since modified the Kirigami PCB so don't need to use the mount for that (see future post).


I had to slightly straighten the Kirigami bed from stock, but only very minimally. I also had to re-align the Z-axis linear rails to the extrusion so that both were perfectly parallel and dropped under it's own weight. Probably unnecessary but more for my OCD.


Edited by indi8six
  • Like 2
Link to comment
Share on other sites

The Motion System

The A/B drives were a doddle. Thankfully I never dropped any of the stacks, but I did use a temporary M3 nut for this exact reason. One issue is that the shims supplied with the kit had thicknesses ranging from single Macdonald's Happy Meal Cheeseburger (0.4mm) to Burger King's Double Whopper (0.6mm). A bit of trial and error was required to have even stack heights.



The X Axis

The X rail was a supplied as a specific preloaded rail by Formbot, marked "light preload". Whilst only "light", the difference was very noticeable versus the other rails. All bolts were used for mounting this, with TL-242 threadlocker. In fact, I used thread locker for all the frame extrusions and rail mounts.

I was very surprised and happy to find that there was no sticking or resistance on the Y-rails. The squaring paid off!


The Belts

The Gates-branded belt was supplied as a single 3m length. I used the method from Ballistic Tech's build series i.e. a dry run through one side and I cut at 10cm overlap. I then paired this for the second length for a matching pair. With both belts fed and tightened to the same leftover length, I made the magic 110hz. I later found out that I would need to trim these flush to mount the Dragon hotend.



My current final progress is that I have made it to the electronics and will be flashing firmware tomorrow. However, I am awaiting a modified version of the Kirigami PCB from JLCPCB. Perhaps I'll post the updates anyway but right now I need a whiskey.

Edited by indi8six
  • Like 3
Link to comment
Share on other sites

I have now been printing for a while now, so a big update is due! BallisticTech has also recently released an unofficial guide to the Formbot 0.2r1 kit which may help future builders. A lot of the info there is gained through his video build series.



The Toolhead

The vast majority of this went together very smoothly. However some intolerances within the printed parts meant a bit of handy work was required with a 3mm drill to make sure the hinge screws spun freely. It has been mentioned that Formbot may not be printing to Voron's official spec (0.4mm Nozzle, 4x walls). However, I haven't found any issues other than this and the filament sensor (see later!). Formbot uses Phaetus aeWorthy™ ABS-GF, in other words ABS reinforced with 10% glass fibre, which would account for the fuzzy texture of the parts.

IMG_1044.jpg.410b8abd03ad7d91361a03eb149cea08.jpg IMG_1045.jpg.b042ff23136a8acfd127c53607603fd9.jpg

The guidler's pin socket was very tight so a bit of slow and manual reaming was required (slow as to avoid enlarging the socket too much). I also sized the 4x2mm bowden tube using the V0 "swiss army jig" multi-tool.

IMG_1043.jpg.5a3c1d1ea1b82df8ca32a53dc6494057.jpg IMG_1046.jpg.29694c587341843204b61c4bede048d4.jpg

The V0.2 Umbilical toolhead mount was not included in the Formbot kit, surprisingly. Again, I used PETG for now, printing in ABS later.





Edited by indi8six
  • Like 1
Link to comment
Share on other sites

The Electronics Bay

I didn't fancy using VHB tape to mount the electronics and I also kept mobility of the boards in mind. Similarly the PSU was mounted using this adapter rather than tape. I had originally planned to use this DIN rail mount but the push-fit was so tight that I snapped the part. Instead I used this instead with printed M3 washers. I also needed an two extra fans for the MFNano and the bed fan, so a Klipper Expander was required. I could potentially have used Canbus (BTT EBB 36) which would free up a similar amount of ports, but the expander was cheaper and I wanted to use the stock configuration with the option of upgrading later. I modified this Klipper Expander duct, removing the central pillars to fit the DIN rail.

IMG_1056.jpg.f244e5107926d8ca95198f54e5fe9de5.jpg IMG_1057.jpg.9a1d6bf4d4616b2864a5de3ef77f9793.jpg

To save USB space (expander, display, canbus port etc) and optimise the link, I decided to connect the BTT Klipper Pi to the SKR Pico via UART. This required a new 5 pin JST connector which I wrapped in matching braided sleeve. This cable also powers the Pi from the Pico.


Flashing the firmware wasn't too bad. The process was roughly as follows:

  • Image BTT Klipper Pi. I used the full image, rather than the minimal image, as it included the GPU acceleration by default. I gave this a static IP and hostname via my router's DHCP service.
  • I set the BoardEnv.txt to console=serial to enable UART via /dev/ttyS0.
  • I updated Debian using Putty and updated the rest of Mainsail and it's plugins via the web interface .
  • Flash BTT SKR Pico. I used the PC method as detailed in the Voron documentation. I used serial interface instead of USB.
  • I connected the Pi to the MCU with the homemade UART cable.
  • Flash Klipper Expander: I used the default Github method with the module attached to the Pi via USB.
  • Flash Voron Display: Again, a very easy to follow guide on Github, via USB.
  • Configuration Files: The default Voron Klipper configuration files were pretty self-explanatory. I separated out the macros into macro.cfg with everything (klipperexpander.cfg, pico.cfg, macro.cfg) individually called from printer.cfg. /dev/ttyS0 was used as the Pico UART device (NB: Official Raspberry Pi would use /dev/ttyAMA0).

