The Voron 2.4 350 Carbon Filter That Doubles as a Chamber Heater
We calculate the final price from your selected file. You review and accept it before payment.
Saved on this device.
External carbon filter for the Voron 2.4 350: tool-free cartridge, no draft on prints, and it warms the chamber.
- Rating
- 0 / 5
- Likes
- 29
- Downloads
- 33
- Makes
- 0
- Views
- 710
- Reviews
- 0
Updated at source:
Model information
- Source
- Printables
- Category
- Voron Parts & Upgrades
- License
- GNU General Public License v3.0
An external carbon filter for the Voron 2.4 350 (LDO kit) that also works as a chamber heater. Most carbon filters sit inside the chamber, where the heat slowly cooks the carbon and shortens its life. This one mounts on the outside of the rear panel, so the carbon stays cool. The filtered air goes back into the chamber near the deck, or under the heated bed with the optional diffuser.
Warms the chamber: +6–8 °C measured, from about 53 °C to 60 °C at 50% fan speed (ASA, 245 °C nozzle, 100 °C bed).
No draft on the print: the air comes back in low, not over the part.
Tool-free: the cartridge and the rear panel with the return duct both come off without tools, like my Panel Glide System.
Tested with real data: I logged VOC levels with sensors inside and outside the enclosure over months of ASA printing (see the end of this page).
How the air flows
Hot air is drawn from the top of the chamber into the filter.
It passes through the carbon first. The carbon sits before the fan, so the fan only moves clean air.
The blower pushes the air down the external return duct.
The air comes back in through the deck inlet, near motor B.
From there it rises into the chamber, or passes under the heated bed first if you use the under-bed diffuser.

Why it heats the chamber
Without the diffuser: in a closed enclosure the hot air collects at the top. The fan pulls it from the top and brings it back in at the bottom, so the whole chamber ends up warmer and more even. The fan doesn't add heat, it moves heat that would otherwise sit unused at the top.
With the under-bed diffuser: the air passes under the bed, picks up heat from the underside (100–110 °C for ASA) and then rises into the chamber. This works as a real heat exchanger, and it's where the +6–8 °C comes from.

Use the Klipper macro
Because the fan both filters and heats, running it at 100% for a long print can push the chamber well above 60 °C. That's too hot for sharp corners and small features, and I don't want to overheat the electronics either. The macro in the PDF Carbon_Filter_Chamber_Temperature_Auto-Control_V3 handles it in three phases:
Heat-up: fan at 50% while the bed heats, to spread the heat around the chamber.
Printing: it checks the chamber temperature every 10 seconds. Fan at 100% up to 58 °C, down to 40% at 62 °C, and a smooth change in between. It never stops completely, so the filter keeps working when the fumes are at their peak.
After the print: 5 minutes at 80% to clear the air, then off.
You need a chamber temperature sensor in Klipper. The PDF has the full macro, the two lines to change (fan pin and sensor name), where to call it from PRINT_START and END_PRINT, and troubleshooting. I strongly recommend it, especially with the under-bed diffuser.
Files and materials
File | What it is | Material |
|---|---|---|
carbon_filter_body | The removable cartridge. Holds about 370 ml of activated carbon. | ASA |
fill_plug_TPU | Plug on the cartridge. Pull it out to pour in fresh carbon. | TPU 95A |
mesh_retaining_ring | Holds the mesh screen in place. | ASA |
printed_mesh_alternative | Printed screen, if you don't have stainless steel mesh. Print it with no top or bottom layers and 20% gyroid infill. | ASA |
filter_flange | Flange around the air intake on the rear panel. | ASA |
flange_support | Support on the rear panel that the cartridge slides into. | ASA |
flange_gasket_TPU | Seals the flange against the panel. | TPU 95A |
fan_gasket_TPU | Seals the blower fan. | TPU 95A |
fan_duct_coupler_TPU | Connects the fan outlet to the return duct. | TPU 95A |
retur_duct_upper / retur_duct_lower | The external return duct, in two pieces that press fit together. | ASA |
deck_inlet | Sits on the deck panel and takes the air from the return duct back into the chamber. | ASA |
under-bed _diffuser | Optional. Spreads the air under the heated bed. | ASA |
drilling_template | Full-size reference of the rear panel. Print only the sections you need. | any |
The two PDFs: Back_panel_cutout (all cut-outs and hole positions) and Carbon_Filter_Chamber_Temperature_Auto-Control_V3 (the macro).