Here it is all connected prior to the Klipper Expander (i.e. Formbot stock) with the homemade UART cable.


It took a couple of hours to run through Ellis' Tuning Guide, and with that, she was finished and printing the first mod almost perfectly.

IMG_1069.jpg.783cbc67686210042ddb8fdcd1a031cb.jpg IMG_1075.jpg.15fcda34f6a46d1f8cc81d99a29bc137.jpg

Edited by indi8six
  • Like 1
Link to comment
Share on other sites

First Prints / Mods

The printer is lightweight but awkward to pick up and so the first print were Stealth Handles. I had pre-loaded extra nuts for these handles, but ultimately preferred the more subtle look. Using eSUN ABS+, I settled on 230C hotend and 105C bed (240C and 110C first layer). I let chamber reach 40-50C which meant I could use the part cooling fan between 40-80% depending on the shape of the print (overhands, corners, size etc). Pressure advance 0.028 and flow 97%.

To help with moving the machine, I printed this lockable spool holder, using a spare extruder spring for the mechanism.


During this time, I noticed that the filament sensor was unpredictable and very dependent on the bend of the filament itself (or the diameter tolerance of the filament). Annoyingly, this caused a lot of paused prints. The micro-switch seemed a bit of an awkward fit on the Formbot part so I decided to reprint the entire part. A genuine Omron switch is in the post to replace the generic bundled switch, but thankfully for now no further false-readings appear to be happening. Ironically, an initial unattended first run of the filament sensor failed due to another filament sensor mis-read (the bed got too cold to resume the part). I turned the sensor off to complete the finished print (!!).



Kirigami PCB / Bed-Fan / Neopixels

The lengthy heating time of the chamber annoyed me, so I used a hairdryer to boost the chamber. This was not acceptable for the rest of the family (all female) but luckily I spotted the recent release of the Kirigami PCB. I didn't think a full umbilical + umbilical PCB was particularly helpful as it would be running through the cable chain. It would also have required Molex connectors which I didn't have. I therefore modified the KiCad files for a customised solution - a plug in breakout board attached underneath the bed with a 30mm bed fan and Neopixel point. This only requires JST connectors and 2-pin screw terminals. The screw terminals allow me to get rid of the Wagos and the JSTs let me run wires from the chain to the PCB, and from the PCB to the parts around the Kirigami (thermistor, heater, thermal fuse, bed fan, Neopixel). Not a compulsory upgrade but very convenient. The main goal was the fan!


I ordered a minimum quantity through JLCPCB so I have 3 boards remaining if anyone is interested.

Edited by indi8six
  • Like 2
Link to comment
Share on other sites

2 hours ago, indi8six said:

During this time, I noticed that the filament sensor was unpredictable

The Filament Runout Sensor Foot has a huge design flaw. Sorry for the Voron Team, but they failed at designing it. They considered that an internal right angle can be printed sharp. Of course, such an angle does not exist in the wild. As a result, the switch is not seated properly.

I had to redesign and it now works properly.

Also, there's not enough meat for the screws, and the tie strap is on the wrong side.

Left : Voron design. Right : better design.



  • Like 3
Link to comment
Share on other sites

Are you interested by this remix ?

I wasn't planing to publish mods too soon because most of them were flawed, unfinished or even pointless (like 90%), but this one has been in use for 80 hours without any problem. Still have to create a series of go nogo test prints for choosing the right model for the pneufit ; the Voron one didn't work well, despite a perfectly tuned machine. (for example wall printing oreder could make a difference...)

Could upload to Github within 2 or 3 days. Let me know.

Link to comment
Share on other sites

  • 4 weeks later...
  • 1 month later...


I went for the Rainbow On A Matchstick (RGB) mod (https://github.com/VoronDesign/Voron-Hardware/tree/master/Daylight/Rainbow_on_a_matchstick), mounted under each Y-axis gantry. Most of my prints are quite short in height and I was happy keeping the lighting at the top, plus the lighting position paired well with my camera mount.

The connection for this was via my Klipper Expander and Umbilical PCB.

[neopixel Matchstick]
pin: expander:PB1
color_order: GRB
chain_count: 20

I used this wiring diagram albeit with Strip 3 unused and Strip 2 without the blue output wire. https://github.com/VoronDesign/Voron-Hardware/blob/master/V0-Umbilical/Images/FramePCB_PixelWiring.png





Link to comment
Share on other sites


I used a OV2659 module (120 deg FOV) as my camera, connected directly to the BTT Pi via USB. I went through a few mounting options but finally settled on the PanzerObserver system, mounted at the top of . I used a mount for 1515 extrusion available from Printables, but hartk1213 also modified one. Finally I used this modified cover as the OV2659 doesn't have a flash. The picture below is taken with the tophat open, the mount is actually behind the extrusion (not below) and barely noticeable from the front.



The 120 degree field of view perfectly captures the build area and is set with manual focus.