Hardware
Stainless steel woven mesh (optional, or use the printed screen)
3 × M4 heat-set inserts (fan mount) and 12 × M3 heat-set inserts
M3 and M4 screws
About screw lengths: I can't list an exact length for every screw, sorry about that. The easiest way is to keep an M3/M4 screw assortment on hand, or buy long screws and cut them to length. If you find the right length for a part, share it in the comments to help others.
Build steps
Print the parts from the table above.
Mark and drill the rear panel. Print the section of the drilling template you need and check that it matches your panel, since printer calibration can scale it slightly. Drill all intake holes at 2 mm first, then open them to 6 mm. Drill from the outer face, at low to medium speed with light pressure: acrylic (LDO kits) or printed ABS/ASA panels chip if you force it. Deburr lightly.
Cut the opening for the return duct in the lower rear panel. Positions are in Back_panel_cutout.pdf.
Install the heat-set inserts: 3 × M4 for the fan, 12 × M3 in the ASA parts.
Make the screen. Cut the stainless steel mesh into a circle with tin snips and fix the retaining ring on top with two drops of CA glue. The printed screen goes in the same way, with CA glue or acetone.
Join the return duct. The upper and lower parts press fit together, no glue needed. Use acetone if you want a permanent bond.
Mount the flange support on the rear panel and slide the cartridge in.
Place the deck inlet on the deck panel and, if you want it, slip the under-bed diffuser onto it. Both just rest on the deck, no screws into the frame.
Check the Z belt guard. If it touches the diffuser, file it a little until it clears.
Wire the fan to a fan output on your mainboard and set up the Klipper macro.
Changing the carbon
To refill you don't need to remove the cartridge: pull out the TPU plug, pour in the carbon (about 370 ml) and put the plug back.
To empty it, pull the cartridge up and out of its support, empty it, refill it while it's out, and slide it back in.
How often depends on how much you print. In my garage there's no smell at all, so I just change the carbon from time to time.
Other printers and frame sizes
The return duct is sized for the 350 mm frame. For other sizes or for double panels (DoomCube), or to cut a panel for another printer, I've uploaded STEP files for the five parts that touch the panels: retur_duct_upper, retur_duct_lower, deck_inlet, under-bed_diffuser and drilling_template. The filter body doesn't touch the panels, so it shouldn't need changes. I don't have time to support conversions, but the STEP files should be enough to adapt the ducts.
Works well with
On my printer this filter runs together with three other mods. They all aim at the same thing, a well-sealed chamber you can open without tools, so the filter only moves the chamber air:
Panel Glide System: gravity-retained TPU panels, tool-free access and a complete seal.
LDO Deck Seal with Umbilical Clip: seals the deck and blocks debris.
Rear Panel Seal & Filament Passthrough: TPU rear seal and PTFE passthrough.
Optional: VOC monitoring
To check that the filter really works, I added real-time VOC monitoring during ASA prints. It's a separate electronics project:
Sensors: 2 × BME280 (temperature and humidity) and 2 × SGP41 (VOC), one pair in the chamber and one in the room.
Electronics: a Raspberry Pi Pico 2 W running MicroPython, connected by USB to the Voron's Raspberry Pi. Data goes through MQTT (Mosquitto) to Moonraker and to a custom web dashboard.
What it shows: VOC levels (Sensirion Gas Index), temperature and humidity in the chamber and in the room, filter efficiency, and a 24-hour trend to see when the carbon wears out.
What I learned: fan speed matters more than how fresh the carbon is. At 30% fan speed the filter couldn't keep up with ASA. At 100%, the filter delta went from consistently negative to +60–100%. The carbon wasn't saturated, it just wasn't getting enough air. That's why the macro never drops the fan below 40% during a print.
For the VOC version I modified the filter housing to fit the four sensors and the Pico 2 W, and made a new cable with all the wires. It's too complicated to document properly, so I'm not publishing it. If you're building something similar, ask in the comments and I'll share what I can.
Changelog
2026-09-15: Added STEP files for the five parts that touch the panels (for DoomCube and other printers).
2026-04-23: Klipper macro V3, with proportional fan control.
2026-04-05: Added the back panel cut-out PDF.
About this description: I design, print and test everything myself. I'm a hands-on maker, not a writer, so I use AI (Claude, by Anthropic) to help me organize and polish the text. The Klipper macro and the VOC dashboard software were also written with Claude's help.