Edited by indi8six
  • Like 2
Link to comment
Share on other sites


Due to very minor bed levelling issues and laziness, I decided to install some form of auto-bed-levelling. I went for ZeroClick due to it's simple design and preserved print-area versus Klicky. This also allowed me have an almost perfect Z-offset as Z-homing was taken care of using this method. I used Klicky's macro scripts and my config's probe's z_offset is set at 9.400 (Dragon SF with standard Phaetus brass nozzle, printing ABS in a heated chamber). I simply run -0.05 to -0.07 Z-offset for PLA in a cold chamber.


I used ZeroClick's rear mount so that I could still mount the MFNano at the back. However, I found that the X-gantry would interact with it and interfere with sensorless homing. Instead of re-designing the part, I filed down 10mm x 5mm from the rear mount so that the X-gantry could still home all the way and leave enough room for the docking/undocking of the probe.


Due to lack of mini-SB mount, I used a modified mini-SB cowl but inlaid it with black for the Voron logo. The electrical feed is through the umbilical via the X-endstop connections.

Overall it took quite some time to have everything set up perfectly, but the quality of life benefits since then have been worth it. Here is my full klicky-specific.cfg

pin: ^gpio22
x_offset: -19
y_offset: 14
z_offset: 9.400
speed: 3
lift_speed: 7
samples: 5
samples_result: median
sample_retract_dist: 2
samples_tolerance: 0.01
samples_tolerance_retries: 10
#drop_first_result: true

speed: 100
horizontal_move_z: 20
mesh_min: 15,17
mesh_max: 100,105
algorithm: lagrange
probe_count: 5,5
zero_reference_position: 80,50
# relative_reference_index: 4
fade_start: 1
fade_end: 10
fade_target: 0
move_check_distance: 3
split_delta_z: 0.0125
mesh_pps: 2,2
#bicubic_tension: 0.2

screw1: 79,0             #For Long probe
screw1_name: front screw
screw2: 120,100          #For Long probe
screw2_name: back right
screw3: 26,100            #For Long probe
screw3_name: back left
horizontal_move_z: 20
speed: 100
screw_thread: CW-M3
Edited by indi8six
  • Like 2
Link to comment
Share on other sites


I have the V0 often on the same desk as my workspace and so it was important to have some sort of VOC filtering. I settled with the MFNano which fitted very stealthily in black at the back without interfering with the toolhead. The Bento Mini alternative seemed a bit too bulky for me.



I hooked this up to the Klipper Expander's mosfet outputs:

[fan_generic MF_Nano]
pin: expander:PA3
cycle_time: 0.010
shutdown_speed: 0
kick_start_time: 0.5

So far, the ABS smells seem to have been greatly reduced.

  • Like 1
Link to comment
Share on other sites

Kirigami Upgrades

The aluminium Kirigami bed which came with the Formbot kit was actually pretty decent. No warping and pretty good shape. I did not encounter any of the issues which Brian did in his build videos. The only downside was that there was some minor flex, which occasionally led to adhesion issues at the very front of the bed. I also wanted a more sturdy plate for ZeroClick to do it's job. I bit the bullet and bought a steel version from AE (Siboor) and sold my aluminium version on eBay for a cost-free upgrade. The steel version, weighing in at 210g, is so much more robust and barely flexes. Ultimately due to the cantilever design, it would be preferable if all Kirigami beds were made out of steel.


I also mentioned the Kirigami PCB which I modified, above. Changing to the steel Kirigami gave me an opportunity to take some photos of the PCB in place.


I run the bed fan using one of Klipper Expander's mosfet ports. Heating the bed to 110C with the stock 60W heater and bed fan set gradually from 15% to 25% gets my umbilical's thermistor 20 Celsius above ambient in about 20 minutes. Totally worth it for the huge reduction in heat soak time.


Finally I am running Maple Leaf Maker's Stealth Bumper with the Neopixel connection to the BTT Pico. My hope is to add the adaptive LED status macros.


  • Like 3
Link to comment
Share on other sites

16 hours ago, indi8six said:


I went for the Rainbow On A Matchstick (RGB) mod (https://github.com/VoronDesign/Voron-Hardware/tree/master/Daylight/Rainbow_on_a_matchstick), mounted under each Y-axis gantry. Most of my prints are quite short in height and I was happy keeping the lighting at the top, plus the lighting position paired well with my camera mount.

The connection for this was via my Klipper Expander and Umbilical PCB.

[neopixel Matchstick]
pin: expander:PB1
color_order: GRB
chain_count: 20

I used this wiring diagram albeit with Strip 3 unused and Strip 2 without the blue output wire. https://github.com/VoronDesign/Voron-Hardware/blob/master/V0-Umbilical/Images/FramePCB_PixelWiring.png


I have my Rainbow on a Matchstick strips connected the same way to the LDO Picobilical board. Only the three wires though and they work fine; might be that LDO eliminated the need for that wire in their updates to the PCB.. BTW, mounting them on the Y gantry extrusions keeps the lights pointed at the nozzle, which is really where you want them aimed anyway. 🙂

  • Like 2
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Create New...